2020 Book DevelopmentsAndAdvancesInDefen PDF

Smart Innovation, Systems and Technologies 152
Álvaro Rocha
Robson Pacheco Pereira Editors
Developments
and Advances
in Defense and Security
Proceedings of MICRADS 2019
Smart Innovation, Systems and Technologies
Volume 152
Series Editors
Robert J. Howlett, Bournemouth University and KES International,
Shoreham-by-sea, UK
Lakhmi C. Jain, Faculty of Engineering and Information Technology,
Centre for Artificial Intelligence, University of Technology Sydney,
Broadway, NSW, Australia
The Smart Innovation, Systems and Technologies book series encompasses the
topics of knowledge, intelligence, innovation and sustainability. The aim of the
series is to make available a platform for the publication of books on all aspects of
single and multi-disciplinary research on these themes in order to make the latest
results available in a readily-accessible form. Volumes on interdisciplinary research
combining two or more of these areas is particularly sought.
The series covers systems and paradigms that employ knowledge and intelligence
in a broad sense. Its scope is systems having embedded knowledge and intelligence,
which may be applied to the solution of world problems in industry, the environment
and the community. It also focusses on the knowledge-transfer methodologies and
innovation strategies employed to make this happen effectively. The combination of
intelligent systems tools and a broad range of applications introduces a need for a
synergy of disciplines from science, technology, business and the humanities. The
series will include conference proceedings, edited collections, monographs, hand-
books, reference books, and other relevant types of book in areas of science and
technology where smart systems and technologies can offer innovative solutions.
High quality content is an essential feature for all book proposals accepted for the
series. It is expected that editors of all accepted volumes will ensure that
contributions are subjected to an appropriate level of reviewing process and adhere
to KES quality principles.
** Indexing: The books of this series are submitted to ISI Proceedings,
EI-Compendex, SCOPUS, Google Scholar and Springerlink **
More information about this series at http://www.springer.com/series/8767

Álvaro Rocha Robson Pacheco Pereira
•
Editors
Developments and Advances

in Defense and Security
Proceedings of MICRADS 2019
123
Editors
Álvaro Rocha Robson Pacheco Pereira
Departamento de Engenharia Informática Military Engineering Institute
Universidade de Coimbra Rio de Janeiro, Brazil
Coimbra, Portugal
ISSN 2190-3018 ISSN 2190-3026 (electronic)

Smart Innovation, Systems and Technologies
ISBN 978-981-13-9154-5 ISBN 978-981-13-9155-2 (eBook)
https://doi.org/10.1007/978-981-13-9155-2
© Springer Nature Singapore Pte Ltd. 2020
This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part
of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations,
recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission
or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar
methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are exempt from
the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this
book are believed to be true and accurate at the date of publication. Neither the publisher nor the
authors or the editors give a warranty, expressed or implied, with respect to the material contained
herein or for any errors or omissions that may have been made. The publisher remains neutral with regard
to jurisdictional claims in published maps and institutional affiliations.
This Springer imprint is published by the registered company Springer Nature Singapore Pte Ltd.
The registered company address is: 152 Beach Road, #21-01/04 Gateway East, Singapore 189721,
Singapore
Organization
Honorary Chair
Álvaro Rocha, University of Coimbra, Portugal
Honorary Co-chair
Lakhmi C. Jain, University of Canberra, Australia
General Chairs
Robson Pacheco Pereira, Instituto Militar de Engenharia—IME/RJ, Brazil

José Avelino Moreira Victor, University Institute of Maia, Portugal
Local Organizing Chair
Luiz Eduardo Pizarro Borges, Instituto Militar de Engenharia—IME/RJ, Brazil

Luiz Augusto Cavalcante Moniz de Aragão Filho, Instituto Militar de Engenharia—
IME/RJ, Brazil
Filipe Silva Semaan, Universidade Federal Fluminense, Brazil
Bárbara Dias de Castro Pacheco, Universidade Federal Fluminense, Brazil
v
vi Organization
Scientific Committee
Abderrazak Sebaa, University of Bejaia, Algeria

Alex Fernando Jimenez Velez, Fuerza Aérea Ecuatoriana (FAE), Ecuador
Amita Nandal, UIST—University for Information Science and Technology
“St. Paul The Apostle”, Ohrid, Macedonia
Anacleto Correia, CINAV/Escola Naval, Portugal
Andrea D’Ambrogio, University of Rome Tor Vergata, Italy
Andrea Visconti, Università di Milano, Italy
Angelo Borzino, Brazilian Army Technological Center, Brazil
Angelo Brayner, Federal University of Ceará, Brazil
António Abreu, ISCAP/IPP, Portugal
Antonios Andreatos, Hellenic Air Force Academy, Greece
Asanka Pallewatta, University of Kelaniya, Sri Lanka
Bruno de Pinho Silveira, ESPE—University of the Armed Forces, Ecuador
Bruno Madeira, CTEx, Brazil
Calogero Vetro, University of Palermo, Italy
Chiara Braghin, Università degli Studi di Milano, Italy
Chinenye Ajibo, University of Nigeria, Nsukka, Nigeria
Damir Blazevic, Faculty of Electrical Engineering, Computer Science and
Information Technology Osijek, Croatia
Daniel Sampaio, University Institute of Maia, Portugal
Daniela Suzuki, Institute of Biomedical Engineering (IEB-UFSC), Brazil
Darwin Manolo Paredes Calderon, ESPE—University of the Armed Forces,
Ecuador
Derya Yiltas-Kaplan, Istanbul University, Turkey
Diana Patricia Arias Henao, Universidad Militar Nueva Granada, Colombia
Diego Paes de Andrade Peña, Universidade Federal do Maranhão, Brazil
Dimitrios Dalaklis, World Maritime University (WMU), Sweden
Eduardo Izycki, Presidência da República do Brasil, Brazil
Eduardo Siqueira Brick, Universidade Federal Fluminense—Instituto de Estudos
Estratégicos, Brazil
Emanuele Bellini, University of Florence, Italy
Enrique Carrera, ESPE—University of the Armed Forces, Ecuador
Fabian Ramirez Cabrales, Colombian Naval Academy “Almirante Padilla”,
Colombia
Felipe Torres Leite, Universidade do Estado do Rio Grande do Norte—UERN,
Brazil
Fernando Almeida, University of Campinas (UNICAMP), Brazil
Filipe Sá, Câmara Municipal de Penacova, Portugal
Flávio de Barros Vidal, University of Brasilia, Brazil
Franklin Johnny Dominguez Ruiz, Universidad da Coruna, Ecuador
Franklin Manuel Silva Monteros, Universidad de las Fuerzas Armadas—ESPE,
Ecuador
Organization vii
Gabriele Virzi’ Mariotti, University of Palermo, Italy

Galo Ricardo Andrade Daza, AGUENA Centro de Estudios Estratégicos de la
Armada, Ecuador
Geert de Cubber, Royal Military Academy, Belgium
Giovana Garrido, Universidad Tecnológica de Panamá, Panamá
Glauber Cruz, Federal University of Maranhão, Brazil
Guilherme Pussieldi, Universidade Federal de Viçosa—Campus Florestal, Brazil
Hachem Slimani, University of Bejaia, Algeria
Helio de Oliveira, Universidade Federal de Pernambuco, Brazil
Henrique Miguel Gouveia Silva, University Institute of Maia, Portugal
Henry Omar Cruz Carrillo, Centro de Investigacíon de Aplicaciones Militares—
CICTE, Ecuador
Hiroshika Premarathne, FCT, University of Kelaniya, Sri Lanka
Hugo Peixoto, Algoritmi Research Center, University of Minho, Portugal
Ioannis Chatzigiannakis, Sapienza University of Rome, Italy
Isabel Lopes, Instituto Politécnico de Bragança, Portugal
Ivan Machado, Federal University of Bahia, Brazil
Jaime Meza Hormaza, Universitat Politecnica de Catalunya, Spain
Jeimy Cano, Universidad del Rosario, Colombia
Jiahn-Horng Chen, National Taiwan Ocean University, Taiwan
João Reis, Portuguese Military Academy, Portugal
João Vidal de Carvalho, ISCAP/IPP, Portugal
José Antonio Apolinario Junior, Military Institute of Engineering (IME—SE/3),
Brazil
José Avelino Moreira Victor, University Institute of Maia, Portugal
José Carlos Dias Rouco, Portuguese Military Academy, Portugal
José Luís Reis, IPAM—Instituto Português de Marketing, Portugal
José Miguel de Carvalho Cerqueira, Instituto Universitário Militar/Academia
Militar/Swedish Defence University, Portugal
Jurij Mihelič, University of Ljubljana, Slovenia
Kalinka Kaloyanova, Sofia University, Bulgaria
Lúcio Agostinho Barreiros dos Santos, Instituto Universitário Militar, Portugal
Luis Alvarez Sabucedo, University of Vigo, Spain
Luis Dieulefait, Universitat de Barcelona, Spain
Luis Eduardo Palacios Aguirre, Fuerza Aerea Ecuatoriana, Ecuador
Luis Anido Rifón, Universidade de Vigo, Spain
Luiz Goncalves Junior, Sao Paulo State University—UNESP, Brazil
Manolo Paredes, Universidad de Las Fuerzas Armadas, Ecuador
Manuel Francisco González Penedo, Centro de Investigación CITIC, UDC, Spain
Manuel Tupia, Pontificia Universidad Catolica del Perú, Perú
Manuel Vilares Ferro, University of Vigo, Spain
Marcelo Henrique Prado da Silva, Instituto Militar de Engenharia—IME, Brazil
Marcos Barreto, Universidade Federal da Bahia, Brazil
María Carolina Romero Lares, World Maritime University, Sweden
viii Organization
Maria Manuela Martins Saraiva Sarmento Coelho, Military University Institute,

Portugal
Mario Bernabé Ron Egas, Universidad de las Fuerzas Armadas—ESPE, Ecuador
Maroi Agrebi, University of Polytechnique Hauts-de-France, France
Martín López Nores, University of Vigo, Spain
Matheus Pinheiro Ferreira, Instituto Militar de Engenharia, Brazil
Mauricio Loachamín Valencia, Universidad de las Fuerzas Armadas—ESPE,
Ecuador
Maximo Jr. Q. Mejia, World Maritime University, Sweden
Mohammed Mahmood Ali, Muffakham Jah College of Engineering and
Technology, Osmania University, India
Muhammed Ali Aydin, Istanbul University, Turkey
Nina Figueira, Brazilian Army, Brazil
Oscar Barrionuevo Vaca, Naval High School, Ecuador
Pastor David Chávez Muñoz, Pontificia Universidad Catolica de Peru, Peru
Paulo Afonso Silva, Military Institute of Engineering, Brazil
Pedro Ferreira, Centre for Marine Technology and Ocean Engineering (CENTEC),
Portugal
Rafael Timoteo de Sousa Junior, University of Brasilia, Brazil
Rashed Mustafa, University of Chittagong, Bangladesh
Regilberto Girão, Ministério Público Federal, Brazil
Renato Jose Sassi, Universidade Nove de Julho, Brazil
Reza Malekian, University of Pretoria, South Africa
Robert Beeres, Netherlands Defence Academy, The Netherlands
Robson Pacheco Pereira, Instituto Militar de Engenharia—IME/RJ, Brazil
Ronaldo Salles, Military Institute of Engineering—IME, Brazil
Rosalba Rodríguez Reyes, ESPE—University of the Armed Forces, Ecuador
Rui Carreira, Maia Polytechnic Institute, Portugal
Rui Jorge Palhoto Lucena, Military University Institute—Military Academy,
Portugal
Sarat Mohapatra, Centre for Marine Technology and Ocean Engineering
(CENTEC), Instituto Superior Tecnico, University of Lisbon, Portugal
Sebastião Alves Filho, Universidade do Estado do Rio Grande do Norte—UERN,
Brazil
Sidnei Alves de Araújo, Universidade Nove de Julho, Brazil
Silvio Melo, Universidade Federal de Pernambuco, Brazil
Simona Safarikova, Palacky University, Department of Development and
Environmental Studies, Czech Republic
Sonia Cárdenas Delgado, Universidad Politécnica de las Fuerzas Armadas—ESPE,
Ecuador
Suzana Paula Gomes Fernando da Silva Lampreia, Portuguese Navy, Portugal
Taybi Chakib, Faculty of Sciences, Mohammed First University, Morocco
Telmo Bento, ISMAI/IPMAIA, Portugal
Teresa Guarda, ESPE—University of the Armed Forces, Ecuador
Uma N. Dulhare, MJCET, Hyderabad, India
Organization ix
Veronica Rossano, Department of Computer Science—University of Bari Aldo

Moro, Italy
Wagner Tanaka Botelho, Federal University of ABC, Brazil
Zohra Bakkoury, Ecole Mohammadia d’Ingénieurs, Morocco
Preface
This book contains a selection of papers accepted for presentation and discussion at
The 2019 Multidisciplinary International Conference of Research Applied to
Defense and Security (MICRADS’19). This Conference had the support of IME
(Military Institute of Engineering) of Brazil, ESPE (University of Armed Forces) of
Ecuador, and AISTI (Iberian Association for Information Systems and
Technologies). It took place at Rio de Janeiro, Brazil, during May 8–10, 2019.
The 2019 Multidisciplinary International Conference of Research Applied to
Defense and Security (MICRADS’19) is an international forum for researchers and
practitioners to present and discuss the most recent innovations, trends, results,
experiences, and concerns in the several perspectives of defense and security.
The Program Committee of MICRADS’19 was composed of a multidisciplinary
group of more than 100 experts from 35 countries around the world and those who
are intimately concerned with Research Applied to Defense and Security. They
have had the responsibility for evaluating, in a ‘double-blind review’ process, the
papers received for each of the main themes proposed for the Conference:
(A) systems, communication and defense; (B) strategy and political–administrative
vision in defense; and (C) engineering and technologies applied to defense.
MICRADS’19 received about 100 contributions from 11 countries around the
world. The papers accepted for presentation and discussion at the Conference are
published by Springer (this book) and by AISTI and will be submitted for indexing
by ISI, Ei Compendex, Scopus, and/or Google Scholar, among others.
We acknowledge all of those that contributed to the staging of MICRADS’19
(authors, committees, workshop organizers, and sponsors). We deeply appreciate
their involvement and support that was crucial for the success of MICRADS’19.
Rio de Janeiro, Brazil Álvaro Rocha

May 2019 Robson Pacheco Pereira
xi
Contents
Part I Cybersecurity and Cyberdefense

Reducing the Attack Surface of Dynamic Binary Instrumentation
Frameworks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Ailton Santos Filho, Ricardo J. Rodríguez and Eduardo L. Feitosa
Single Sign-on Implementation: Leveraging Browser Storage
for Handling Tabbed Browsing Sign-outs . . . . . . . . . . . . . . . . . . . . . . . 15
Lokesh Ramamoorthi and Dilip Sarkar
Cybernetic Dependency Capacity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
Jorge Barbosa
Portuguese Concerns and Experience of Specific Cybercrimes:
A Benchmarking with European Citizens . . . . . . . . . . . . . . . . . . . . . . . . 39
João Vidal Carvalho, Álvaro Rocha, António Abreu
and Avelino Victor
Classification of Phishing Attack Solutions by Employing Deep
Learning Techniques: A Systematic Literature Review . . . . . . . . . . . . . 51
Eduardo Benavides, Walter Fuertes, Sandra Sanchez
and Manuel Sanchez
Is Cyber Warfare an Alternative? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
Jorge Barbosa
Memory Auditing for Detection of Compromised Switches in
Software-Defined Networks Using Trusted Execution Environment . . . . 77
Filipe Augusto da Luz Lemos, Rubens Alexandre de Faria,
Paulo Jose Abatti, Mauro Sergio Pereira Fonseca
and Keiko Veronica Ono Fonseca
Mobile Communication Systems: Evolution and Security . . . . . . . . . . . . 87
Teresa Guarda, Maria Fernanda Augusto, Isabel Lopes,
José Avelino Victor, Álvaro Rocha and Lilian Molina
xiii
xiv Contents
Part II Computer Networks, Mobility and Pervasive Systems

Evaluating Trail Formation in Collaborative UAV Networks
with Lethal Threats . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
Nícolas Pereira Borges, Cinara G. Ghedini
and Carlos Henrique Costa Ribeiro
Visual Analytics for the Reduction of Air Pollution on Real-Time
Data Derived from WSN . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
Dorys Quiroz, Byron Guanochanga, Walter Fuertes, Diego Benítez,
Jenny Torres, Freddy Tapia and Theofilos Toulkkeridis
Part III Defense Engineering

Toward the Development of Surveillance and Reconnaissance
Capacity in Ecuador: Geolocation System for Ground Targets
Based on an Electro-Optical Sensor . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
Zurita C. Marco Antonio, Aguilar C. Wilbert Geovany
and Enríquez C. Victor Xavier
Fuzzy Logic for Speed Control in Object Tracking Inside
a Restricted Area Using a Drone . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
Richard Navas Jácome, Harley Lovato Huertas,
Patricia Constante Procel and Andrés Gordón Garcés
Part IV Health Informatics in Military Applications

Micro-controlled EOG Device for Track and Control Military
Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149
Nayana L. M. Viana, José Ailton L. Barbosa Junior
and Francisco A. Brito-Filho
Part V Leadership and e-Leadership

Multilevel Military Leadership Model: Correlation Between
the Levels and Styles of Military Leadership Using MLQ
in the Ecuadorian Armed Forces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161
Celio Humberto Puga Narváez, Alex Fernando Jimenez Vélez,
Rafael Caballero Fernández and Osvaldo Fosado Téllez
E-leadership Using WhatsApp, A Challenge for Navy
Organizations: An Empirical Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171
Rolando P. Reyes Ch., Luis Recalde Herrera, Galo Andrade Daza,
Victor Gómez Bravo and Hugo Pérez Vaca
Contents xv
Part VI Planning, Economy and Logistics Applied to Defense

Career Anchors for the Portuguese Army’s Volunteers and Contract
Personnel: Using the Career Orientations Inventory . . . . . . . . . . . . . . . 185
Lúcio Agostinho Barreiros dos Santos
and Maria Manuela Martins Saraiva Sarmento Coelho
Part VII Simulation and Computer Vision in Military Applications

Implementation of Dubin Curves-Based RRT* Using an Aerial
Image for the Determination of Obstacles and Path Planning
to Avoid Them During Displacement of the Mobile Robot . . . . . . . . . . . 205
B. Daniel Tenezaca, Christian Canchignia, Wilbert Aguilar
and Dario Mendoza
Machine Learning and Multipath Fingerprints for Emitter
Localization in Urban Scenario . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 217
Marcelo N. de Sousa, Rafael L. Cardoso, Henrique S. Melo,
José W. C. Parente Jr. and Reiner S. Thomä
Virtual Rehabilitation System Using Electromyographic Sensors
for Strengthening Upper Extremities . . . . . . . . . . . . . . . . . . . . . . . . . . . 231
Z. Andrea Sánchez, T. Santiago Alvarez, F. Roberto Segura,
C. Tomás Núñez, P. Urrutia-Urrutia, L. Franklin Salazar,
S. Altamirano and J. Buele
Part VIII Strategy, Geopolitics and Oceanopolitics

The Portuguese Special Operations Forces as Instrument
of Foreign Policy: The Case Study of Afghanistan . . . . . . . . . . . . . . . . . 245
João Reis and Sofia Menezes
The Internal–External Security Nexus: EU Operation Sophia
Through the Lens of Securitization . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257
Ana Paula Brandão
The Evolution of EU’s Maritime Security Thinking . . . . . . . . . . . . . . . . 269
João Almeida Silveira
The Obstacles Women Face in Gaining Access to Special
Operations Forces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 281
João Reis, Rafael Gonçalves, Sofia Menezes and Manuela Kaczynska
The Transformation of the Defense and Security Sector
to the New Logistics 4.0: Public–Private Cooperation
as a Necessary Catalyst Strategy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 293
Manuel A. Fernández-Villacañas Marín
xvi Contents
Part IX Safety and Maritime Protection

An Autonomous Airship Swarm for Maritime Patrol . . . . . . . . . . . . . . 307
Constantino G. Ribeiro, Luciano Santos Constantin Raptopoulos
and Max Suell Dutra
Assessing the Location of Search and Rescue Stations
on the Portuguese Coast . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 321
Anacleto Correia, Ricardo Moura and Miguel Fonseca
Operational Scenarios Identification and Prioritization in Systems
of Systems: A Method and an Experience Report in the Defense
Domain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 333
Carlos Eduardo de Barros Paes and Valdemar Vicente Graciano Neto
Science Diplomacy: Strategic Initiative to Create a Buffer Zone
in the Caribbean Colombian Marine Protected Area Seaflower . . . . . . . 345
Fabián Ramírez-Cabrales and Sergio Iván Rueda Forero
Part X Wearable Technology and Assistance Devices

Quantum Well Infrared Photodetector for the SWIR Range . . . . . . . . . 363
Pedro Pereira, Lesslie Guerra, G. M. Penello, M. P. Pires,
L. D. Pinto, R. Jakomin, R. T. Mourão, M. H. Degani,
M. Z. Maialle and P. L. Souza
Independent Feeding of People Affected with Osteoarthritis
Through a Didactic Robot and Visual Control . . . . . . . . . . . . . . . . . . . . 371
Arturo Jiménez, Katherine Aroca, Vicente Hallo, Nancy Velasco
and Darío Mendoza
Adjustable Structure in Height for Assisted Feeding in Patients
with Osteoarthritis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 383
Katherine Aroca, Arturo Jiménez, Nancy Velasco, Vicente Hallo
and Darío Mendoza
Voice-Controlled Assistance Device for Victims of Gender-Based
Violence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 397
Miguel A. Domínguez, David Palomeque, Juan M. Carrillo,
José Mª Valverde, Juan F. Duque, Bruno Pérez and Raquel Pérez-Aloe
Part XI Analysis and Signal Processing

Analysis and Evaluation of the Positioning of Autonomous
Underwater Vehicles Using Acoustic Signals . . . . . . . . . . . . . . . . . . . . . 411
Enrique V. Carrera and Manolo Paredes
Contents xvii
Part XII Chemical, Biological and Nuclear Defense

Evaluation of 3D Printing Parameters on the Electrochemical
Performance of Conductive Polymeric Components for Chemical
Warfare Agent Sensing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 425
Joseane R. Barbosa, Pedro H. O. Amorim, Mariana C. de O. Gonçalves,
Rafael M. Dornellas, Robson P. Pereira and Felipe S. Semaan
Part XIII Information and Communication Technology

in Education
Interactive System to Improve the Skills of Children with Dyslexia:
A Preliminary Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 439
Jorge Buele, Victoria M. López, L. Franklin Salazar, Jordan-H. Edisson,
Cristina Reinoso, Sandra Carrillo, Angel Soria, Raúl Andrango
and Pilar Urrutia-Urrutia
Use of Information and Communication Technologies
in the Classroom: An Exploratory Study in Professional
Military Education . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 451
Nuno Alberto Rodrigues Santos Loureiro
and Lúcio Agostinho Barreiros dos Santos
Using Augmented Reality for Learning Naval Operations . . . . . . . . . . . 475
Anacleto Correia, António Gonçalves and Marielba Zacarias
Communication of the Quality of Higher Education Institutions
of Ecuador . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 487
Guillermo Santa María, Carina Rey, Lilian Molina and Arturo Clery
The Four Dimensions of the GDPR Framework: An Institutional
Theory Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 497
Isabel Maria Lopes, Teresa Guarda and Pedro Oliveira
Metrics and Indicators of Information Security Incident Management:
A Systematic Mapping Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 507
Alyssa Cadena, Franklin Gualoto, Walter Fuertes, Luis Tello-Oquendo,
Roberto Andrade, Freddy Tapia and Jenny Torres
Virtual Environment for Remote Control of UGVs Using
a Haptic Device . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 521
F. Roberto Segura, Pilar Urrutia-Urrutia, Z. Andrea Sánchez,
C. Tomás Núñez, T. Santiago Alvarez, L. Franklin Salazar,
Santiago Altamirano and Jorge Buele
Author Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 533
About the Editors
Álvaro Rocha holds a Habilitation (postdoctoral degree) in Information Science,

Ph.D. in Information Systems and Technologies, M.Sc. in Information Management,
and B.C.S. in Computer Science. He is currently a Professor of Information Systems
and Software Engineering at the University of Coimbra, researcher at CISUC (Centre
for Informatics and Systems of the University of Coimbra) and a collaborator
researcher at LIACC (Laboratory of Artificial Intelligence and Computer Science)
and CINTESIS (Centre for Research in Health Technologies and Information
Systems). His main research interests are in Information Systems Planning and
Management, Maturity Models, Information Systems Quality, Online Service
Quality, Intelligent Information Systems, Software Engineering, e-Government,
e-Health, and Information Technology in Education. He is also President of the AISTI
(Iberian Association for Information Systems and Technologies), Chair of the IEEE
Portugal Section Systems, Man, and Cybernetics Society Chapter, and
Editor-in-Chief of both JISEM (Journal of Information Systems Engineering &
Management) and RISTI (Iberian Journal of Information Systems and Technologies).
Professor Rocha has also served as a Vice-Chair of Experts in the European
Commission’s Horizon 2020 initiative, as an Expert at the Italian Ministry of
Education, University and Research, and as a General Chair of the WorldCIST
(World Conference on Information Systems and Technologies).
Robson Pacheco Pereira holds a degree in Chemistry (2000) and a doctorate in

Physics/Chemistry (2005) from the Federal University of Rio de Janeiro. He is
currently a Professor at the Military Institute of Engineering (IME/RJ), where he
coordinates the Nanotechnology Group and Physics and Chemistry, and develops
projects in the areas of nanotechnology, materials, electrochemistry, biosensors and
theoretical chemistry, intended for use in the area of Defense. The Group chiefly
pursues work on nanotechnology, physicochemistry, polymeric materials and
inorganic and nano-structured nanoparticles, kinetics, electrochemical solids, cor-
rosion, and power conversion and storage devices. Professor Pereira has also
published extensively in international journals.
xix
Part I
Cybersecurity and Cyberdefense
Reducing the Attack Surface of Dynamic
Binary Instrumentation Frameworks
Ailton Santos Filho, Ricardo J. Rodríguez and Eduardo L. Feitosa
Abstract Malicious applications pose as one of the most relevant issues in today’s
technology scenario, being considered the root of many Internet security threats. In
part, this owes the ability of malware developers to promptly respond to the emer-
gence of new security solutions by developing artifacts to detect and avoid them. In
this work, we present three countermeasures to mitigate recent mechanisms used by
malware to detect analysis environments. Among these techniques, this work focuses
on those that enable a malware to detect dynamic binary instrumentation frameworks,
thus increasing their attack surface. To ensure the effectiveness of the proposed coun-
termeasures, proofs of concept were developed and tested in a controlled environment
with a set of anti-instrumentation techniques. Finally, we evaluated the performance
impact of using such countermeasures.
Keywords Anti-instrumentation · Analysis-aware · Malware · Dynamic binary

instrumentation · Anti-analysis
1 Introduction
The number of software specially crafted with malicious intentions (commonly

referred to as malware) has increased in quantity and complexity during the last
years [2]. This fact supposes an issue for the anti-virus companies that need to dis-
pose of an up-to-date database of known malware to protect their customers. Thus,
malware analysts face every day to an increasing number of malware samples. For
A. S. Filho (B) · E. L. Feitosa

Instituto de Computação, Universidade Federal do Amazonas (UFAM), Manaus, Brazil
e-mail: assf@ufam.edu.br
E. L. Feitosa
e-mail: efeitosa@ufam.edu.br
R. J. Rodríguez
Centro Universitario de la Defensa, Academia General Militar, Zaragoza, Spain
e-mail: rjrodriguez@unizar.es
© Springer Nature Singapore Pte Ltd. 2020 3

Á. Rocha and R. P. Pereira (eds.), Developments and Advances in Defense
and Security, Smart Innovation, Systems and Technologies 152,
https://doi.org/10.1007/978-981-13-9155-2_1
4 A. S. Filho et al.
instance, Kaspersky (a well-known anti-virus company) stated that they analyzed

more than 360.000 malware samples per day in 2017 [11].
The process of analyzing a piece of software and determining if it is malicious can
be done manually or automatically. A manual analysis requires to first analyze the
program assembly code in a static way (without executing it) and then to execute it to
analyze the program behavior when interacting with the operating system (internal
behavior) and with the network (external behavior). However, this manual approach
is a very intensive and time-consuming process (in fact, it is usually referred to as
an art instead of science). To cope with the increasing trend of malware samples,
automation of malware analysis tasks in isolated analysis environments as sandboxes
or hypervisors has emerged in recent years [9]. This automatic approach allows the
anti-virus companies to keep updated their databases in a timely manner.
Unfortunately, malware writers have also started to incorporate small pieces of
code into their software to detect analysis environments. This is mainly motivated
because the longer the malware is undetected, the more revenue the cybercriminals
achieve. The kind of malware that behave differently depending on where they are
executed is referred to as evasive malware, analysis-aware malware, or split per-
sonality malware [3, 13, 19, 21, 23]. The incorporation of those evasion techniques
allows a malware to check where it is being launched and thus behave benignly when
an analysis environment is detected. As a consequence, the malware that is wrongly
identified as benign software can easily disseminate and penetrate a target system [5].
Special care has to be taken when analyzing an evasive malware to prevent de-
tection of the analysis environment. Some authors proposed to execute them first
on a bare-metal system and then compare their behavior when executed on other
emulation- and virtualization-based analysis systems [12]. Other authors proposed
tools such as PinVMShield [21] or Arancino [19] detect or circumvent the eva-
sion techniques used by malware to hinder their analysis. Both tools are based on
dynamic binary instrumentation, which allow the malware analyst to insert arbitrary
code during malware execution. Although this kind of analysis is feasible, it can
also be detected through different techniques, as widely reported in the literature (we
reviewed all these works in Sect. 2).
In this paper, we evaluate three techniques to demonstrate how a framework for
dynamic binary analysis (in particular, Intel Pin) can be detected. Those techniques
were previously proposed in the literature, but only from a theoretical perspective.
Here, we have implemented them to test their detection efficiency. Those evasion
techniques increase the attack surface of a dynamic binary instrumentation frame-
work, since they can be used by malware to detect that analysis environment. More-
over, we also provide the corresponding countermeasures to mitigate those evasion
techniques. Those countermeasures have been integrated in PinVMShield [21],
a plugin-based tool specially designed to circumvent evasive techniques performed
by malware. Furthermore, they have also been integrated in the benchmark-like tool
eXait [7], which serves to verify if a DBI framework is recognizable. Finally, to prove
the effectiveness of our countermeasures, we evaluate their efficacy and performance
using SPEC CPU 2006 [6]. The results showed that although the evasive behavior
Reducing the Attack Surface of Dynamic Binary Instrumentation Frameworks 5
can be circumvented, the use of current frameworks of dynamic binary analysis has
a great overhead, as already claimed by other authors [20].
This paper is organized as follows. Section 2 reviews the literature. Section 3
is devoted to previous concepts needed to follow the rest of this work, as the DBI
framework developed by Intel or the architecture of the PinVMShield tool. Sec-
tion 4 introduces the new techniques for detecting those analysis environments, as
well as the proper countermeasures. Finally, Sect. 5 concludes the work and states
future work.
2 Related Work
This section presents in chronological order the works related to techniques for
detecting a DBI framework, detailing also the countermeasures proposed in each
work. We first review the related works in the industry and then in the academy.
Most of the existing studies come from industry-related security conferences.
Falcón and Riva introduced in [7] the first techniques to detect the presence of
the Intel Pin DBI framework in a Windows OS. In particular, thirteen techniques
with the corresponding proofs of concepts (PoCs) were presented. Those PoCs were
distributed as a sort of a benchmark-like tool called eXait. This tool is useful to verify
if a DBI framework is recognizable.
Two years later, Li and Li [14] showed that the DynamoRIO DBI framework can
be detected in both Windows and Linux environments through ten new evasion tech-
niques. Furthermore, they remarked the transparency problem of DBI frameworks in
some cases. Similarly, Hron and Jermář proposed in [10] six new evasive techniques
against the Pin and DynamoRIO, providing also PoCs of those techniques. Sun et
al. also presented in [22] other six new evasive techniques aimed at the DBI Pin and
DynamoRIO frameworks. In addition, they introduced the idea of escaping of the
sandbox-like environment provided by a DBI framework, thus compromising the
analysis environment.
Regarding the academic literature, Rodríguez et al. [21] reviewed and provided a
taxonomy of the existing anti-DBI and countermeasures techniques up to that time.
Moreover, they released an extensible Pin-based tool called PinVMShield that
enabled to analyze applications with those evasive techniques by means of circum-
venting them. Later, Polino et al. [19] introduced a review and classification of some
anti-DBI techniques mainly focused on the Pin DBI framework. Furthermore, they
developed a set of approaches to mitigate evasion techniques and used the eXait
tool [7] to validate their effectiveness. They also introduced a Pin-based tool named
Arancino, which incorporates the proposed countermeasures and allows to retrieve
the original binary form of malware programs protected by software packers.
Recently, Zhechev published a Master’s thesis [24] raising the question whether
DBI frameworks are appropriate tools for analyzing malware and other potentially
evasive artifacts. Moreover, he introduced thirteen new techniques to detect the pres-
ence of Intel Pin DBI framework in a Linux OS and also demonstrated that escaping
of such a DBI framework is feasible due to the shared memory model used by DBI
frameworks. In this regard, Zhechev stated that the isolation and the stealthiness of the
analysis code under DBI frameworks are not guaranteed, and thus, DBI frameworks
are unsuitable for building any security-related application.
As shown, DBI frameworks are gaining popularity among security researchers
as a way to insert arbitrary code during program execution. Following this trend, in
this paper, we provide PoCs of three evasion techniques already documented in the
literature (but without PoCs) and we give the proper countermeasures.
3 Previous Concepts
3.1 Intel Pin DBI Framework
The Pin DBI framework (or Pin for short) enables to build easy-to-use, portable,
transparent, and efficient dynamic instrumentation tools. Pin was designed by Intel in
2005 and gives support for the three major desktop operating systems (i.e., Windows,
Linux, MacOS X).
Pin is composed of the three typical DBI components (depicted in Fig. 1): (1)
the application to be instrumented; (2) a dynamic binary analysis tool developed
with Pin, normally termed as Pintool; and (3) the DBI engine. The DBI engine
consists of a virtual machine (VM), a code cache, and an instrumentation application
programming interface (API) invoked by the Pintool. The VM takes as input the native
executable code of the application to be instrumented and uses a just-in-time (JIT)
compiler to insert the instrumentation code, prior to execution. Then, the resulting
instrumented code is saved in the code cache and the execution is transferred to it.
After execution, the JIT compiler fetches the next sequence of instructions to be
executed and generates more code. The emulator unit is in charge of instructions that
cannot be directly executed, such as system calls that require special handling from
the VM [16]. As shown in Fig. 1, the instrumented application, the Pintool, and the
DBI engine are executed in the same memory address space.
Pin has been widely used in the scientific community. As a result, a lot of interest-
ing Pintools have been released. In this regard, it is worth mentioning Pintools that
enable debugging for instrumented applications, such as PinADX [15] or the tool
introduced in [18].
3.2 PinVMShield
PinVMShield [21] is a tool to detect and circumvent evasion techniques used by

analysis-aware malware. The tool follows a plugin architecture, enabling an easy
extension for covering other evasion techniques than the currently supported ones.
Fig. 1 Pin’s software Pin same address space
architecture (adapted
from [16]) Application Pintool
Virtual Machine
Code cache
(VM)
JIT Compiler
Instrumentation
APIs
Dispatcher
Emulation
Unit
Operating System (OS)
It was released under GNU GPL version 3 license and its source code is publicly
available at https://bitbucket.org/rjrodriguez/pinvmshield/.
PinVMShield is mainly focused on Windows OS and uses two granularity in-
strumentation levels, at the routine level and at the instruction level. This allows the
tool to detect evasive behavior based on Windows APIs, such as checking the pres-
ence of a software debugger though IsDebuggerPresent or CheckRemote
Debugger, or based on specific assembly code instructions (e.g., sidt, sgdt, or
sldt [8]).
PinVMShield currently addresses the evasive behavior performed by software
binaries to detect virtual environments (in particular, VirtualPC, VMWare, and Vir-
tualbox), debuggers (WinDBG, OllyDBG, and ImmunityDebugger), and sandboxes
(WinJail, Cuckoo Sandbox, Norman, Sandboxie, CWSandbox, JoeSandbox, and
Anubis).
4 Evasive Techniques and Countermeasures
In this section, we address some of the evasion techniques that a malware can in-
corporate into its code to detect when it is being executed inside a DBI framework.
Furthermore, to the best of our knowledge, we are the first to provide countermeasures
for those evasion techniques, which are introduced next.
4.1 Evasive Techniques
An exhaustive list of techniques to detect DBI frameworks is discussed in [22, 24].

There, small pieces of code are given as PoCs, as well as proper countermeasures,
specially designed for the Intel Pin DBI framework and Linux OS running on top of
Intel x86-64 architectures.
Among those evasive techniques for which no countermeasures are provided,
there are few of them particularly relevant, since they may be used as an attack
vector by malware against the analysis environment. In particular, those techniques
are the neglecting of the No-eXecute bit, TLS detection, and code cache signatures
detection. In the rest of this section, we describe each of those evasion techniques
in detail. Furthermore, we also provide PoCs and countermeasures. Unlike [24], we
focus on the Windows OS, as it is the most prevalent system attacked by evasive
malware [21].
To demonstrate the effectiveness of these evasion techniques against Pin, our PoCs
have been integrated as plugins for eXait [7]. Those plugins are publicly released
under GNU GPLv3 and freely available online1 Moreover, our proposed counter-
measures have been integrated in PinVMShield [21].
Neglecting No-eXecute Bit. The No-eXecute (NX) bit is a defense mechanism added
at hardware level to prevent the execution of data memory pages by the processor.
Roughly speaking, the main idea is that no memory zone is simultaneously writable
and executable. This feature is incorporated by almost all the processor’s manufac-
turers, although referred to under different nomenclature (e.g., execute disable bit
in Intel, enhanced virus protection in AMD, or execute never bit in ARM). Recall
that a DBI framework uses a JIT compiler that “recompiles” the application code
in conjunction with the instrumented code (see Sect. 3.1). That is, it needs to write-
then-execute certain memory zones. Therefore, as suggested by Zhechev [24], any
program instrumented by a DBI framework in JIT mode has this protection disabled.
Therefore, an application can detect a DBI framework by allocating a new heap
space, placing the valid code on it, and then executing it. When the application is not
being instrumented, then the execution crashes since the heap memory page has no
permission to execute. However, the execution under a DBI framework will continue
normally.
TLS Detection. Thread Local Storage (TLS) is a feature that allows a developer to
provide unique data for each thread, in vector format, accessed by the process using
a global index. Roughly speaking, TLS variables can be seen as global variables only
visible to a particular thread and not the whole program. In the case of Windows,
those per-thread global variables are maintained in the TLS directory, which is a part
of the Portable Executable (PE) header of an executable image. The PE header is the
header of any Windows executable file. The minimum number of positions in the
directory is guaranteed to be at least 64 for any system, while the maximum number
is 1088 [17]. According to [22], DBI frameworks as Pin allocate and use positions
of that data structure for internal purposes.
1 See https://github.com/ailton07/eXait_Plugin_PinDetectionByDEPNeglect, https://github.com/

ailton07/eXait_Plugin_CodeCacheDetectionByFEEDBEAF, https://github.com/ailton07/eXait_
Plugin_PinDetectionByTLS.
Since allocated indexes are shared by any thread in the process, an application
can inspect the number of positions allocated in the TLS directory, revealing the
presence of a DBI framework.
Code Cache Signatures Detection. As reported in [19], there are several artifacts
that a DBI tool unavoidably leaves in memory. In this regard, the authors in [22]
shown that those memory artifacts can be easily detected by any application under
analysis. In the case of Pin, one of those memory artifacts is the hexadecimal pattern
0xFEEDBEAF. This byte sequence is repeated through the memory zones used as
code cache by Pin, as an identification fingerprint.
Therefore, an application can detect those memory patterns by means of memory
scans. The presence of those byte patterns in the memory allocated by a process can
be used as a criterion to indicate the presence of Pin in this particular case.
4.2 Countermeasures
For the development of the proposed countermeasures, we used the tool

PinVMShield since its plugin-based architecture makes easier for a developer
to incorporate new mitigation techniques against an evasion technique. To foster
research in this area, the developed countermeasures are also publicly released an
available online.2 In the following, we briefly describe each countermeasure.
Neglecting No-Execute Bit. As countermeasure, we can rely on the set of Windows
APIs that allow a developer to check for specific memory addresses and permissions.
The developed countermeasure detects when an attempt to execute a piece of code
from an invalid region (non-executable memory) occurs, and allows the binary analyst
to take action afterward. A more transparent implementation would raise an exception
with status STATUS_ACCESS_VIOLATION, imitating the behavior outside the DBI
framework.
TLS Detection. Since the TLS access is made exclusively using APIs provided by
Windows OS, the proposed countermeasure relies on monitoring those APIs and
logically isolating the positions used by Pin. Thus, an attempt to read or modify a
position already allocated by the DBI tool is redirected to a position unrelated to it.
The PinVMShield tool already provides facilities for intercepting and modifying
Windows APIs, through the ReplaceWinAPI method.
Code Cache Signatures Detection. As suggested in [19], a process can scan its entire
memory space performing an inspection of every allocated memory page. In Win-
dows OS, this can be done through APIs as VirtualQuery or NtQueryVirtualMemory.
To mitigate this evasion technique, it is necessary to establish a memory access policy,
telling which memory addresses are protected memory addresses. This can be done
by creating a blacklist of memory addresses related to the DBI framework which
2 See https://github.com/ailton07/PinVMShield.
cannot be accessible by the instrumented application. In this regard, Pin provides

a set of APIs, such as CODECACHE_AddTraceInsertedFunction, which allows an
analyst to monitor the creation of code caches and thus get their addresses, enabling
the creation of that blacklist upon execution.
Once the blacklist is initialized, those Windows APIs that allow memory page
inspection can be monitored to bypass the queries of any protected memory regions.
Our developed countermeasure intercepts those calls when an attempt to obtain infor-
mation from a protected memory region is done and returns a zero value, indicating
that the function failed.
4.3 Efficacy of the Countermeasures and Performance

Impact
These countermeasures were incorporated in PinVMShield to evaluate their effec-

tiveness. We developed the small pieces of code implementing evasive techniques and
execute them with PinVMShield and our countermeasures enabled. As a result,
TLS Detection and Code Cache Signature Detection were mitigated successfully.
Similarly, Neglecting No-Execute Bit was detected during the execution attempt,
allowing the analyst to take appropriate action at run time.
To measure the impact caused by the use of the PinVMShield tool in the
application performance, we used the benchmark tool SPEC CPU2006 [6] widely
used to evaluate performance measurements on DBI tools [1, 4, 16]. While executing
SPEC with PinVMShield, features not related to the countermeasures presented
here were disabled to avoid performance degradation by other means.
As experimentation environment, we used a virtualized environment on top of a
KVM processor 8 cores 3.41 GHz, with 16GB of RAM memory. The virtual machine
was running a Windows 7 SP1 x64 with Intel Pin 2.14-71313 and Microsoft 32bit
C/C++ compiler v16 for SPEC2006 compilation.
Table 1 summarizes the results obtained for each one of benchmark tools of
SPEC. The first column lists the names of the benchmarks. The second column tells
the execution time (in seconds) of Pin, while the third shows the execution time (in
seconds) of Pin with the PinVMShield tool. Finally, the fourth column presents
the overhead when PinVMShield is used. On average, the overhead was 59.73%
with a standard deviation of 98.68%.
5 Conclusions and Future Directions
Malicious software able to detect an analysis environment is a potential threat, since

they can behave differently to evade the classification as malware when they detect
an analysis environment.
Table 1 Overhead introduced by PinVMShield with our countermeasures

Benchmark Instrumentation time PinVMShield time PinVMShield
(s) (s) overhead (%)
400.perlbench 484 534 10.3305
401.bzip2 560 562 0.3571
403.gcc 443 1150 159.5936
429.mcf 192 209 8.8541
445.gobmk 521 537 3.0710
456.hmmer 680 718 5.5882
458.sjeng 606 607 0.1650
464.h264ref 1069 4071 280.8231
471.omnetpp 294 819 178.5714
473.astar 339 326 –3.8348
483.xalancbmk 287 326 13.5888
In this paper, we have studied three evasion techniques of DBI frameworks. Those
techniques are used as attack vectors by malware to detect an analysis environment.
We have provided small pieces of code as proof of concepts of those techniques, as
well as their countermeasures to reduce the attack surface of DBI frameworks. Those
countermeasures were developed on top of the PinVMShield tool. Both proof of
concepts and countermeasures have been evaluated in a virtualized environment to
prove their effectiveness. Our experiments showed a performance overhead close to
60% by the use of developed countermeasures.
Unlike the work in [24], we have shown that DBI frameworks are suitable for se-
curity purposes: when equipped with the appropriate tools, the requirements needed
in a security analysis context (i.e., stealthiness and isolation) are achieved. How-
ever, the performance overhead introduced by the use of DBI frameworks is still
an issue. Nonetheless, we argue that it is necessary to keep observing the growth
of evasion techniques focused on DBI tools. Despite efforts done to develop DBI
evasion techniques countermeasures, there are still evasion techniques that threaten
the stealthiness of DBI frameworks and pose a challenge to system security pro-
fessionals. For instance, the intrinsic problem of overhead detection that basically
affects to any dynamic analysis tool.
Acknowledgements The research of A. Santos Filho and E. L. Feitosa supported in part by the
FAPEAM Proc. No. 009/2017 and by the Federal University of Amazonas (UFAM). The research
of R. J. Rodríguez was supported in part by the University, Industry and Innovation Department of
the Aragonese Government under Programa de Proyectos Estratégicos de Grupos de Investigación
(project references T21-17R).
References
1. Arafa, P.: Time-aware dynamic binary instrumentation. Ph.D. thesis, University of Waterloo
(2017)
2. AV-TEST GmbH: The AV-TEST Security Report 2017/2018 (2018)
3. Balzarotti, D., Cova, M., Karlberger, C., Kirda, E., Kruegel, C., Vigna, G.: Efficient detection of
split personalities in malware. In: Proceedings of the Network and Distributed System Security
Symposium (NDSS) (2010)
4. Bruening, D., Zhao, Q., Amarasinghe, S.: Transparent dynamic instrumentation. ACM SIG-
PLAN Not. 47(7), 133–144 (2012)
5. Carpenter, M., Liston, T., Skoudis, E.: Hiding virtualization from attackers and malware. IEEE
Secur. Priv. 5(3), 62–65 (2007)
6. CPU2006, S.: Standard performance evaluation corporation. https://www.spec.org/cpu2006/
(2006) (Online)
7. Falcón, F., Riva, N.: Dynamic binary instrumentation frameworks: I know you’re there spying
on me (2012)
8. Ferrie, P.: Attacks on virtual machine emulators. Symantec Adv. Res. Threat. Res. 1–13 (2007)
9. Greamo, C., Ghosh, A.: Sandboxing and virtualization: modern tools for combating malware.
IEEE Secur. Priv. 9(2), 79–82 (2011)
10. Hron, M., Jermář, J.: SafeMachine malware needs love, too. https://www.virusbulletin.com/
uploads/pdf/conference_slides/2014/sponsorAVAST-VB2014.pdf (2014) (Online)
11. Kaspersky lab: Kaspersky lab detects 360,000 new malicious files daily—up 11.5% from
2016. https://www.kaspersky.com/about/press-releases/2017_kaspersky-lab-detects-360000-
new-malicious-files-daily (2017) (Online)
12. Kirat, D., Vigna, G., Kruegel, C.: Barecloud: Bare-metal analysis-based evasive malware de-
tection. In: 23rd USENIX Security Symposium (USENIX Security 14), pp. 287–301. USENIX
Association, San Diego, CA (2014)
13. Kumar, A.V., Vishnani, K., Kumar, K.V.: Split personality malware detection and defeating in
popular virtual machines. In: Proceedings of the 5th International Conference on Security of
Information and Networks (SIN), pp. 20–26. ACM (2012)
14. Li, X., Li, K.: Defeating the transparency features of dynamic binary instrumentation. BlackHat
US (2014)
15. Lueck, G., Patil, H., Pereira, C.: PinADX: An interface for customizable debugging with
dynamic instrumentation. In: Proceedings of the 10th International Symposium on Code Gen-
eration and Optimization (CGO), pp. 114–123. ACM, New York, NY, USA (2012)
16. Luk, C.K., Cohn, R., Muth, R., Patil, H., Klauser, A., Lowney, G., Wallace, S., Reddi, V.J.,
Hazelwood, K.: Pin: Building customized program analysis tools with dynamic instrumen-
tation. In: Proceedings of the 2005 ACM SIGPLAN conference on Programming Language
Design and Implementation, PLDI ’05, pp. 190–200. ACM, New York, NY, USA (2005)
17. Microsoft: Thread Local Storage. https://msdn.microsoft.com/en-us/library/windows/
desktop/ms686749(v=vs.85).aspx (2018) (Online)
18. Pan, H., Asanović, K., Cohn, R., Luk, C.K.: Controlling program execution through binary
instrumentation. SIGARCH Comput. Archit. News 33(5), 45–50 (2005)
19. Polino, M., Continella, A., Mariani, S., D’Alessio, S., Fontata, L., Gritti, F., Zanero, S.: Mea-
suring and Defeating Anti-Instrumentation-Equipped Malware. Detection of Intrusions and
Malware and Vulnerability Assessment, pp. 73–96. Springer International Publishing, Cham
(2017)
20. Rodríguez, R.J., Artal, J.A., Merseguer, J.: Performance evaluation of dynamic binary instru-
mentation frameworks. IEEE Lat. Am. Trans. (Rev. IEEE Am. Lat.) 12(8), 1572–1580 (2014)
21. Rodríguez, R.J., Gaston, I.R., Alonso, J.: Towards the detection of isolation-aware malware.
IEEE Lat. Am. Trans. 14(2), 1024–1036 (2016)
22. Sun, K., Li, X., Ou, Y.: Break Out of The Truman Show: Active Detection and Escape of
Dynamic Binary Instrumentation. Black Hat Asia (2016)
23. Vishnani, K., Pais, A.R., Mohandas, R.: Detecting & defeating split personality malware. In:
Proocedings of the 5th International Conference on Emerging Security Information, Systems
and Technologies (SECURWARE), pp. 7–13 (2011)
24. Zhechev, Z.: Security evaluation of dynamic binary instrumentation engines. Master’s thesis,
Department of Informatics Technical University of Munich (2018)
Single Sign-on Implementation:
Leveraging Browser Storage for
Handling Tabbed Browsing Sign-outs
Lokesh Ramamoorthi and Dilip Sarkar
Abstract Organizations provide multiple web-based services to its users. To manage

easy provisioning and deprovisioning of users, they offer single sign-on (SSO) service
as an access control mechanism. To access any SSO service, users need to remember
only one set of credentials. These credentials are managed by organization’s identity
and access management system (IAM). The identity provider (IDP) is the federated
identity management platform to authenticate users, for service providers (SPs), to
access and use the services. Users may access the services provided by organization
via web browsers. Modern web browsers provide a feature called tabbed browsing,
where tabs are widgets within the browser window, so that the users can stay organized
when browsing multiple Web sites. In this paper, we analyzed how SSO works in
a tabbed browsing environment. Our analysis shows that, in some scenarios, a user
may not sign out from some services that he or she was using. This situation may lead
to information security attacks. Also, we propose a solution that can be implemented
on the browser to ensure a safe sign-out process.
Keywords Information security · Access control · Federated identity

management · Authentication · Authorization · Identity provider · Service
provider · Browser security
1 Introduction
Today, every entity (university, organizations, and businesses, etc.) provides multiple
web-based services. The user population of an entity changes dynamically, that is,
existing users leave and new users are added. For user conveninces, the Information
Technology (IT) team of the organization provides Single Sign-on (SSO) service to its
L. Ramamoorthi (B) · D. Sarkar

University of Miami, Coral Gables, FL 33146, USA
e-mail: lokeshr@miami.edu
D. Sarkar
e-mail: sarkar@miami.edu

https://doi.org/10.1007/978-981-13-9155-2_2
16 L. Ramamoorthi and D. Sarkar
users. A user of an entity that provides SSO service needs only one set of credential
(say, username and password) to access all available services.1
In a SSO service system, a (physical or logical) Identity Provider (IDP) verifies
credentials of all users who want to receive any web-based service provided by the
entity. While a user is verified for one of the many web-based services, the user can
access all other services authorized to the user because the IDP maintains an active
IDP session of the user in the user’s browser.
Currently, SSO service has expanded to individual Web sites, where popular social
networking sites (such as Facebook, Google, or LinkedIn) work as IDP. This concept
is referred to as social login. All the security issues common to SSO are also present
in social login system.
In this paper, we investigate how single sign-on implementation on tabbed brows-
ing scenarios may risk the information security. The rest of the paper is organized
as follows, key components of SSO, how tabbed browsing works, common browser
storage mechanisms, how multiple tabs handle information during SSO process,
security pitfalls, possible solutions to avoid information security incidents during
SSO.
1.1 Key Components of SSO Implementation
Many commercial implementations of SSO follow OAuth 2.0 standard [1], a Secu-
rity Assertion Markup Language (SAML)-based reference architecture. SAML is a
XML-based framework [2] for communicating user authentication, entitlement, and
attribute information between Identity Provider (IDP) and Service Provider (SP) via
web browser.
Three main components in a SAML-based SSO implementation are:
(i) Web browser of the user,
(ii) Identity provider—(often) authentication server of the user’s organization, and
(iii) Service provider—the application that the user gets service.
Figure 1 illustrates the interaction model of the components in the SSO imple-
mentation. Details about SSO components and related information security issues
have been presented in earlier work (see [3] and references therein).
2 Tabbed Browsing
Observing users’ tendency of accessing multiple Web sites simultaneously by open-

ing multiple browser windows, one tab for each Web site, the browser software
companies introduced tabbed browsing. Tabs are browser widgets, where in a single
1 For instance, federated identity management is a widely used SSO mechanism.

Single Sign-on Implementation: Leveraging Browser Storage … 17
Fig. 1 Browser-centric SSO

process
Fig. 2 A tabbed browser

window
browser window users can open multiple web pages simultaneously, one for each
web page. With the help of tabbed browsing, users can stay organized and it provide
them with an organized and a convenient browsing experience.
When a user signs in to a Web site on one tab and opens another page from the
same Web site on a different tab, both the tabs share the same session information
and the user is not required to sign in again. The user adaptation for tabbed browsing
is also very positive, because he/she can keep all their browsing activities organized
in a single window [4].
Figure 2 shows how a tabbed browser appears to users. Whenever the user clicks
on browser icon, such as Mozilla Firefox, Internet Explorer, or Google Chrome, the
browser application opens with the default tab. The default tab is highlighted in the
Fig. 2 as ‘Browser Tab 1.’ If the user would like to open additional web pages, it
can be accomplished by opening more tabs. Each tab displays a web page. If the
user decides to have any of the tabs separated in a new window, he/she can do so
by dragging it as out of the main window; this transforms the selected tab to a new
browser window.
3 Client-Side Storage
To provide an enhanced personalized web browsing experience, the service providers

send some information to the web browsers about the users’ activities. The browser
may store this information on the client, (user’s computer or other devices used for
browsing) and retrieves the information when requested. This storage is called client-
side storage or browser storage. The information storage and retrieval is enabled by
Javascript API functions. Table 1 gives a summary of various browser storage options.
A good understanding of all the Web site storage options helps in selecting the
most secure and efficient storage solutions for the browsers and their users. In the
Table 1 Standard browser storage options

Type of storage Description Data format Advantages Disadvantages
stored
Cookies Flat file storage Key value Traditional, Highly susceptible to
widely used, information security
simple format attacks such as
cross-site scripting
(XSS) and cross-site
request forgery
(CSRF)
WebStorage Browser inbuilt Key value [5] Stored within Not encrypted.
API—Local storage, data the browser, Complex datatypes
storage persisted in the scope restricted not supported
system even to the Web site
after session
WebStorage Browser inbuilt Key value Stored within the Not encrypted.
API—Session storage, data is browser, scope Complex datatypes
storage lost after the restricted to the not supported
tab/browser browser session
closure
IndexedDB Browser inbuilt No SQL Multimedia Encryption is not
storage storage, NoSQL objects can also inbuilt, but can be
format, be stored, integrated with
multimedia indexed for WebCryptoAPI [6],
supported faster access data retrieval may
depend on the
indexing and
implementation
CacheAPI and Effective for Multiple formats Useful for Encryption not
service workers offline offline access inbuilt, efficiency
experiences, and performance is
HTML, CSS can dependent on
be stored the implementation
forthcoming sections, we will discuss how current SSO implementations rely more
on cookies and the problems associated with it. We will also discuss few possible
solutions leveraging other client-side storage methods such as IndexedDB API.
4 Process Sequence in Sign-in, Sign-out Activities
In this section, we discuss the process flow during a single sign-on, sign-out activity
as seen from a web browser’s perspective. For the discussion here, we assume that
the SPs and the IDP belong to different platforms (either physically or logically).
Whenever a user attempts to sign-on to any of the SPs, the user is redirected to the
IDP.
For ease of description, complete discussion is divided into cases:
(i) the user has no active session with any of the SPs and
(ii) the user has one or more active session with some SPs.
.
4.1 User Sign-in Activity: No Active Signed-in Sessions
Here, the user has no active session with any SP and the browser has no record of
active SSO. Once a user requests for a service via the browser, the SP checks for
possible active service session with it; because it is an initial request, the SP requires
that the user be authenticated. For this, the SP finds the source of request (Where are
you from?—WAYF) and redirects the browser to the user’s IDP.
As soon as the browser is redirected from the SP to the IDP for authentication, the
IDP checks the browser for existing active IDP session. Since no IDP session exists,
the IDP challenges the user for valid credentials and displays the sign-on page; then,
the user provides sign-in credentials for possible sign-on. Once the user provides
valid credentials (which could be a simple user ID and password or a complicated
multi-factor authentication), IDP creates a new active session and asserts it on the
browser. IDP stores the information in the browser cookies.
In addition, the IDP creates a SAML record response that contains assertion
information of the authenticated user for the browser. The assertion information in the
SAML record contains information about users organization as well as authorization
level. The browser forwards the IDPs SAML record to the service provider. After
receiving the SAML assertion info, the SP creates a session for the user; at this point,
a service session is initiated. The service provider stores the information either in
local storage or in the browser cookies. Figure 3 shows a logical view of the browser
storage after the service session is initialized. The tab shown in the figure is the
default tab, where the SP session is accessed by the user.
Fig. 3 Browser storage

structure with IDP and
session info
As presented, in the SSO approach for an initial sign-in, the browser coordi-
nates communication between the SP and IDP. Moreover, browser maintains session
information for both the SP and IDP. As we discussed in Sect. 2, the user can access
multiple pages of the same SP via. different browser tabs.
4.2 User Sign-in Activity: Tabbed Browsing
Let us consider two scenarios where a user accesses services in a tabbed browsing
fashion,
(1) Scenario 1: the user access to a new web page of the SP where she is currently
signed-in and
(2) Scenario 2: the user tries to access a new web page of a SP where she is not
currently signed-in, but IDP of the new SP is same as the IDP of the SP where
she is currently signed-in.
Scenario 1: Once a user signs in to a service provider’s Web site, she can access
another page of the same Web site opening a new browser tab. Figure 4 shows storage
structure when two browser tabs of one SP is accessed. Both the tabs Tab1 and Tab2
access the SP1 session information.

structure for one SP and two
tabs

structure for two SPs and
three tabs
Scenario 2: In this tabbed browsing scenario, the user signs in for a service and opens
a new tab for another service. Let us consider services SP1 and SP2 , both leverage
authentication of the same IDP. Initially, user has signed on to service SP1 , and a
while later, user wishes to open another browser tab and access another service say
SP2 . The service SP2 also leverages the authentication provided by the same IDP of
the organization.
Figure 5 shows the logical view of the browser storage and how the session infor-
mation is accessed by multiple tabs.
Even though the browser allows access to SP2 without requiring user to provide
credentials, there is a risk in this situation, where the user might sign out of one service
provider and assume that she will be signed out from all other service providers.
The critical consideration here is the status of IDP session may remain active even
after user signs out from SP. First, we consider possible situations that arise, because
of sign-out by the user.
4.3 Sign-out Activity—Tabbed Browsing with Parallel

Sessions
In this section, we will discuss the sign-out activity by the user in a tabbed browsing
scenario. The key reason we discuss the sign-out activity for parallel tabbed sessions
is that, the user does not explicitly sign in when she opens a request in new tab. It is
possible that the user can assume sign-out activity and also does not require explicitly
signing-out from all the related tabs. Here, there are two cases of sign-out, the user
might encounter in the tabbed browsing session.
Case 1: When a user signs out from a service from a browser tab, the SP session is
invalidated. However, when the user accesses the same service via another tab, the
browser does not require user to sign in again, because the IDP session is still active.
So during the second time request of a service via a different tab, the service retrieves
SAML from the IDP session and does not explicitly challenges the user for sign-in
credentials.
Hence, the scenario is bit confusing for the user because the user may assume that
the sign-out from one tab of a service signs out all the active sessions, whereas in the
actual scenario of SSO implementation as long as IDP session is active, the service
is still accessible without requiring user credentials.
Case 2: Assuming two services SP1 and SP2 leverages the service of the same IDP,
when the user requests for SP2 after signing-out from SP1 , the browser does not
prompt for user sign-in credentials. This is because SP2 could get the information
from the IDP session and will not prompt the user for sign-in credentials.
As shown in Fig. 5, irrespective of how many different services being accessed,
the IDP session is shared by all the services.
5 Solutions to Avoid Undesired Active Sessions
There are many advantages of SSO. But still there are concerns about the implemen-
tation that revolves around the security and privacy of user information [7]. In this
section, we discuss how application developers and SSO service developers prevent
hidden active sessions.
5.1 Notifying the User During Sign-Out
During a SP sign-out activity, the SP can display a notification to the users,

You are signing-out from the Service, but you may have other active sessions, including the
Identity Provider sessions that was created by your authentication server.
This would forewarn the user that there might be other active and open SP sessions
which the user has to manually sign out.
This notification would be very helpful in a tabbed browsing scenario where when
the user signs out from one tab the service provider can notify that there might be
other tabs that are active. The sign-out confirmation link can also include a trigger
or link that invalidates the IDP session. This will reduce the risk to a greater extent
when the user opened multiple tabs for multiple service sessions.
5.2 Triggering IDP Session Invalidation During Service

Provider Sign-out
Another solution is, whenever the user signs out of a service, the SP application
sends a notification to the IDP that triggers a sign-out process of the IDP session. So
in this case, the IDP session becomes inactive. Any other users trying to access the
service has to sign in with their own set of credentials.
5.3 Automatically Sign-out at Browser Window Closing Event
With modern browsers and advanced browser scripting technologies, the applica-
tions can be implemented where in case if the user closes the browser window,
the SP session variables are invalidated. Not many users are aware of the fact that
closing a browser window is not the same as a sign-out activity. Implementing auto-
invalidation of a session during a window closing event helps where the users do
not perform explicit sign-out. This scenario is very appropriate implementation in
situations where multiple users accessing a common computer, such as in public
library.
The auto-trigger IDP sign-out during a service provider sign-out is very effective
during tabbed browsing scenario as well, where the browser can give a quick noti-
fication to the user and triggers a IDP sign-out. This way the user can feel secure
about the sign-out activity that no one can access any services without a user sign-in
credentials.
So, when new users open the browser, they are prompted for their credentials.
6 Specification for Handling SSO Sign-out Scenarios via. A

Browser-Based SSO Manager Module
The following proposal emphasizes on browser to handle the security gaps discussed
above. In this specification, we create a new module called SSO Manager Module.
This module can be implemented as a browser extension. This module does not
manipulate or add any additional features to IDP or SP applications. This module
provides a browser-based mechanism that ensures a safe sign-out process.
SSO Manager module uses a data object named Active SSO SessionsInfo that
contains essential information about SP and IDP sessions. Whenever a sign-out
event happens, the module will effectively handle the event either by automatically
signing-out the user from the session, or displaying a notification for input from the
user.
6.1 SSO Manager Module : Data Structure
The SSO Manager module is a web browser extension. It saves the information
about the active IDP Sessions and the SP sessions that leverage the IDP. To store the
information, SSO manager leverages IndexedDB storage of the web browser.
Fig. 6 Active SSO

sessions—schema
The main advantage of IndexedDB is that it helps in organizing the IDP-SP hier-
archy. Following is the proposed schema for the SSO Manager application.
The SSO Manager stores the IDP Session metadata in the IndexedDB.
Figure 6 shows us the schema of data stored by the SSO Manager module. Every
record holds metadata about the IDP, name of the IDP, IDP’s sign-out URL, and IDP’s
timeout threshold. Each IDP Info record contains an array of service provider’s
meta data. Every SP metadata contains the information about the service provider
such as name or id, SP’s timeout threshold, and SP’s sign-out URL. As we discussed
in Sect. 2, users may access the services via multiple browser tabs. Hence, the SP
Info data contains the list of tabs that the SP is currently accessed. The structure
Tabs contains tab id, parent window id, and the URL accessed by the user.
Whenever a user tries to sign in via. SSO, the SSO Manager gathers information
from the browser and stores it in its schema. The functionality of the SSO Manager
module is described in the following sections.
6.2 SSO Manager Module Implementation : Initial Sign-in

Scenario
Listing 1.1 Database Entry after IDP Session Initiation

1 {
2 " IDP Name " : " U n i v e r s i t y A - ID and Auth " ,
3 " IDP T i m e o u t T h r e s h o l d " : " 4 5 m "
4 " IDP Sign - out URL " : " https : // univA . com / auth / s i g n o u t " ,
5 " SP Info " : [ ]
6 }
Listing 1.2 Database Entry after SP Session Initiation

1 {
3 " IDP T i m e o u t T h r e s h o l d " : " 4 5 m " ,
5 " SP Info " : [

6 {
7 " SP Name " : " S m a r t S t u d y LMS " ,
8 " SP T i m e o u t T h r e s h o l d " : " 3 5 m " ,
9 " SP Sign - out URL " : " https : // lms . s s t u d y . com / s i g n o u t " ,
10 " Tabs " : [
11 {
12 " Tab ID " : " 0 0 1 " ,
13 " P a r e n t W i n d o w ID " : " 0 0 1 " ,
14 " URL " : " https : // u n i v A . s s t u d y . com / h o m e "
15 }
16 ]
17 }
18 ]
19 }
Let us consider a scenario of user X, a student of university A. To access the

services such as learning management system, enrollment financial systems, shared
file repository, X has to use the university id. Here, the university is the IDP.
X logs into a computer’s web browser and tries to access the learning management
system (LMS). When X tries to access the university’s LMS URL, the LMS as a
service provider redirects the browser to university’s Identity and Auth Platform
(IDP).
Once the IDP session is created, the SSO Manager creates a new record in
the browser’s IndexedDB. The SSO Manager records the following information in
the IndexedDB. And during the sign-in process, after the SP session is created, in
Sect. 4.1, the record is updated as follows.
Hence after the sign-on is complete, we could infer that SSO Manager recorded
the information in IndexedDB of the web browser. A sample entry for the above
student sign-on scenario is depicted in Listing 1.2.
6.3 SSO Manager Module Implementation : Additional Tabs
In continuation of above scenario, user X would like to access assignments and course
information simultaneously on two browser tabs. So she opens the LMS via another
tab. The SSO Manager recognizes the event of open tab and modifies the existing
entry to add the information about newly opened tab.
Listing 1.3 shows the additional information that has been added about the newly
opened browser tab.
Listing 1.3 Database Entry with additional tabs
1 {
5 " SP Info " : [
6 {
10 " Tabs " : [
11 {
12 " Tab ID " : " 0 0 2 " ,

13 " P a r e n t W i n d o w ID " : " 0 0 1 " ,
14 " URL " : " https : // u n i v A . s s t u d y . com / c o u r s e s "
15 },
16 {
17 " Tab ID " : " 0 0 3 " ,
18 " P a r e n t W i n d o w ID " : " 0 0 1 " ,
19 " URL " : " https : // u n i v A . s s t u d y . com / a s s i g n m e n t s "
20 }
21 ]
22 }
23 ]
24 }
Listing 1.4 Database Entry with multiple SPs

1 {
5 " SP Info " : [
6 {
10 " Tabs " : [
11 {
12 " Tab ID " : " 0 0 2 " ,
13 " P a r e n t W i n d o w ID " : " 0 0 1 " ,
14 " URL " : " https : // u n i v A . s s t u d y . com / c o u r s e s "
15 },
16 {
17 " Tab ID " : " 0 0 3 " ,
18 " P a r e n t W i n d o w ID " : " 0 0 1 " ,
19 " URL " : " https : // u n i v A . s s t u d y . com / a s s i g n m e n t s "
20 }
21 ]
22 }
23 {
24 " SP Name " : " E n r o l l m e n t A p p l i c a t i o n " ,
26 " SP Sign - out URL " : " https : // u n i v A . e n r o l l . com / s i g n o u t " ,
27 " Tabs " : [
28 {
29 " Tab ID " : " 0 0 4 " ,
30 " P a r e n t W i n d o w ID " : " 0 0 1 " ,
31 " URL " : " https : // u n i v A . e n r o l l . com / h o m e "
32 }
33 ]
34 }
35 ]
36 }
6.4 SSO Manager Module Implementation : Additional

Services
Similar to the additional tabs, additional services are handled by the SSO Manager.
In the scenario, where the user would like to access another service leveraging the
same IDP, the information is updated.
If the user X opens the another service (SP2 ) on a new tab, the service application
does not requires the user to sign-in again. It leverages the information available
from existing active IDP session.
Here, in this scenario, the SSO Manager module adds the information of SP2 to
the SSO Sessions Info entry. The sample entry is listed in Listing 1.4.
6.5 SSO Manager Module Implementation : Sign-out

Scenario
The key purpose of the SSO Manager is to efficiently manage the sign-out scenarios.
If the user signs out of a service from a tab, SSO Manager interprets the event and
displays a prompt to the user about other active tabs of the same service provider.
In the student user example, if the student chooses to sign out from the assignments
tab, SSO Manager prompts a message for input from the user regarding all the active
tabs and services for the IDP. Once the user confirms about signing-out of all the
services, SSO Manager removes the SSO Sessions Info entry, once the user signs
out from the IDP and SPs.
SSO Manager can also handle situations that arise from accidental closing of
browser or tab windows. If there are no browser windows in open state, the SSO
Manager can be configured to invoke the sign-out routine of the IDP and all the
respective services.
7 Conclusion
With growing cybersecurity attacks in the digitized world, everyday hackers and
malicious actors are looking for exploiting access credentials of naive users. Getting
an access to one user’s account opens a multitude of possibilities for hackers, such
as phishing attacks, ransomware attacks, data corruption, and so on. In this paper,
we analyzed single sign-on (or federated identity management) process is handled
in tabbed browsing environment, identified potential security compromise scenarios,
and proposed a solution to avoid such scenarios. Therefore, it is essential to com-
pletely understand the process of single sign-on and eliminate the gaps for a safer
access control.
References
1. Hardt, D.: The OAuth 2.0 authorization framework, RFC 6749. Internet Engineering Task Force
(2012). https://tools.ietf.org/html/rfc6749
2. Kemp, J., Cantor, S., Mishra, P., Philpott, R., Maler, E.: Assertions and protocols for the OASIS
security assertion markup language (SAML) v2.0. OASIS (2015). http://saml.xml.org/saml-
specifications
3. Ramamoorthi, L., Sarkar, D.: Single sign-on demystified: security considerations for developers
and users. In: Rocha, Á., Adeli, H., Reis, L.P., Costanzo, S. (eds.) Trends and Advances in
Information Systems and Technologies, pp. 185–196. Springer International Publishing, Cham
(2018)
4. Dubroy, P., Balakrishnan, R.: A study of tabbed browsing among Mozilla Firefox users. In:
Proceedings of the SIGCHI Conference on Human Factors in Computing Systems, CHI ’10, pp.
673–682. ACM, New York, NY, USA (2010)
5. Mozilla developer network - web storage API. https://developer.mozilla.org/en-US/docs/Web/
API/Web_Storage_API
6. W3c docs - web cryptography API (2017). https://www.w3.org/TR/WebCryptoAPI/
7. Sun, S.T., Pospisil, E., Muslukhov, I., Dindar, N., Hawkey, K., Beznosov, K.: What makes users
refuse web single sign-on?: An empirical investigation of openid. In: Proceedings of the Seventh
Symposium on Usable Privacy and Security, SOUPS ’11, pp. 4:1–4:20. ACM, New York, NY,
USA (2011)
Cybernetic Dependency Capacity
Jorge Barbosa
Abstract The state of development of the information society of a particular country

can have a great influence on the outcome of a cyberwarfare in which it is involved.
This influence, in the possible outcome of a war of this type, can be positive or nega-
tive. In addition, contrary to what might initially be thought, a low development of the
information society can be advantageous to obtain a victory, if one strictly considers
the scenario of cyberwar and the countries in confrontation. This way we analyze
this influence and relationship between the information society and cyberwarfare
in the scope of security and defense. We designate this dependency as Cybernetic
Dependency Capacity. This dependence is a very important factor in the results of
cyberwarfare actions and can be positive or negative for both technologically devel-
oped and less-developed countries. For these reasons, the destiny of any country,
within the scope of a cyberwar, is very influenced by the dependence of this coun-
try of their information society. As a major development of the IT society in one
country will not make it theoretically victorious in a cyberwarfare with a country
less IT developed. This fact, which is apparently paradoxical, is because the larger
information society of a country is more informatics systems have to exist. Then, if
much more informatics systems exist, more cyber-doors are open for cyber-attacks,
because the probability of these systems will be unprotected, from the point of view
of cyber-attacks, increases with the growth in the number of computers systems. For
the same reason, the countries with less dependence on the information society do not
have this cyber-permeability and may be more cyber-resilient. In the limit, theoret-
ically, the least dependent cyber-country can be the winner and the most dependent
and as such, technologically more developed, be the loser in a cyberwar between
the two. These scenarios are also applicable to non-states and terrorist organizations
and for these reasons should be analyzed and considered in the definition of the
cyber-defense policies of any country, namely in the factors like a Cybernetic Attack
Capacity and a Cybernetic Dependency Capacity.
J. Barbosa (B)
Coimbra Polytechnic – ISEC, Coimbra, Portugal
e-mail: jorge.barbosa@isec.pt
Portuguese National Defense Institute, Lisbon, Portugal

https://doi.org/10.1007/978-981-13-9155-2_3
30 J. Barbosa
Keywords Cyber Dependence · Cyber-Warfare · Cyber-Defense ·

Cybersecurity · Cyber Resilience · Information Society
1 Introduction
The benefits to civil societies resulting from the massive use of information tech-
nology in all aspects of that society, namely service provision, government, central
and local administration, banking and finance, transport management, logistics, food
and fuel distribution, etc. are now an undisputed reality in all minimally developed
countries. It can even be said that it is no longer possible at the present for those
countries to survive without such means and facilities provided by the information
society.
The maintenance of the information society and the means necessary for its oper-
ation are therefore of paramount importance. Their allocation or destruction is there-
fore unthinkable in the current development context.
Cyber-criminal activities are usually associated with common cybercrime, i.e.,
activities related to the cybersecurity of individuals or companies. However, similar
actions, with a larger scale and scope, can be used in cyber-actions against state
sovereignty, known as cyberwarfare actions.
Actions, even relatively small and isolated in their range and effects, such as acts
by hackers, which either individually or in small groups, that trigger cyber-actions
against individuals or companies, can have relatively serious consequences.
The issues of great concern to many security and defense agencies and services are
that they do not come from these actors, but are orchestrated and unleashed by other
actors, notably countries, which have specialized units. If concerted and large-scale
action is taken against a country with the exclusive use of computers means, it can
profoundly affect the IT systems of that other country, in particular, those linked to
its critical infrastructures. The situation could become quite serious for the country
under attack and may even lead to its paralysis, subjugation and loss of national
sovereignty.
Cyberwarfare [1, 2] designates this type of action. Given its intimate connection
with the information society, we will then analyze the implications that a lesser or
greater dependence and development of the IT may have in the countries, in case to
be targeted of a cyberwar.
Those cyberwarfare actions can affect not only to military targets and to instal-
lations but also to civilian targets, such as critical infrastructures linked to the func-
tioning of modern civil societies, seriously affecting the lives of such societies. Even
if there is no great physical destruction, both civilian and military, and therefore the
losses do not seem to be significant because they are not visible, for the common
citizen this “modern” war can be as harmful as the classic wars that occurred in past
centuries. Their daily and habitual experiences may be deeply affected, in particular,
by taking away the benefits and comforts they currently enjoy.
Cybernetic Dependency Capacity 31
2 Known Cyber and CyberWarfare Actions
In what concerns to security and defense of states, and in one conventional perspec-
tive, the most relevant actions are those that are carried out with the aim of triggering
actions that undermine the designated sovereignty of the states. In this perspective,
two main groups of actions are considered: Those that can be considered actions of
common criminality, although they can be quite disturbing of the daily experience,
like actions of domestic terrorism; or the actions that disturb the order and the civil
life of the society of that country to the point of endangering the sovereignty of the
states. The border between these two types of actions may be tenuous and so it is
often difficult to establish.
In most democratic states, according to this separation, the first set of these actions
are dealt with by police or internal security agencies (and in the case of actions with
external origin in collaboration with other external security organs). For the second
set of actions, as they are detrimental to the national sovereignty of the states, are
treated by the organs that defend this sovereignty, that is, by the armed forces of the
states.
This model is also usually followed with regard to actions triggered by the use
of cybernetic means, i.e., actions triggered in cyberspace. For the particular char-
acteristics of this environment, there are also very particular and specific consider-
ations that have to be addressed. First, the difficulty of separating and classifying
the actions initiated soon in the establishment of the classification frontier of such
actions. Physical borders, as in the case of conventional external aggressions, do
not limit actions in cyberspace. In addition to the possible use of entities directly
linked to foreign governments, organizations or individuals acting on their own may
also trigger cyber-actions that threaten national sovereignty. They will not be sim-
ple cybersecurity actions, but cyberwarfare actions. The means used in common
cybersecurity actions and in cyberwarfare actions may be similar, but the scope and
latitude of cyberwarfare put the sovereignty of the states into question, hence being
classified in this particular way. Thus, many states have developed within their con-
ventional armed forces cyber corps to deal with such cyber-actions. At the same time,
the internal intelligence and security forces have developed specialized units to deal
with common cybersecurity actions.
In the case of national sovereignty, it may be understood by the states affected
by such actions, not to divulge much the occurrence of cyber-actions of which they
were targeted. In this way, there is not much information about such acts and what
there is may not be reliable. In the literature, by its impact and apparent, there are
references to some of these actions, indicated as having been perpetuated by groups
or individuals against other states or perpetuated states against other states or even
regions of the same state, in situations of, for example, attempts of political separation
by these regions. Many other cyber-attacks are may be due to the rise of “no state”
hackers groups.
The most referred attack methods used in the actions identified with “no state”
groups as Traffic Redirection and Web sites defacement. The most attack methods
32 J. Barbosa
identified with the actions of states are distributed denial of service (DDoS), viruses
and trojans and a specific cyberweapon, such as those used in the famous action
against Iran known as Stuxnet, [3]. The DDoS attacks are also used by “no state”
groups.
In the literature, [1, 4, 5] are reported many cyber-attacks actions: (1) USA and
South Korea: On the July 4, 2009; (2) Iran: During the disputed Iranian presidential
elections of June 14, 2009 and the most cyber-action know by Stuxnet; (3) Tatarstan:
In June 2009, the president of Tatarstan’s Web site was knocked offline and Internet
access was lost; (4) USA: On April 21, 2009, the Wall Street Journal reported that
security around the Pentagon’s multi-billion-dollar Joint Strike Fighter project was
compromised; (5) Kyrgyzstan: On January 18, 2009, a DDoS attack shuttered two
to three of the nation’s four ISPs for several days; (6) Israel and the Palestinian
National Authority: In December 2008 along with Israel’s military action designated
Operation Cast Lead against Hamas; (7) Zimbabwe: In December 2008, a action
reported by Concerned Africa Scholars; (8) Myanmar: On September 23, 2008, the
government launched DDoS attacks against three Web sites that support the monks
and many others.
The above actions can be considered at the border between common cybersecurity
actions and cyberwarfare actions. Better separation will be achieved if it is precisely
determined those as put in danger the national sovereignty of the affected states.
Many other cyberwarfare actions are allegedly attributed to Russia against Estonia,
Georgia and Ukraine. These cyberwarfare actions were triggered using DDoS, SQL
Injection, SQLi, and cross-site scripting (XSS) attacks.
In the strict sense of cyberwarfare, are used specific cyberweapons. These cyber-
weapons are developed based on the exploration of computer systems vulnerabilities,
namely not yet known vulnerabilities. These vulnerabilities are known as zero-day
vulnerabilities, ZDV, and based on these vulnerabilities are developed specific appli-
cations named zero-day exploits, ZDE, which can affect and even paralyze the com-
puter system affected by this ZDV. The most known cyberwarfare action, known
as Stuxnet, used cyberweapons, which exploit ZDV’s in the computer systems of
nuclear centrifuges of Iran, [3].
3 Cybernetic Dependency Capacity
As far as conventional wars are concerned, on a primary level, the military and
political strategists and leaders of a country equate the potential, in the face of a
possible and determined enemy, through two factors: the “Attack Capacity” and the
“Defense Capacity” of both your country and that of the eventual enemy country.
From the consideration of these capacities, they infer who, in the case of a con-
ventional war between these two countries, theoretically, can become the winner or
the loser of the same.
In the case of a cyberwar, these considerations have to be different, given the
much more complex scenarios. The key factors to be taken into account should be
not only the two above-mentioned, adapted to the cyberwar, but three factors, namely:
“Cybernetic Attack Capacity,” “Cybernetic Defense Capacity” and a third and new
factor regarding conventional warfare, the “Cybernetic Dependency Capacity” of the
countries involved [6].
Then, the Cybernetic Dependency Capacity of a country, as the dependency that
this country and its society have on the IT organization and the complexity and
density of connections and interoperability between their IT systems. The larger this
organization and the complexity and density of connections of its computer systems,
the greater the country’s dependence on IT and the greater vulnerability to eventual
computer crashes. A serious computer fault in an important critical computer system
can be very severe for any modern country and, in extreme cases, can even paralyze
the country or sectors thereof.
Modern and developed Western societies, as well as those of some Asian countries,
have a very high Cybernetic Dependency Capacity and as such are extremely sensitive
to problems in their computer systems, in particular, those resulting from cyberwar.
Less-developed societies are normally often underdeveloped in terms of the mas-
sive use of information technology in the various organizational aspects of that soci-
ety. This causes them to have a very low dependency cybernetic capacity. Because
they do not rely heavily on computing for their day-to-day operations, they are not
so sensitive to computer glitches, nor do they have computer “targets” to attack.
4 Influence of Information Society in a Cyber Warfare
It can even be said that it is no longer possible, at present, for the countries to survive
without such means and facilities provided by the information society.
Then, the maintenance of the information society and the means necessary for its
operation are, therefore, of paramount importance. Their allocation or destruction is
therefore unthinkable in the current development context. Actions, even relatively
small and isolated in their range and effects, such as acts by hackers, which either
individually or in small groups, trigger cyber-actions against individuals or compa-
nies, can have relatively serious consequences, [1, 7].
The issue of great concern to many security and defense agencies and services is
that they do not come from these actors, but are orchestrated and unleashed by other
actors, notably countries, which have specialized units. If concerted and large-scale
action is taken against a country by means of the exclusive use of computerized
means, it can profoundly affect the IT systems of that other country, in particular,
those linked to its critical infrastructures, [5, 8].
Given its close dependence, the level of development of the information society
in a given country may make it very likely to be heavily affected by cyberwarfare.
This aspect is amplified by the fact that even in technologically advanced countries
and, as such, very dependent on computer systems for the most diverse and banal
activities, a large part of its population still has a very big illiteracy in computer
34 J. Barbosa
operation. This makes many of the computer systems unprotected or poorly protected
by seemingly harmless actions of their users and as such permeable to cyberwarfare.
The vulgar actions of hacking are often minimized and typical considered as
actions in the scope of cybersecurity, related to attacks against personal or corporate’s
computers. However, they may be hiding real cyberwarfare actions. In particular,
they may be masking pre-preparation actions for future cyberwarfare actions, such
as DDoS attacks on government, military or critical infrastructure installations. The
computer systems thus affected stay then in standby. If they necessary to use them
for a cyberwar will already be prepared.
Due to these minimizations of the importance of the attacks or their misclassifica-
tion, this preparation is made without the knowledge of the legitimate owners of these
systems. This failure to correctly consider or classify such acts, further increases the
criticality and permeability of systems greatly affecting the Cybernetic Dependency
Capability.
Thus, a country’s less or greater dependence on the information society and the
computer skills of its population, namely your education and sensitivity for the
computer security, has a direct implication on the degree of cyberwarfare permeability
of the cyber-attacked countries.
The situation could become quite serious for the country under attack and may
even lead to its paralysis, subjugation and loss of national sovereignty, [4].
5 Cybernetic Dependency Capacity and Cybernetic

Resilience
Considering only the cyberwarfare, i.e., not considering conventional wars at all, due
this great dependence of the information society, an underdeveloped country can be
a cyberwar winner and a highly developed country can be a loser in the same war,
[6]. This is a very interesting paradox.
This may occur if the less-developed country has minimal cyber-attack and cyber-
defense capabilities in such a way that it can inflict significant damage on the com-
puter systems, especially the critical ones, in the opposing country. This may also
have good cyber-attack capabilities, but as the other country does not have a min-
imum of IT structures, they do not have what to attack in the other less-developed
country in order to paralyze it from a computer point of view. However, the developed
country, attacked in the cyberwar, can get their systems deeply affected and since it
is very dependent on these systems, it can even paralyze.
Of course, the consideration of the conventional war component in this equation
can greatly alter this end result, especially if, as usual, the most developed country
has minimum conventional warfare capabilities and so can retaliate conventionally
and inflict significant damage on its opponent. However, this question about the
use of conventional means of warfare can be critical if the two countries are not
geographically close. The displacement and positioning of conventional media can
be difficult, time-consuming and late in the theater. Military logistics may have even
been affected by cyber-attacks, which would further affect conventional operations.
From a mere speculative point of view and considering some recent developments
in relations between these two countries, such a situation could hypothetically occur
in a strictly cybernetic confrontation between the USA and North Korea.
American society is a lot, if not totally, dependent on information technology for
its functioning. For its part, and to the best of our knowledge, North Korean society
is not.
In the country, the informatics is considered incipient and even the small use of
computing means is practically done by the governmental members, which are very
controlled.
In addition, the computer connected to the outside of this country is practically
non-existent, which means that there are practically no “doors” to the entry of cyber-
attacks. According to intelligence, North Korea has relatively developed cyberwar-
fare capabilities and there are several specialized and highly trained military units
for cyberwarfare, i.e., cyber-acts. The hypothetical considerations made in respect
of these two countries could be made on a few others.
An obvious example of the influence of Cybernetic Dependency on cyberwar-
fare is the case of Estonia. Shortly after Estonia became independent of the Soviet
Union, it promoted a policy of development and modernization of the country. One
of the biggest stakes was the massive introduction of information and knowledge
technologies. With this policy, it has become one of the countries with most modern
computerized public services. Numerous public administration services, banks and
services have emerged, based on computer systems with which Estonian citizens
can interact directly. Considering Cybernetic Dependency, Estonia became a coun-
try with a high cybernetic capacity for dependency, i.e., a country potentially very
vulnerable to cyber-attacks. On 27 April 2007, cyber-actions was launched in the
form of violent DDoS attacks targeting a number of IT systems in Estonia, including
important government Web sites, parliament, banks and the communications sys-
tem. These actions became known as the “Estonian Cyberwar.” It is assumed that for
these DDoS attacks, a net bot was created consisting of 85,000 servers having the
attacks lasted three weeks and having been attacked sixty Web sites and the almost
stoppage of the country, [3]. These cyber-actions have never been officially assigned
to any country. However, the Estonian Foreign Minister attributed the attacks to Rus-
sia, which always denied their official participation in them and considered them as
actions of individuals who acted as “Patriotic Hackers.” Eventually, if the country
had not promoted this development policy of its IT, it would not have become a
desirable and possible target for these actions, which were facilitated by the fact that
the country had a high Cybernetic Dependency Capacity.
The use of computerized means in cyberwars could also be considered by non-
states or by terrorist organizations. Theoretically, these states or organizations do not
have the capabilities they can use to trigger conventional attacks to countries that are
minimally capacitated from a conventional point of view. This lack of conventional
capabilities, on the part of such countries and organizations, could result from the
36 J. Barbosa
lack of means, structures, organization, and even logistics to trigger conventional

attacks, conduct, and sustain them for as long as such conventional wars lasted.
On the contrary, it may theoretically be relatively easy for them to launch cyber-
wars against some countries, even large conventional powers. The fact that they are
not highly structured and not very dependent states or organizations leads, according
to the previous analysis, to a very low or even zero Cybernetic Dependency Capac-
ity because they do not have behind a civil society and even military organization
dependent on the modern information technologies. As such, they are not very likely
to be, as well, cyber attacked.
Another feature that could create interest in this type of non-states and especially
in terrorist organizations is that their targets, in addition to their immediate areas of
operation and influence, usually extend to other countries, especially Western ones.
As these latter countries are not generally geographically close to the theaters of
action of such organizations, the use of cyber-media would be more viable for these
organizations as well.
From the operational point of view, however, there would be many constraints that
we consider impeding or at least very restrictive in triggering large-scale cyberwarfare
against these countries, by these organizations. First of all, the need for human and
technological resources, which are fundamental to trigger these large-scale actions,
would be very large, and as such it would be very difficult for them to operationalize
such actions. In other words, such states and organizations theoretically have a very
low or even zero cyber-attack capacity.
However, this aspect will not be a deterrent or an obstacle to the launching of
small-scale cyberwarfare, which will nonetheless be disturbing in the societies of
the countries under attack. Cyber-attacks well targeted and directed at critical infras-
tructure in any country can be quite nefarious. We could compare these types of
actions to nuclear attacks. For similar reasons, non-states or terrorist organizations
do not have the capacity to launch large-scale nuclear attacks. However, theoreti-
cally, they could launch isolated and limited critical actions, for example, using the
so-called dirty bombs, which are not necessarily intended to cause large-scale phys-
ical destruction, but only to contaminate an area use for a long period and hinder the
normal functioning of the country. These organizations can theoretically and with
the same objective launch cyber-“dirty bombs” actions.
These non-states and organizations have to be monitored from the point of view
of their cybernetic activities and cannot be neglected by the cyber-defense structures
of the countries. Policies to define the cyber-defense capacity should consider these
possibilities and be a concern to be taken into account, in order to make more resilient
societies organized and very dependent on the information society given the great
Cybernetic Capacity of Dependence of such societies.
In the same way as in common cybersecurity actions, protective measures can be
used in computer systems and action and prevention plans established [2, 5, 8].
The difficulty with regard to cyberwarfare is the unpredictability of such actions
and the need to, quickly, take decisions that stop or at least minimize the effects of such
actions, in particular, if are attacked Critical Infrastructures, such as hydroelectric
dams, banking systems and so on.
However, the main problem concern with cyberwarfare, and what the usual mea-
sures of protection are difficult to prevent, is the use of specific weapons against
the computer systems that control these critical structures, in particular the cyber-
weapons developed to exploit the vulnerabilities not known in those systems, i.e.,
the zero-day exploit. It may even be impossible to take preventive and protective
measures for these cyberweapons.
The use of ZDE is a major challenge to the cybernetic resilience of any country.
6 Conclusion
The advantages to civil societies of countries, resulting from the massive use of
information technology, are now an indisputable reality in all minimally developed
countries. One can even say that it is no longer possible for modern civil societies to
survive without the means and facilities provided by the information society.
Then, the maintenance of the information society and of the means necessary
for its operation is therefore of paramount importance, since the countries could be
paralyzed by their affectation or even their destruction.
This dependency has created what we call as Cybernetic Dependency Capacity.
This factor, in our approach, allows us to consider the influence in the daily experience
of a civil society due to the affectation, by cyberwarfare actions, of country’s computer
systems, namely those related to the control of its critical infrastructures. The greater
this Cybernetic Dependency Capacity, the more exposed that country is to the effects
of actions affecting those systems. Generally, and only from the point of view of
cyberwarfare, a country very dependent on the information society is, maybe, much
more vulnerable to these actions than others less dependent.
This conclusion leads us to another, which translates into a very interesting para-
dox: strictly considering cyberwar, that is, not considering conventional war at all,
an underdeveloped country can be a winner in a cyberwar and a developed country
may be losing in this same war.
The possible use of cyberweapons by non-states or by terrorist organizations was
analyzed. From this analysis, we concluded that its use, by such organizations is quite
possible and may be desirable. The use of these means to attack certain countries
would be more viable to them, than the use of means of conventional warfare. As
the Cybernetic Dependency Capacity of such non-states and organizations is very
low if not null, cyber-retaliation against these countries and organizations in order
to cause them harm would be practically impossible. Then, the very low Cybernetic
Dependency Capacity becomes an advantage for such groups. However, such coun-
tries would have to overcome many constraints in order to trigger cyber-actions on a
large scale, which is virtually impossible for them. In this way, their cyber-actions, to
occur, should be limited to isolated attacks on well-selected targets, namely critical
infrastructures, which would nevertheless create difficulties for modern societies,
especially Western ones, strongly based on information society and as such with a
great Cybernetic Capacity of Dependency. Then, the cybernetic activities of these “no
38 J. Barbosa
states” groups and organizations have to be permanently monitored by cyber-defense

intelligence.
The possibility of using specific cybernetic weapons based on the vulnerabilities
of the computer systems, namely vulnerabilities known as zero-day vulnerabilities,
ZDV, which allow the development of applications, called zero-day exploit, ZDE,
that, using these vulnerabilities, possibility the attack and even take the control of
the affected computer systems, are the main concerns in the field of cyber-defense.
It may even be impossible to take preventive and protective measures for these
cyberweapons, making the use of ZDE with cyberweapons the biggest challenge to
the cybernetic resilience of any country.
This way, due to this high dependency, in the civil and military areas, of the
information society, the policies of any country should include in the cyber-defense
strategies the consideration of your Cybernetic Capacity of Dependence, as a concern
to be taken into account in order to make the societies more cybernetic resilient.
Acknowledgements I thank the Portuguese National Defense Institute as well as the ISEC/IPC
for the support and facilities granted.
This paper is based on the work Final Report of National Defense Course, done by the author in
the Portuguese National Defense Institute.
References
1. Carr, J.: Inside Cyber Warfare, 2nd edn. O’Reilly Media Inc., Sebastopol, CA, USA (2012)
2. Ranger, S.: What is cyberwar? Everything you need to know about the frightening future
of digital conflict. https://www.zdnet.com/article/cyberwar-a-guide-to-the-frightening-future-
of-online-conflict/ (2018). Acedido em 25 Setembro 2018
3. Zetter, K.: Countdown to Zero Day: Stuxnet and the Launch of the World’s First Digital Weapon.
Crown Publishers, NY (2014)
4. Clark, R.A., Knake, R.K.: Cyber War: The Next Threat to National Security and What to do
About. HarperCollins Publishers Inc, New York (2010)
5. Kostopoulos, G.K.: Cyberspace and Cybersecurity. CRC Press, Taylor & Francis Group, New
York (2013)
6. Barbosa, J.:Pequenas potências militares convencionais, Grandes potências militares cibernéti-
cas - Abordagem da utilização de meios informáticos na defesa/ataque militar moderno. IDN,
Lisboa (2018)
7. Singer, P.W., Friedman, A.: Cybersecurity and Cyberwar—What Everyone Needs to Know.
Oxford University Press, NY (2014)
8. Wittkop, J.: Building a Comprehensive IT Security Program. Apress, Bolton, USA (2016)
9. Ayala, L.: Cyber-Physical Attack Recovery Procedures: A Step-by-Step Preparation and
Response Guide. Apress, New York (2016)
Portuguese Concerns and Experience
of Specific Cybercrimes:
A Benchmarking with European Citizens
João Vidal Carvalho, Álvaro Rocha, António Abreu and Avelino Victor
Abstract The increasing use of a variety of devices that can access the Internet has
contributed to the overall growth of Internet use. The growth of Internet use also leads
to a general increase in cybercrime concerns as well as the number of users’ victims
of this permanent threat. This reality is also very present in Portugal where there is a
sense of lack of control and combat of this kind of cyber-threats. In this context, it is
important to know some indicators associated with cybercrime, in order to take the
best measures/actions in a reasoned way, to combat this phenomenon. This article
presents the results of a study carried out in 2017 by the European Commission,
whose analysis helps to understand the Portuguese positioning in the concerns and
experiences about cybercrime, in relation to the average of the European Commu-
nity citizens. Analysing this study, it was possible to perceive that the Portuguese’s
concerns about cybercrime are in line with the average of the European citizens,
although the percentage of Portuguese experienced these attacks is, for now, below
average. Based on these results, it is possible to identify some public policies that
aim to mitigate this phenomenon.
Keywords Cyber security · Cybercrime · Cyber-threats · Information security ·

Online fraud · Scam · Cyber extortion
J. V. Carvalho (B) · A. Abreu

Politécnico do Porto, ISCAP, CEOS.PP, São Mamede de Infesta, Portugal
e-mail: cajvidal@iscap.ipp.pt
A. Abreu
e-mail: cajvidal@iscap.ipp.pt
Á. Rocha
Departamento de Engenharia Informática, Universidade de Coimbra, Coimbra, Portugal
e-mail: amrocha@dei.uc.pt
A. Victor
Instituto Universitário da Maia, Instituto Politécnico da Maia, Maia, Portugal
e-mail: javemor@ismai.pt

https://doi.org/10.1007/978-981-13-9155-2_4
40 J. V. Carvalho et al.
1 Introduction
It is generally accepted that computerization and the introduction of information

systems and technologies (IST) in all social activities significantly increase the effi-
ciency and effectiveness of the activities of each person or organization. In addition,
this process enhances the quality of life of an individual through the ability to access
from any place and at any time, essential information in services such as electronic
administration, electronic health, education, leisure and shopping. However, the con-
sequent opening of databases and information systems of public administration and
other business entities inevitably leads to the possibility of such openness being
exploited by malicious organized individuals or groups.
The possibilities for abuse are large ranging from unauthorized access to system
data and resources, to identity theft of users to actions that significantly or completely
degrade the functionality of information systems. Such activities not only affect users
and information systems that are the target of the attack, but to a greater or lesser
extent undermine the confidence of users and business entities, which ultimately
leads to a slowdown in technological development. Indeed, the advancement of IST
and digital economy has great potential for the well-being of citizens and for business
growth, but this new paradigm also entails cyber security challenges that can have
high economic impacts. Cybercrime is estimated to cause the loss of billions of
euros per year and is placing an increasing strain on law enforcement response
capability [1].
Cybercrime is a borderless problem, consisting of criminal acts that are commit-
ted online by using electronic communications networks and information systems,
including crimes specific to the Internet, online fraud and forgery, and illegal online
content such as child pornography [2]. It is estimated that more than one million peo-
ple worldwide become victims of cybercrime every day [3]. The European Union is
leading the effort to regulate defence against cyber threats in both normative/legal
and strategic areas. Through this strategy, the European Commission sought to create
legislation to criminalize such crimes, increase cyber security capacity and promote
the exchange of information between countries (in particular on recorded incidents)
[4]. In fact, given the development of cybercrime in recent years, the European Com-
mission has designed a coordinated policy in close cooperation with European Union
member states and the other European Union institutions.
As part of the European Community, Portugal is highly involved in this European
strategy. In this sense, also in Portugal there is a growing concern about cybercrime,
both at the level of the state and at the citizen level. Nevertheless, it is important
to understand how the Portuguese citizens face this new paradigm and especially to
understand how far they are compared to other Europeans. This comparison is made
at the level of concerns about cybercrime (and some of the most recurrent threats
associated with this concept) as well as the experiences they have experienced in
the recent past. The analysis of this comparison will allow Portugal to position itself
against the other European countries in this sensitive and important area for today’s
citizens and businesses. In sum, the study presented here examines the experience
Portuguese Concerns and Experience of Specific Cybercrimes … 41
of cybercrimes that EU citizens and the Portuguese in particular have and the level
of concern they feel about this type of crime.
In this article, we will initially present the description of a study carried out in
Europe in 2017 about cyber security. In the next section, the values of concerns
and experiences of ten cybercrimes will be compared between Portuguese and the
European citizen’s average. Finally, we will present the discussion about the results
of the study, with a view to identify measures that have mitigated this phenomenon.
2 Cyber Security Study in Europe
The Special Eurobarometer series on cyber security is the most important resource for
learning about cybercrime in Europe [1]. The importance of this resource stems from
the treatment and analysis of representative data of different types of cybercrimes
collected in the last five years of the 28 member states of the European Commu-
nity. The most recent report covers a wide range of threats and aims to understand
European citizens’ experiences and perceptions of cyber security issues. That is,
the survey adopted in this report analyses the nature and frequency of Internet use
by citizens; their awareness and experience of cybercrime; and the level of concern
they feel about this type of crime. In this paper, we analyse only the perceptions of
Internet users to cyber security, whether they have experienced or been a victim of
cybercrime and the level of concern they feel about it.
This survey adopted in this report was carried out between the 13th and the 26th
June 2017, by TNS opinion & social,1 carried out the wave 87.4 of the Eurobarometer
survey, on request of the European Commission. The wave 87.4 covers the population
of the respective nationalities of the European Union member states, resident in each
of the 28 member states and aged 15 years and over. In total, 28,093 respondents
(1075 from Portugal) from different social and demographic groups were interviewed
face-to-face at home in their mother tongue on behalf of the Directorate-General for
Home Affairs. The methodology used is that of Eurobarometer surveys as carried out
by the Directorate-General for Communication (“Strategic Communication” Unit).2
The findings from this survey update a previous survey which was carried out in
2013 [5] and 2015 [2]. The 2017 survey repeats most of the questions asked in 2015
in order to provide insight into the evolution of knowledge, behaviour and attitudes
towards cyber security in the European Union.
1 TNS opinion & social is a consortium created between TNS political & social, TNS UK and TNS
opinion.
2 http://ec.europa.eu/public_opinion/index_en.htm.
2.1 Questions Used in the Study for Cyber Security Concerns

and Experience
In order to identify the cyber security indicators applied in the survey, we have
relied on the most important types of cybercrime. Table 1 presents these different
types of cybercrime and a summary description of each. It should be noted that these
descriptions were presented to the respondents in the same terms. To assess the levels
of concern of respondents, they were asked how concerned they were about falling
victim to cybercrimes. Thus, the following question was asked:
“Cybercrimes can include many different types of criminal activity. How concerned
are you personally about experiencing or being a victim of the following situations?”
Having been asked about their concerned of cybercrimes, respondents were then
asked whether they had been the victim of any such crimes. Thus, the following
question was asked:
“How often have you experienced or been a victim of the following situations?”
Table 1 Description of cyber security indicators

Cyber security indicator Description
Identity theft Somebody stealing your personal data and
impersonating you
Scam emails or phone calls Receiving fraudulent emails or phone calls
asking for your personal details (including
access to your computer, logins, banking or
payment information)
Online fraud Online fraud where goods purchased are not
delivered, are counterfeit or are not as
advertised
Offensive material and child pornography Accidentally encountering child pornography
online
Material promoting racial hatred or religious Accidentally encountering material which
extremism promotes racial hatred or religious extremism
Access to online services Not being able to access online services like
banking or public services because of
cyber-attacks
Email account hacking Your social network account or email being
hacked
Online banking fraud Being a victim of bank card or online banking
fraud
Cyber extortion Being asked for a payment in return for
getting back control of your device
Malicious software Discovering malicious software (viruses, etc.)
on your device
In any of the questions, the “following situations” refer to the ten different types
of cybercrime that were applied in this survey and are presented in the following
table.
2.2 Concerns with Cybercrimes
Europeans’ concerns about cyber security have been growing over the last few years.
In fact, almost nine in ten (87%) respondents see cybercrime as an important chal-
lenge, a significant increase in the 80% recorded in March 2015. The rise is even more
significant when looking at the proportion of respondents who see cybercrime as a
very important challenge: 56% compared with 42% in 2015. Less than half (49%)
of the respondents agree or mostly agree that law enforcement is doing enough to
combat cybercrime, with the proportion of respondents who totally agree being gen-
erally low across member states. Moreover, a significant proportion (14%) does not
know if enough is being done to combat cybercrime.
In this last study of 2017, we can see that the percentage of respondents wor-
ried about any type of threat is always higher than those that do not have concerns
(see Table 2). Respondents have become more concerned about “Identity theft”,
“Malicious software” and “Online banking fraud” that present percentages close to
70%. On the other hand, “Material promoting racial hatred or religious extremism”
presents only a little more than 50% concern by the average European respondents.
Comparing the levels of concern of the Portuguese, we can see that they are
in line with the average of the Europeans (Fig. 1). However, there are threats that
concern more the Portuguese, namely “Offensive material and child pornography”
and “Material promoting racial hatred or religious extremism”.
2.3 Experience of Cybercrimes
Europeans’ experience about cyber security has been growing over the last few years.
In this last study of 2017, we can see that the percentage of respondents with concrete
experiences with these types of threats associated with cyber security has increased
(see Table 3). However, there is a significant variation among member states on the
proportion in this type of experience. For instance, in general the Portuguese present
a percentage of cases in which they were victims, inferior to the majority of the
member states. The average European respondents report more cases in which they
were victims associated with “Malicious software” and “Scam emails or phone calls”
that present percentages close to 50%. On the other hand, “Identity theft”, “Offensive
material and child pornography” and “Cyber extortion” present percentages below
10%.
Table 2 Concerns of Portuguese and Europeans average in cybercrimes

Very Fairly Not very Not at all Don’t know
concerned concerned concerned concerned (%)
(%) (%) (%) (%)
Identity EU 33 36 22 8 1
theft PT 35 34 15 16 0
Scam EU 26 34 26 13 1
emails or PT 27 33 18 22 0
phone calls
Online EU 22 36 28 12 2
fraud PT 21 38 22 17 2
Offensive EU 27 26 26 19 2
material and PT 30 31 16 23 0
child
pornogra-
phy
Material EU 21 30 29 18 2
promoting PT 24 31 19 25 1
racial hatred
or religious
extremism
Access to EU 22 35 28 13 2
online PT 25 30 18 25 2
services
Email EU 27 36 24 11 2
account PT 30 36 18 15 1
hacking
Online EU 32 34 21 11 2
banking PT 31 31 18 18 2
fraud
Cyber EU 25 30 28 15 2
extortion PT 28 30 16 24 2
Malicious EU 29 40 22 8 1
software PT 29 42 22 7 0
Comparing the experience of the Portuguese with the average of the European
respondents (Fig. 2), we can see that in some cases the percentage of cases in which
respondents are victims are very similar, namely “Identity theft”, “Offensive mate-
rial and child pornography”, “Access to online services” and “Cyber extortion”.
However, there are cyber security threats whose percentages of cases are more diver-
gent between the Portuguese and the European average. Indeed, “Material promoting
racial hatred or religious extremism” presents a 10% difference and especially “Scam
emails or phone calls” which present a 22% difference.
Fig. 1 Comparison between Portuguese and Europeans average concerns in cybercrimes indicators
Table 3 Experience of Portuguese and Europeans average in cybercrimes

Has been a victim (%) Never (%) Don’t
know (%)
Identity theft EU 8 91 1
PT 4 96 0
Scam emails or phone calls EU 38 61 1
PT 11 89 0
Online fraud EU 16 83 1
PT 7 92 1
Offensive material and child EU 7 92 1
pornography PT 7 93 0
Material promoting racial EU 18 81 1
hatred or religious extremism PT 8 91 1
Access to online services EU 11 87 2
PT 7 92 1
Email account hacking EU 14 85 1
PT 6 93 1
Online banking fraud EU 11 8 1
PT 3 96 1
Cyber extortion EU 8 91 1
PT 5 95 0
Malicious software EU 42 57 1
PT 33 66 1
Fig. 2 Comparison between Portuguese and Europeans average concerns in cybercrimes indicators
It should be noted that only a minority of respondents have experienced any of

these crimes. Most of the attacks described here were experienced by less than 15%
of respondents, with the exception of “Scam emails or phone calls” and “Malicious
software” that presented more representative values.
2.4 Institutions Perceived as Responsible to Provide

Assistance to Citizens for the Different Types
of Cybercrimes
In face of the concerns and cases experienced by European citizens, it is important to

know which entities they resort to, if they are involved in these threats. In this context,
citizens were questioned for each of the indicators, witch entity that would resort to
if they were victims of one of these threats. The applied question would be “If you
experiences or were a victim of the following situations, who would you contact?”.
Regarding the possibilities of choice, these would be: Police; Website/Vendor; Con-
sumer protection Organisation; Your Internet service provider; Other; No one; and
Don’t know. The vast majority of respondents would go to the police if they were
victims of any kind of cybercrime represented here, as can be seen in Table 4. The
only exception is related to “Malicious software” in which the average European
citizens prefers to use the Internet service provider (29%) and only 25% resort to the
Police.
All threats of cybercrime considered in this study have considerable percent-
ages associated with the police as institutions perceived as responsible to provide
assistance to citizens for the different types of cybercrimes. However, we can also
Table 4 Institutions perceived as responsible to provide assistance to citizens

Police (%) Internet service provider (%)
Identity theft EU 85
PT 88
Scam emails or phone calls EU 53
PT 58
Online fraud EU 52
PT 61
Offensive material and child EU 76
pornography PT 67
Material promoting racial hatred or EU 59
religious extremism PT 50
Access to online services EU 40
PT 47
Email account hacking EU 40
PT 38
Online banking fraud EU 76
PT 84
Cyber extortion EU 70
PT 76
Malicious software EU 25 29
PT 32
conclude that the Portuguese resort to this entity in a percentage higher than the
average of the other Europeans.
The following figure (Fig. 3) presents the spider chart showing the percentage of
the Portuguese and European citizens that have been victims of one of cyber threats
and go to the police. It is observed in Fig. 3 that in most of the types of threats the
Portuguese resort to police in a greater percentage, except “Offensive material and
child pornography” and “Material promoting racial hatred or religious extremism”.
3 Discussion
The increasing use of a growing variety of devices that can access the Internet has
contributed to the overall growth of Internet use. This reality also leading to a general
increase in cybercrime concerns as well as the number of cases associated with this
phenomenon where more and more users are involved. In this context, there is an
increase in citizens’ concerns in most of the different types of threats, with particular
focus on the “Identity theft”, “Malicious software” and “Online banking fraud” that
Fig. 3 Comparison between Portuguese and Europeans average about institutions perceived as
responsible to provide assistance to citizens
present percentages close to 70%. Comparing the experience of the Portuguese with
the average of the European respondents, we can see that the European citizens have
experienced these threats more often than the Portuguese in most of the different
types of cybercrimes. However, only a minority of respondents have experienced
any of these crimes. Anyway, “Malicious software” stands out as the threat that both
the Portuguese and the Europeans more mentioned. In the case where European
citizens have been victims of one of these threats, the majority of them have resort to
the police. The Portuguese have the same approach, although with slight differences
in the various types of threats.
Based on the data analysed by this study, there is a growing need to combat this
phenomenon. In fact, knowing that there is a growing concern about cybercrimes and
an increase in cases where people are victims justifies intervention through public
policies to take action to combat these threats. In this sense, new and more restric-
tive legislation has been implemented, which in the medium term may contribute
to the improvement of these results. In 2013, the European Commission presented
a proposal for a Directive concerning measures to ensure a high common level of
network and information security across the Union. The Directive (EU) 2016/1148
of the European Parliament and of the Council was approved on the 6 July 2016
and is currently undergoing the transposition phase to the internal legal systems of
the EU countries [6]. In addition to new legislation, the European Commission has
established in 2013 a European Cybercrime Centre (EC3)3 to help protect European
citizens and businesses against these growing cyber-threats. This centre leads Euro-
pean entity to combat cybercrime and focus on illegal online activities by organized
3 https://www.europol.europa.eu/about-europol/european-cybercrime-centre-ec3.
crime groups, particularly those that generate large criminal profits, such as online
fraud involving credit cards and bank credentials [3]. However, the most important
institutions in the European Union in the field of network and information security
are European Agency for Network and Information Security (ENISA), established
in 2004, and European Computer Reaction Initiative (CERT-EU), formed in 2012.
Also, Portugal creates a Centre (CNCS),4 that is the operational coordinator and the
Portuguese national authority specialized in cyber security working in this field with
state entities, operators of critical infrastructures, operators of essential services and
digital service providers [6]. Another way of improving the results of this study is
related to the level of knowledge and information of the citizens themselves. Some of
the threats most experienced by citizens may be reduced by an additional caution that
they must have. Actually, users can do more to protect themselves, because the two
most frequently experienced forms of cybercrime, infection with malicious software
and fraudulent obtaining of personal information, are actions that a well-informed
public can do much to prevent themselves [1].
4 Conclusion
Despite the permanent development of information technologies and new forms of

potential abuse, Internet users expect that every information system be able to reject
attacks that could jeopardize system data [7]. Additionally, these users, on their own
initiative, publish a large number of their personal information in cyberspace, without
thinking that this information could be used against them. Such behaviour inevitably
leads to a marked lack of trust in systems and an ongoing concern with security
issues. These concerns are well present in the results of the survey analysed here.
The Portuguese and the European citizens in general have the same degree of
concern about cyber security and the threats associated with cybercrime. It is noted
that these concerns have not diminished over time, and therefore, actions are justified
to reverse this situation. Actions should be taken in a coordinated manner to combat
large-scale organized cybercrime and impose data protection standards on providers
of online services. Besides that, the findings of this survey also highlight the impor-
tance of greater public education on types of cybercrime, their consequences, and
ways in which their impact can be avoided or mitigated.
In the near future, it will be important to carry out a new survey with the same
format, which allows to know the evolution of these cybercrime indicators over time,
as well as, to check if the combat measures implemented are taking effect or require
a reinforcement or adjustment.
4 CNCS—Centro Nacional de Cibersegurança (National Cybersecurity Centre).

References
1. European Commission: Europeans attitudes towards cyber security. Special Eurobarometer 464a
(2017)
2. European Commission: Cyber security. Special Eurobarometer 423 (2015)
4. Barros, G.: A Cibersegurança em Portugal. Gabinete de Estratégia e Estudos - Estudos Económi-
cos, p. 56 (2018)
6. CNCS: Centro Nacional de Cibersegurança, 13 Dec 2018. Available from https://www.cncs.
gov.pt/en/about-us/ (2018)
7. Vilic, V.: International and Serbian legal framework of the right to privacy in cyberspace. MEST
J. 6(1), 119–131 (2018)
Classification of Phishing Attack
Solutions by Employing Deep Learning
Techniques: A Systematic Literature
Review
Eduardo Benavides, Walter Fuertes, Sandra Sanchez and Manuel Sanchez
Abstract Phishing is the technique by which the attacker tries to obtain confidential
information from the user, with the purpose of using it fraudulently. These days, three
ways to mitigate such attacks stand out: Focus based on awareness, based on black-
lists, and based on machine learning (ML). However, in the last days, Deep Learning
(DL) has emerged as one of the most efficient techniques of machine learning. Thus,
this systematic literature review has been aimed to offer to other researchers, read-
ers and users, an analysis of a variety of proposals of other researchers how to face
these attacks, applying Deep Learning algorithms. Some of the contributions of the
current study include a synthesis of each selected work and the classification of anti-
phishing solutions through its approach, obtaining that the uniform resource locator
(URL)-oriented approach is the most used. Furthermore, we have been able to clas-
sify the Deep Learning algorithms selected in each solution, which yielded that the
most commonly used are the deep neural network (DNN) and convolutional neural
network (CNN), among other fundamental data.
Keywords Phishing · Deep Learning · Social Engineering · Machine Learning ·

Cybersecurity
1 Introduction
Phishing is the most commonly used attack on social engineering. Through such
attacks, the phisher tries to obtain confidential information from the user, with the
purpose of using it fraudulently against himself or its organization [1]. The most
E. Benavides (B) · W. Fuertes

Escuela Politécnica Nacional, Quito, Ecuador
e-mail: diego.benavides@epn.edu.ec
E. Benavides · W. Fuertes · S. Sanchez
Universidad de Las Fuerzas Armadas, Sangolqui, Ecuador
M. Sanchez
Universidad de Alcalá de Henares, Madrid, Spain
https://doi.org/10.1007/978-981-13-9155-2_5
52 E. Benavides et al.
common way to become a victim to these attackers is by entering a Web site with a
fraudulent URL [2]. Currently, three ways to mitigate social engineering attacks stand
out: The first way is through user awareness and training to deal with these attacks
[1]. The second and most commonly used way is to have a blacklist of phishing sites
[3]. However, the disadvantage of this approach is that it must first know the sites that
have already been detected as phishing. The third way to prevent these attacks, being
more effective than the previous one, is to detect them in their early appearance,
using machine learning algorithms, which are able to predict if a page is phishing,
based on the previously analyzed characteristics of many other pages that already
have been adulterated. The problem with the ML approach is that, its need to involve
the knowledge of an expert, and a considerable period of time, in order to select the
characteristics that dictate that a page is phishing or not. A particular technique of
machine learning is Deep Learning [4]. With this approach, based on Deep Learning
algorithms, it is not necessary that the user or an expert indicates the characteristics
of a contaminated page, because the algorithm is able to detect them automatically
and independently [5].
Currently, ML anti-phishing approaches can be classified into two groups: Phish-
ing email-based approaches and those based on phishing Web sites. Among the solu-
tions approaches to phishing Web sites, the one based on URLs stands out. Thus, our
work is oriented to give a complete vision to the techniques of mitigation of phishing
URLs by means of Deep Learning algorithms.
Based on the aforementioned, our study aims to analyze the state of the art related
to the detection of social engineering attacks associated with phishing URLs, using
Deep Learning techniques. In order to fulfill such attempt, a systematic literature
review (SLR) has been conducted [6], with the purpose of analyzing the studies
related to the subject and evaluating the proposals of other researchers, in order to
determine their impact and find potential solutions.
Some of the contributions of the current study include a SLR, whose preliminary
result provides a general approach on the application of Deep Learning techniques
in the process of detection and mitigation of social engineering attacks, in order to
reduce the given risk.
The rest of this article has been organized as follows: Sect. 2 sets the background
of this research. Next, in Sect. 3, the phases of the SLR are explained, and a synthesis
of the most relevant articles is presented. In Sect. 4, the obtained results are discussed.
Finally, the conclusions and future work are described in the last section.
2 Background
2.1 Phishing
Phishing is the combination of social engineering and technical exploitations,

designed to convince a victim to provide personal information such as account
Classification of Phishing Attack Solutions by Employing … 53
numbers, passwords and credit card details, and with the usual purpose created in
order to obtain a monetary gain by the attacker. Most phishing attacks are realized
by fake emails, containing a uniform resource locator (URL). In case of its activation
by a click, such URL leads to a fake malicious Web site. Despite the important atten-
tion, it has received over the years, there is still no definitive solution to solve this
problem [2]. In [3], authors define “a Phishing page, like any web page, that without
permission, claims to act on behalf of a third party; with the intention of confusing
the spectators in the performance of an action.”
Anti-phishing studies may be classified; these may be classified, according to their
objective, in two large groups being those aimed at mitigating phishing emails, and
those aimed at mitigating attacks on Web sites. Those oriented toward Web sites may
be classified mainly into four different classes: (1) focused on the URL of the Web
site; (2) focused on the HTML content; (3) focused on behavior; and (4) focused on
a hybrid approach.
2.2 Deep Learning
According to [4], Deep Learning is defined as a machine learning technique, where

many layers of information processing stations are exploited, by classification pat-
terns and characteristics, or by learning by representation. In fact, Deep Learning
is implemented by neural networks, with many layers that are not something new.
However, it has become popular recently, due to three factors: First, there is a notable
increase in processing capabilities (e.g., video cards, graphical processors, etc.); sec-
ond, by affordable computer hardware; and third, due to recent advances and devel-
opments in Deep Learning research. Deep learning algorithms may be classified into
three subgroups, which depend on the fact if the algorithms are trained to yield the
obtainable results or not. The three subgroups are classified as unsupervised, super-
vised and hybrid. In [5], the consistency of each of the Deep Learning algorithms
has been described in detail (Fig. 1).
3 Methodology
Our study followed the SLR methodology proposed by Barbara Kitchenham [6].
Five consecutive steps have been conducted: (1) define the research questions; (2)
search the relevant documents; (3) select the primary studies; (4) analyze the abstracts
and extract keywords and data; and (5) map the selected primary studies. First, the
search of the related articles has been accomplished and as a consequence, the relevant
results have been listed. Then, the data extraction has been achieved and the reports
of the encountered information have been finally produced. A brief description of
each phase is described below:
Deep Learning
Unsupervised Hybrid Supervised
Sum Product Recurrent Neural Deep Neural ConvoluƟonal

Autoencoder Boltzman Machine
Network Network Network Neural Network
(AE) (BM)
(SPN) (RNN) (DNN) (CNN)
Stacked AE Deep BM Restricted BM

(SAE) (DBM) (RBM)
Deep Belief
Denoising AE
Network
(DAE)
(DBN)
Fig. 1 Classification of deep learning techniques [5]
3.1 Define the Research Questions
What are the techniques of Deep Learning that are currently used in primary studies
and how do researchers use these Deep Learning techniques, in order to mitigate
phishing attacks?
3.2 Search the Relevant Documents
Our strategy consisted of first defining the search chain, which consisted of two parts,
being URL phishing and Deep Learning. Next, the search chain has been modified
by synonyms of each of the two parts. With this chain, we proceeded to search in all
articles related to the given subject.
3.2.1 Inclusion Criteria
Once the search keywords have been defined, only the following documents have
been included being those: (1) whose main topic is phishing and Deep Learning; (2)
which have been written exclusively in English; (3) its publication has been only of
at most the last five years; (4) its solution method is about URL-oriented.
3.2.2 Exclusion Criteria
(1) articles whose subject has been only phishing or only Deep Learning, without
combing both topics; (2) being in a language other than English; (3) articles published
prior to 2010; (4) lacking of methods applied to URL-oriented.
3.2.3 Quality Criteria
In order to comply with the quality criteria of the search, it has been decided to
conduct the scrutiny only on pages recognized at the scientific level, which are also
from the area of information technology. The indexed databases chosen have been:
IEEE Explore, Taylor and Francis, Springer Linker, ACM Portal and Science Direct.
3.3 Selecting the Primary Studies
After searching for the needed information, 59 candidate articles have been encoun-
tered, and a group of 19 primary studies has been established. Based on the synthesis
performed in the previous step, we proceeded to conduct the report, and to demon-
strate the information gathered by the analysis.
4 Primary Studies
Id Used Used Applied Phishing Applied Main features

technique algorithm methodology objective technique
[7] a TML support Comparative Malign N/A Authors evaluate some
vector Analysis Email algorithms: Support vector
machines Features machines (SVM), neural
(SVM), networks, self-organizing
neural maps and K-means.
networks, Among the studied
self- methods, we observed that
organizing the most accurate has been
maps and SVM with an accuracy of
K-means 97.99%
(continued)
(continued)
[8] b HMDL Deep Comparative Malign Multi-stage Normal or abnormal
Learning Analysis and Phishing Malware- behavior of the programs
Algorithms A new proposal Software detection executed in the operating
and ML → DL system has been
Machine monitored. If for some
learning reason and based on the
algorithms rules of machine learning,
this software has been
suspicious, then it passed
through an evaluation
phase using Deep Learning
[9] c DL CNN A new proposal Homoglyph CNN is This solution uses a
at URL. i.e., trained with technique applied in the
svchost.exe the rendering of images. In
replaced Euclidean this way, a CNN learns
with distance features that are used to
svch0st.exe between detect visual similarities of
their the rendered strings.
respective Methodologically, first the
features domain name text strings
are transformed into
images and then these are
passed through the SCNN
[10] DL CNN A new proposal URL eXpose With this technique, we
obfuscation neural emphasize that the data
network entered are short strings of
raw data (data without
format or without
classification). Then, it
learns to extract
characteristics and classify
them, using incrustations at
the character level with a
convolutional neural
network
[11] DL CNN A new proposal URL Bag of The researchers had
obfuscation Bytes previously assigned a
hexadecimal value to each
character in the URL
string. Then, those values
have been matched with
each value on the right.
Later, these datasets in
hexadecimal
configurations are passed
through the algorithm of
the neural network model
for a subsequent
classification. eXpose is
improved
(continued)
(continued)
[12] ML a TML Comparative URL N/A First, traditional methods
DL algorithms Analysis obfuscation of machine learning are
logistic evaluated, and then it
regression compares them with the
with methods of Deep Learning.
bigrams In conclusion, it is deduced
CNN that the Deep Learning
CNN- LSTM method is the most
LSTM accurate, with an accuracy
of 98%
[13] N/A N/A User Awareness N/A N/A The researchers propose to
mitigate the attacks by
teaching how to prevent
these. This proposal is
rather practical, applying
the concept of big data
analytics, in order to
mitigate the attacks of
social engineering
[14] ML AdaBoost, A new proposal URL (1) A syn- The premise is used that a
bagging. obfuscation tactic set of characters on a
random analysis phishing page is similar to
forest, of the a legitimate page. It also
sequential text; and performs an evaluation of
minimal (2) A statis- the statistics of the
optimization tical characteristics of the
(SMO) analysis phishing pages. That
of the occurs if it is a new page
Web or, for example, if it has a
site. bad reputation, etc.
Semantic
features
[15] ML Support Comparative URLs in This system While these algorithms
DL vector Analysis and a new general recom- used a public dataset of
machines proposal mended the around 2.4 million URLs,
(SVM), use of this system recommended
random numerical the use of numerical
forest, characteris- characteristics
Naive tics for
Bayes, training, in
decision order to
trees, obtain a
K-nearest better rate in
neighbors prediction
and accuracy.
multilayer
perceptron
(continued)
(continued)
[16] DL Long A new proposal URL Feature Phishing pages are
Short-Term obfuscation extraction collected, and then the
Memory LSTM recurrent neural
Recurrent network algorithm is
Neural trained with 12
Network characteristics. This finally
LSTM- produces a 99.14%
RNN effectiveness in the
detection of new false sites
[17] DL Deep A new proposal Unbalanced Borderline- First, three groups of
Boltzmann data SMOTE characteristics of the pages
Machine (Synthetic of phishing are extracted
(DBM) Minority being these the
Oversam- characteristics of URL, of
pling Web pages and features of
Technique) the images. With this data,
algorithm the algorithm is trained and
afterward an effectiveness
over 99% in the detection
of new affected sites is
obtained
[18] DL Deep belief Framework to detect Internet It focuses mainly on the
network phishing pages Protocol application of a framework
(DBN) Flow to detect phishing pages.
Firstly, two types of
characteristics of the false
sites are classified, being
the original ones and the
interaction ones. Hereby, a
model based on Deep
Belief Network is proposed
[19] DL Neural Comparative URL Before This is a comparative study
networks, Analysis obfuscation being among traditional ML
support analyzed, approaches and DL
vector the URLs approaches at phishing
machine, are detection. As a result, the
decision tree translated DL technique achieved the
DNN. into the best detection rate with an
Stacked ASCII code. accuracy of 80%,
Autoen- In addition, regardless of the amount of
coder it creates a data entered for training
technique determina-
(SAE) tion of the
incidence
percentage
of each
characteris-
tic in a
phishing
page
(continued)
(continued)
[20] DL RNN A new proposal URLs in Gated An advantage of such
General recurrent study is that it has been
neural able to also identify which
networks type of attack on URLs is
(GRUs) being attempted. Hereby it
classifies that URL in
Legitim, SQL Injection,
XSS Attack, Sensitive File
Attack, Directory
Traversal, or another type
of attack
[21] DL CNN – Comparative Malicious – Feature This approach uses a CNN
study and URLs Selection scheme based on character
– A new proposal Malicious Scheme level. After using this
DNSs and scheme, a comparison is
– CNN at performed with other
the word schemes, obtaining by
and using a feature selection
character scheme, finding some 282
level failures. Later, using CNN
at the word level, 158
failures have been
encountered, while, using
CNN at the character level,
some 40 failures were
detected
[22] DL CNN A new proposal DNS flow D-FENDS D-FENS identifies
LSTM (DNS malicious domain names in
Filtering real time. This system runs
& inside a DNS server.
Extraction D-FENS opt for a Deep
Network Learning approach that
System) learns the characteristics
automatically from the
input data. An unusual
method is where detection
is applied directly to traffic
in a DNS
[23] ML Random A new proposal Domain WordGraph The authors propose to use
DL forest CNN generation a tool called WordGraph,
algorithms which generates
(DGAs) and dictionaries similar to
fraudulent those used by DGAs
URLs
(continued)
(continued)
[24] ML Random Comparative Study • URL To identify This study has been
forest J48, obfusca- the phishing included, because it
logistic tion sites based explains in detail the
regression, features • on the different characteristics
Bayes Third- Features that may be extracted from
network, party- extracted a deceptive site
multilayer based from URL,
perceptron, features • Web site
sequential Hyperlink- content and
minimal based third-party
optimiza- features services
tion,
AdaBoostM1,
SVM
[25] DL DNN Behavior People People DeepSeq First, you get the
Study characteristic profile of
people who are commonly
pro-thought to be phishers.
For this, based on the
obtained logs, the intrinsic
data of the people is
compared (age, sex,
occupation, etc.), versus
the data of the visited sites
(business, art, social
media, etc.). Finally, after
performing an analysis by
means of DNN, a profile is
obtained
[26] DL DNN Address Bat The researchers propose
bar-based, Algorithm the TDLBA method
abnormal- feed- (Tuning Deep Learning
based forward using Bat Algorithm). This
features, neural method combines swarm
HTML- and networks intelligence approaches by
JavaScript- configuring the parameters
based, of Deep Learning networks
domain-
based
features
a TML = traditional machine learning. b HMDL = hybrid of machine and Deep Learning. c DL = Deep Learning
5 Discussion of the Selected Primary Studies
After performing the analysis on the initially found 59 studies, only 19 relevant works
related to phishing attacks have been selected through the use of Deep Learning
algorithms. Table 1 documents that most of them have been published in Springer
(See Fig. 2). We also encountered that most research about deep learning applied to
combat phishing attacks has been conducted in China (See Fig. 3).
Table 1 Articles versus Article URL Content Behavior

approaches
[5] No No Yes
[8] No No Yes
[9] Yes No No
[10] Yes No No
[11] Yes No No
[12] Yes No No
[14] Yes Yes No
[15] Yes No No
[16] No No No
[17] Yes Yes No
[18] Yes Yes No
[19] Yes No No
[20] Yes No No
[21] Yes No No
[23] Yes No No
[24] Yes Yes Yes
[26] Yes Yes Yes
Fig. 2 Articles published by 8

indexed database 7
6
5
4
3
2
1
0
Fig. 3 Articles published by 6

the country
5
4
3
2
1
0
Table 2 Articles per deep Category Deep learning algorithm Articles found
learning algorithm
Unsupervised SAE—stacked AE
SDAE—stacked DAE
SPN—sum product
network
RNN—recurrent neural [12, 16, 20]
network
DBM—deep BM [5, 17]
DBN—deep belief network [18]
Hybrid DNN—deep neural [19, 23, 25, 26]
network
Supervised CNN—convolutional [9, 10, 11, 21]
neural network
In addition, we observed that the main anti-phishing approaches applying Deep

Learning, has been able to be classified according to the place of the Web page, in
which the analysis of the data has been performed. For instance, in: URL, Content
(HTML Content), Behavior (Behavior of the website) (see Table 1). In Table 2, we
observed and classified the proposed solutions according to the use of Deep Learning
algorithms that are Not Supervised, supervised and hybrid.
6 Conclusions and Future Work
The current study is a systematic literature review on current Deep Learning solu-
tions in order to combat phishing attacks, even latest threats, such as spear phishing
attacks. To achieve such attempt, the different Deep Learning solutions have been
characterized, classified and analyzed. As a result, it has been evident that there is
still a great gap in the application of Deep Learning algorithms in the detection of
cyber threats in general that has not been exploited yet.
As future work, a comparative study is expected to determine which Deep Learn-
ing algorithm is the most effective in determining malicious sites. In addition, it has
been planned to design models or algorithms to combat phishing attacks, analyzing
the similar characteristics in the pages generated by recurring bots. A new algorithm
will also be proposed to accelerate performance in the execution of Deep Learning
algorithms, in phishing detection.
References
1. Hajgude, J., Ragha, L.: Phish mail guard: Phishing mail detection technique by using textual
and URL analysis. In: 2012 World Congress on Information and Communication Technologies,
pp. 297–302 (2012)
2. Marchal, S., Armano, G., Grondahl, T., Saari, K., Singh, N., Asokan, N.: Off-the-Hook: An Effi-
cient and Usable Client-Side Phishing Prevention Application. IEEE Trans. Comput. 66(10),
1717–1733 (2017)
3. Whittaker, C., Ryner, B., Nazif, M.: Large-Scale Automatic Classification of Phishing Pages
4. Deng, L.: A Tutorial Survey of Architectures, Algorithms, and Applications for Deep Learning.
APSIPA Trans. Signal Inf. Process. (2014)
5. Selvaganapathy, S., Nivaashini, M., Natarajan, H.: Deep belief network based detection and
categorization of malicious URLs. Inf. Secur. J. A Glob. Perspect. 27(3), 145–161 (2018)
6. Kitchenham, B., Pearl Brereton, O., Budgen, D., Turner, M., Bailey, J., Linkman, S.: Systematic
literature reviews in software engineering – A systematic literature review. Inf. Softw. Technol.
51(1), 7–15 (2009)
7. Basnet, R., Mukkamala, S., Sung, A.H.: Detection of Phishing Attacks: A Machine Learn-
ing Approach. In Soft Computing Applications in Industry, pp. 373–383. Berlin, Heidelberg,
Springer Berlin Heidelberg (2008)
8. Yuan, X.: PhD Forum: Deep Learning-Based Real-Time Malware Detection with Multi-Stage
Analysis. In 2017 IEEE International Conference on Smart Computing (SMARTCOMP),
pp. 1–2 (2017)
9. Woodbridge, J., Anderson, H.S., Ahuja, A., Endgame, D.G.: Detecting Homoglyph Attacks
with a Siamese Neural Network
10. Saxe, J., Berlin, K.: eXpose: A Character-Level Convolutional Neural Network with Embed-
dings For Detecting Malicious URLs, File Paths and Registry Keys (2017)
11. Shima, K., et al.: Classification of URL bitstreams using Bag of Bytes (2018)
12. Vazhayil, A., Vinayakumar, R., Soman, K.: Comparative Study of the Detection of Malicious
URLs Using Shallow and Deep Networks. In 2018 9th International Conference on Computing,
Communication and Networking Technologies (ICCCNT), pp. 1–6 (2018)
13. Epishkina, A., Zapechnikov, S.: A syllabus on data mining and machine learning with appli-
cations to cybersecurity. In 2016 Third International Conference on Digital Information Pro-
cessing, Data Mining, and Wireless Communications (DIPDMWC), pp. 194–199 (2016)
14. Zhang, X., Zeng, Y., Jin, X.-B., Yan, Z.-W., Geng, G.-G.: Boosting the phishing detection
performance by semantic analysis. In 2017 IEEE International Conference on Big Data (Big
Data), pp. 1063–1070 (2017)
15. Vanhoenshoven, F., Napoles, G., Falcon, R., Vanhoof, K., Koppen, M.: Detecting malicious
URLs using machine learning techniques. In 2016 IEEE Symposium Series on Computational
Intelligence (SSCI), pp. 1–8 (2016)
16. Chen, W., Zhang, W., Su, Y.: Phishing Detection Research Based on LSTM Recurrent Neural
Network, pp. 638–645. Springer, Singapore (2018)
17. Zhang, J., Li, X.: Phishing Detection Method Based on Borderline-Smote Deep Belief Network,
pp. 45–53. Springer, Cham (2017)
18. Yi, P., Guan, Y., Zou, F., Yao, Y., Wang, W., Zhu, T.: Web Phishing Detection Using a Deep
Learning Framework. Wirel. Commun. Mob. Comput. 2018, 1–9 (2018)
19. Aksu, D., Turgut, Z., Üstebay, S., Aydin, M.A.: Phishing Analysis of Websites Using Classifi-
cation Techniques, pp. 251–258. Springer, Singapore (2019)
20. Zhao, J., Wang, N., Ma, Q., Cheng, Z.: Classifying Malicious URLs Using Gated Recurrent
Neural Networks, pp. 385–394. Springer, Cham (2019)
21. Jiang, J., et al.: A Deep Learning Based Online Malicious URL and DNS Detection Scheme,
pp. 438–448. Springer, Cham (2018)
22. Spaulding, J., Mohaisen, A.: Defending Internet of Things Against Malicious Domain Names
using D-FENS. In 2018 IEEE/ACM Symposium on Edge Computing (SEC), pp. 387–392
(2018)
23. Pereira, M., Coleman, S., Yu, B., DeCock, M., Nascimento, A.: Dictionary Extraction and
Detection of Algorithmically Generated Domain Names in Passive DNS Traffic, pp. 295–314.
Springer, Cham (2018)
24. Rao, R.S., Pais, A.R.: Detection of phishing websites using an efficient feature-based machine
learning framework. Neural Comput. Appl., 1–23 (2018)
25. Sur, C.: DeepSeq: learning browsing log data based personalized security vulnerabilities and
counter intelligent measures. J. Ambient Intell. Humaniz. Comput., 1–30 (2018)
26. Vrbančič, G., Fister, I., Podgorelec, V.: Swarm Intelligence Approaches for Parameter Setting
of Deep Learning Neural Network. In Proceedings of the 8th International Conference on Web
Intelligence, Mining and Semantics—WIMS ’18, pp. 1–8 (2018)
Is Cyber Warfare an Alternative?
Jorge Barbosa
Abstract The possibility of using computerized means in offensive action against

civilian or military installations from enemy countries is a real. Given the potential
of this type of action, they can be seen by many countries as interesting, and as an
alternative to conventional kinetic warfare. Despite its many theoretical advantages,
obtaining the cyber capabilities needed to implement cyber military units is not easy
and may not be within reach of any country. The necessary financial means may not
be a major obstacle to this achievement because they are relatively low, when com-
pared to those necessary to obtain conventional means. Similarly, obtaining a support
infrastructure is also not theoretically difficult. However, obtaining both human and
technological resources, other than those related to the support infrastructure, may
not be available to any country, namely the designated zero-day exploits, ZDE, that
are needed to explore computer vulnerabilities in informatics systems of the other
countries. There are other types of constraints, time needed to prepare the means
and the planning and development of cyber actions, type of targets to be attained,
specificity of computer systems and even the possible need for cyber media comple-
mentarity with conventional means that can be limiting and impede their widespread
use. Thus, cyber warfare is a very interesting alternative, but given these constraints
and limitations, it may not be a real viable alternative to be used in all situations and
by all countries.
Keywords Cyber warfare · Cyber weapons · Cyber capability · Zero-day exploits
1 Introduction
The use of cyber weapons, which enable cyber warfare to be triggered, is due to
the weaknesses created by the existence of hardware and/or software vulnerabili-
ties, which, in general, all computer systems have. When these vulnerabilities are
J. Barbosa (B)
Coimbra Polytechnic – ISEC, Coimbra, Portugal
e-mail: jorge.barbosa@isec.pt
Portuguese National Defense Institute, Lisbon, Portugal
https://doi.org/10.1007/978-981-13-9155-2_6
66 J. Barbosa
not known, in the sense that they are not publicly known but only known within
a strict group, they are called zero-day vulnerabilities, ZDV. These vulnerabilities,
because they are not publicly known, can be used to great advantage in the construc-
tion of applications called exploits, which if based on ZDV will be called zero-day
exploit, ZDE, which can be used with great success as cyber weapons. These ZDEs
by exploiting the vulnerabilities of computer systems can take control or even par-
alyze the systems affected by those vulnerabilities. If such systems are linked to
the control of so-called critical infrastructures, i.e., systems such as power dams,
banking systems and others, their allocation and even stoppage can have very great
detrimental effects and hence their usefulness as cyber weapons. The interest of this
type of cyber weapon is not, or not only, to be used in the vulgar cybercrime but
essentially because they can be used in actions of cyber war by sovereign states,
against other states.
As this type of action may jeopardize state sovereignty, these actions are referred
to as cyber warfare rather than cybersecurity actions, which will then only be related
to vulgar cybercrime actions.
The use of computerized means to trigger offensive actions against antagonistic
countries made possible by the constant evolution of computing means, both from
the point of view of the means itself and from the increase in the capacity of the
computer networks and their increasing interconnection, is an indisputable reality.
Due to this increased use and diversification of these IT resources, it can be considered
that in modern societies, there are no sectors where they are not used. From critical
infrastructure control systems, dams and other power plants, water treatment and
supply systems, to the management and control of banking and financial operations
and institutions, transport logistics, fuel logistics, etc. Today, practically everything is
based on computer controllers, essentially of the industrial controllers like SCADA or
PLC’s in enterprise installations or sophisticate databases and customer relationships
informatics systems in financial institutions or public services.
The reverse side of this beneficial use is that if such systems are unsafe and if, on the
contrary, they are likely to be penetrated, it could result on catastrophic effects as the
effects of the above systems, their destruction or even their alteration, would paralyze
or would seriously affect modern societies. This dependence is therefore criticized
from this point of view, and thus, these systems are tangible and passable targets
of military action against them. These military actions can be triggered using con-
ventional means, but also the computer itself, through the so-called cyber weapons.
These cyber weapons would act directly on the computer systems destroying, par-
alyzing or altering them in order to stop or interfere with their normal functioning,
and thus, the systems downstream that they control or manage.
The theoretical possibilities offered by this military option may lead to the con-
sideration of its use by a large number of countries and not just by the current con-
ventional powers. Theoretically, the amount of resources required, whether human
or material, in this case technological, are much lower than those required for the
initiation of conventional military operations. Likewise, the economic resources nec-
essary to acquire these technological means and to train the human resources adapted
to this type of activity are substantial, but when compared to the economic resources
Is Cyber Warfare an Alternative? 67
required for conventional actions, they are much smaller. The combination of all
these factors, therefore, makes this option very desirable, and this, theoretically, can
lead to the spread and proliferation of cyber powers ready for cyber warfare as an
alternative to kinetic war.
However, its operationalization may not be easy and may even be impossible for
certain states. These difficulties are essentially related to the aforementioned needs
for human and technological resources. However, in quantitative terms, the needs
are smaller. Compared to those required for a conventional war, the qualitative needs
are much greater and much more demanding. These qualitative constraints are the
most restrictive of any country being able to obtain the capabilities needed to launch
cyber actions.
Considering all these apparent advantages of using military actions based on
computer systems together with the difficulties that may exist in the constitution of
cyber military units, will this option be a true alternative for any country?
2 Critical and Essential Conditions
There are factors critical and essential in constitution of cyber units so that the actions
they bring can be successful. This applies to all countries seeking the acquisition of
cyber warfare capabilities.
It has already been mentioned that the financial requirements for the constitution
of these units are relatively lower than those required for the constitution of modern
conventional units. Then, perhaps for any country, those financial aspects are not
impeditive to the formation of cyber armies.
It can be considered that there are some critical and essential factors to obtain
cyber capabilities for use in a cyber war. These factors are: (i) highly qualified human
resources to integrate the cyber units; (ii) sophisticated technological resources; (iii)
the existence of computer vulnerabilities in the systems of opposing countries and
(iv) the cyber units having knowledge of these failures and have the capacity to
exploit them.
It is essential that the computer systems of opposing countries have vulnerabil-
ities in the hardware and/or software, which allow the use of software specifically
developed, informatics exploits, to control such systems using these failures. If those
systems do not have vulnerabilities cannot be attacked using cyber means because
cyber weapons consist in exploits which use the vulnerabilities to affect the systems.
Time is another important factor in this scenario. Not only the time required for
the preparation of the cyber troops, but also the time necessary for the preparation
and development of the cyber actions, which has to be done with great anticipation.
These specific developed informatics exploits have to be secretly placed into
enemy computer systems, which require a lot of preparation and time. On the other
hand, if there is a long time between the secret placement of these exploits and their
possible use, the other country may have realized those vulnerabilities or may have
changed the physical equipment or the software. This way, purposely or not, it can
68 J. Barbosa
frustrate and prevent the use of those exploits, which, because they are specific to
the previous setup, would thus no longer be useful after these changes. On that case,
the whole process would have to be restarted if that would be possible.
Then, in the scenario of a cyber warfare, the time factor is very important in the
preparation and development of cyber war actions.
2.1 Human Resources
Cyber units cannot be formed with any “recruit” as it is the case of some conventional
armies. There are armies in which the structure and preparation are very different
when compared to the conventional armies of the more developed nations. Often
these armies are no more than tribal or regional groupings that seized power and
after became the army of these countries and did not necessarily evolve into a modern
conventional army. In these conditions, it will be very difficult for these armies to
evolve into the creation of modern cyber units, given the great capacity of organization
and the technical knowledge that would be necessary for this.
The requirements for cyber troops have nothing to do with those needed by most
troops of conventional armies. Excluding, perhaps, the senior officers and some
specific military units, the requirements for military capability of conventional units
are primarily their physical and athletic abilities, which should enable a very good
capacity for the execution of military operations, especially on the ground. The
requirements for the elements of military cyber units are very different. Although
the physical and athletic ability is also important, it is the intellectual capacity that
is paramount.
Maybe the most very important factor that may hinder, or at least make it difficult,
the constitution of cyber units, is the need for highly trained members from the
point of view of computer skills. These cyber troops must have the capacity to
develop computer applications, the designated informatics exploits, capable of taking
advantage of possible hardware or software failures, vulnerabilities, in the computer
systems of the opposing countries. To be used in cyber warfare, these exploits have
the ability to interfere with the normal activities of these systems or even paralyze
them. Then, the knowledge and mastery of information technology in its various
subareas, software, hardware, data networks, databases, programming, etc. of the
cyber troops, should be excellent, and it is fundamental and indispensable.
The above preparation requires a lot of time and maybe not is possible to give to
anyone. The computer aptitude of those involved is fundamental, since the computer
skills and abilities are very innate. On the contrary, the preparation of a substantial
part of the troops for a conventional war is relatively quick, and it is even possible
that, at the height of a conventional attack, a country will be able to form a large part
of these conventional troops on an accelerated basis. In other words, it is relatively
more easy to obtain the skills and abilities necessary for a conventional military unit
than for a cyber unit. For the reasons given, replacing, in case of loss, the elements
of a cyber unit may not be easy or fast and may even be impossible. This can make it
impossible to launch cyber war operations, even hinder, or difficult the continuation
of those already started. The loss of a single element, if it is a fundamental element
due to its capacities and knowledge, can be disastrous for cyber unity, given its great
value of knowledge.
The number of these troops is not of the same order of values of those normally
existing in a conventional army, being smaller. However, given the quantity, diversity
and complexity of the tasks to be performed, have to be a reasonable size.
2.2 Technological Resources
In the field of technical needs, and in addition to the needs of highly qualified human
resources, technological means are necessary to enable such cyber troops to operate
and to contribute to and enable the success of their missions.
Fundamentally, there are two types of important technological means: (i) hard-
ware, physical equipment and computer network necessary to set up a support struc-
ture for the missions; (ii) software, exploits, that allow the control or manipulation
of the systems.
The constitution of the support structure, computer hardware and computer net-
work, is relatively easy to obtain, and it is almost just a financial question whether any
country obtains them or not. These financial needs will not be very large, especially
if compared to those required for the purchase of a modern conventional military
medium, military aircraft, submarine or even a battle tank.
Exploits, in particular ZDE, are much more difficult to obtain because they are
specific to the systems and vulnerabilities in them. In the so-called Deep Web, there
is, however, a vast offer of exploits. However, these exploits that are in the Deep
Web have, in general, little interest for the cyber warfare exactly because already
being publicly known, that is, to say not being ZDE. As such, at least theoretically,
computer systems, especially those linked to critical infrastructures, may already be
protected from attacks based on this failure. They are not, therefore, properly exploits
of “cyber military quality”. Rather, they are only of interest to cybercriminals who
want to launch attacks aimed at individuals and small businesses. We do not want to
say that such attacks are unimportant because they can inflict great damage and are a
great concern. Without these vital resources, the activity, usefulness and necessity of
cyber military units become quite unimportant, and small nations or technologically
less advanced ones are unlikely to succeed in building these units.
The difficulty in obtaining and using exploits, such as the “quality” cyber military
ZDE, could be too one greatest obstacle to the proliferation of cyber military units.
70 J. Barbosa
3 Interest and Need of Cyber Armies
We can question the interest and/or necessity of any country to constitute cyber
armies. “Why?” and “For what?” any country would want to create a cyber army.
It is too required to analyze if this option is a valid alternative to the conventional
means.
Building cyber capacities is not easy for most country, and the conditions to
maintain this capacity are not simple too. If the initial conditions that made possible
obtain cyber capacities are lost, namely if during the cyber warfare get lost of human
or technologically resources, the action can be turned against the aggressor country.
Then, the decision on the constitution of a cyber army must consider not only the
initial conditions and time to develop cyber-attack capacities but also cyber-defense
capacity and the possibility of the countries to maintain those actions.
Many countries may initially depend on other countries for the constitution of its
cyber capacity. To form the necessary frameworks for the constitution of its cyber
army, namely their human resources, they may be obliged to use foreign educational
and research institutions. To get the technological means, for example, ZDEs, initially
they can also have obtained them in foreign countries.
3.1 The Purpose of the Cyber Warfare
In many less developed countries, the conflicts that arise are generally regional con-
flicts to dominate the exploitation of natural resources or tribal or ethnic conflicts with
the same objective of exploiting natural resources in other parts of the same country
or to obtain the supremacy of one ethnic group over another or others existing in that
country.
In this way, triggering a cyber war instead of a kinetic war does not make much
sense in these types of scenarios and countries.
Despite the advantages that a cyber warfare can have in general, capacity building
takes a long time and requires resources, such as human resources, that will not be
readily available for this type of conflict. Many of these conflicts are often triggered
by reaction to close and recent actions, and as such, it is not possible to establish
planning and obtain the necessary human and material resources for a cyber war-
fare. In the same way, the already mentioned conflicts triggered to obtain profits in
the exploitation of natural resources as the primary objective takes the control of
the respective mining complexes and this presupposes the physical and territorial
domination of these places, which does not fit very well with the concept of cyber
warfare.
Not taking as an absolutely correct premise, we consider that in a large number
of conflicts triggered with some frequency, the applicability of a cyber warfare is of
little interest although isolated cyber actions may occur.
Even regional powers, in general, have the concern that they are regional powers to
protect themselves and not necessarily to militarily dominate their neighbors. Even
if they develop cyber resources capable of endowing them with a cyber warfare,
they would not trigger it in the same logic that even though they may have some
conventional ability, they only use it for defense. The Federative Republic of Brazil
and the Republic of South Africa will be examples of this.
In the case of the great global military powers, the USA, Russia and China, cyber
warfare has been strongly considered and even implemented for some time, being
considered as another operational domain. In the case of the USA, one of its first
units linked to cyber warfare, yours first “Cybercorp”, the “Air Force—1st Cyber
Division”, was formed in the nineties of last century.
However, with regard to their use in military attack, these global powers have
considered this alternative as another military option of their vast arsenals, along
with the other conventional means already existing. The philosophy will then be to
employ or not, according to the military theaters and operative situations. It has been
particularly considered as an initial use option, to be used to weaken the enemy and
to prepare the entry of conventional means, for example, to blind air-control radars
or to shuffle the enemy’s military logistics.
In addition to these possible uses of computer means to military attack, these
global powers have given great importance to yours cyber-defense. Much of the US
initiative in cyber warfare area is in the way to increase and consolidate its cyber-
defense capabilities.
This perspective of cyber-defense rather than cyber-attacks is also more in keeping
with the principles set out in the Charter of the United Nations.
One other type of actors, may be capable of considering the employment of cyber
means, is non-states and some terrorist organizations. However, we believe that in
these cases too, the above-mentioned needs, essentially the preparation time and
the necessary resources, would be great obstacles to its realization. This possibility
is more difficult to employ, so that in these situations, such non-states or terrorist
organizations would probably have opted instead for the use of conventional means
of warfare, not excluding of course isolated cyber-attacks.
3.2 Applications of Cyber Weapons
An important aspect related to the issue of the use of cyber weapons is the fact that
cyber warfare may not be of general use. Then, is not suitable for certain scenarios
where conventional warfare may be more advantageous.
For example, if we consider that the initiation of a war against a country may have
as its objective the territorial domination of that other country or part of it, cyber
warfare alone is not adequate, having to be complemented by the use of conventional
means, namely with ground troops who proceed to this territorial occupation.
Obviously, even in these scenarios, cyber warfare is useful and can be used, but
only as a complement to action, either in the preparation of conventional actions
72 J. Barbosa
preceding kinetic warfare, or during or a posteriori of conventional actions for aid or

consolidation of the victory of conventional means.
The cyber warfare can too be seen not as an alternative way to the kinetic war but
as a subsidiarity mean and vice versa, i.e., the conventional means can complement
a cyber offensive and cyber means can be used to complement the kinetic war. This
way, conventional means may also be employed, by the cyber-attacked country, for
a conventional retaliation to a cyber-attack.
3.3 Commitment of Cyber Means
As already mentioned, for various reasons, maybe the use of cyber means is not to a
general application.
Soon, the countries to “attack” will have cyber structures that can be targets for
the cyber weapons of the attacking country. On the other hand, these cyber structures
must be permeable to the use of cyber weapons, that is, they must have vulnerabilities,
hardware and/or software vulnerabilities, which allow the development of computer
applications, the exploits, that using these vulnerabilities may affect those computer
systems and thereby directly or indirectly affect the infrastructures that the systems
control or manage.
These conditions eliminate the use of cyber media in confrontations with many
countries, if they do not have these cyber targets. Since many of the conventional
wars are also small scale and between relatively undeveloped countries, cyber wars
will not be an alternative to them.
If the target is countries in these conditions of weak dependence on information
technology in their society, even countries that are capable of having a high cyber-
attack capacity will be impeded or very limited in the use of their cyber means against
them.
One of the great difficulties for any country is the ability to gauge the cyber
capabilities of another country.
Unlike conventional benchmarking capacities, which can be minimally percepti-
ble through espionage and the use of technical means of espionage, such as satellites
or others, cybernetic capability is not easily measurable. This difficulty exists not
only because cyber media are difficult to observe, namely software-based ones that
are immaterial and therefore invisible, but even if one can perceive their existence
one may not know or be able to determine the effectiveness of the same.
If the determination of the number of human resources linked to existing cyber
armies its possible, that determination alone may not be enough given that in this
particular case, it is not only the number that is important, but also, mainly, its
technical capacity. A large cyber army might not be a good cyber army if the human
resources were not highly trained and cannot therefore know how to take advantage
of the weaknesses of enemy systems.
Another difficulty and possibility are if the country to attack initially does not
have or have little capacity for cyber-defense capabilities, but after the attack can
establish an alliance with some cyber-enabled country that can retaliate against the
attack in favor of the country’s cyber ally. This possibility may totally change the
initial considerations, which the attacking country may have taken into account. It
should be noted that due to the possibilities of cyber media, the allied country does
not have to be geographically close.
The most referred attack methods used as traffic redirection, website deface-
ment, distributed denial-of-service (DDoS), viruses and trojans and a specific cyber
weapons using ZDE, such as those used in the famous action against Iran known as
Stuxnet (Zetter, 2014).
In the strict sense of cyber warfare, are used specific cyber weapons, the ZDE.
The most known cyber warfare action, known as Stuxnet, used cyber weapons, which
exploit ZDVs in the computer systems of nuclear centrifuges of Iran (Zetter, 2014).
The very big difficulty in identifying the aggressor country is very restrictive and
prevents any retaliatory action, whether using conventional means or using cyber
media. Many cyber actions can be triggered by using computers in third countries
that have nothing to do with cyber warfare. The including DDoS attacks it’s a very
good example of this. Maybe can very hard to identify the authors whereas launched
the DDoS attack. A bad identification and retaliation against other country them can
escalate the war and increase exponentially cyber warfare because the country from
which the attack was apparently launched may not be the primary perpetrator of the
attack not having it launched and having nothing to do with it. It could only have been
used as a launching platform for the attack, using computers located in its territory,
whose legitimate owners were unaware of this use. A cyber-attack against this third
country can also cause the same to retaliate, thus escalating the cyber war. A classic
example of this is the so-called Estonian Warfare in 2007, in which apparently many
of the computers used in these attacks would be located in the USA, an allied country
of Estonia.
In the literature, [1, 2] are reported many cyber-attacks actions against more coun-
tries, such USA, South Korea, Iran, Tatarstan, Kyrgyzstan, Israel and the Palestinian
National Authority, Zimbabwe and Myanmar. More of these actions can be consid-
ered at the border between common cybersecurity actions and cyber warfare actions.
Better separation will be achieved if it is precisely determined those as put in danger
the national sovereignty of the affected states.
Many others cyber warfare actions are allegedly attributed to Russia against Esto-
nia, Georgia and Ukraine. These cyber warfare actions were triggered using not only
the mentioned DDoS method but too SQL Injection, SQLi, and Cross-Site Scripting
(XSS) attacks.
On the other hand, cyber actions not being dependent or limited by physical
distances or frontiers. This way the aggressor countries not need be geographically
close to the country to attack. This characteristic can allow too, a type of military
acts for small-scale actions only with intentions of disruption, may be liable to be
launched against particular countries, notably against the large western countries.
74 J. Barbosa
4 Conclusion
The possibility of using cyber weapons, which allows the launching of a cyber war-
fare, is due to the vulnerabilities existing in computer hardware and/or software,
which, in general, all informatics systems have.
When these failures exist in systems, not only in the system directly linked to the
processing of vital activities of coordination or management of entities more directly
related to civil society, but also in those related to the control of facilities, i.e., critical
infrastructures, the negative consequences of cyber-attacks can be devastating for the
country under attack.
The possibility of the occurrence of cyber war is at present very real and that the
total impossibility of its occurrence does not seem very likely very soon, but may
instead increase, since this impossibility would have to stem from the existence of
faultless computer systems, bugs, which seems to us to be utopian in the near future.
Taking into account only the technological aspects referred to above, we can
assume that any nation can acquire cyber military capacities because the creation of
a computer infrastructure supporting cyber war activities, although not easy, is not
financially significant enough to make it impossible for any country.
However, the other necessary resources, namely the need for highly trained human
resources, both for the development of cyber warfare and for the conduct of cyber
warfare, are already much more difficult to obtain. The formation of these human
resources is time consuming, the existence of people with skills for this area and for
this type of activity is rare, and therefore, it is not possible to transform anyone into
a qualified resource, by much training that can be given.
In addition, the volatility and alteration of the systems require that this training
be constant in order to be able to maintain the minimum level of knowledge that
eventually could have been reached when the cyber unit was formed.
Another significant obstacle, although it is theoretically possible to be overcome,
is the need to know the vulnerabilities of the systems, both hardware and software
systems, if you explore them. These vulnerabilities, which must necessarily exist in
the systems in order to be possible to attack them, are referred to as vulnerability, and if
they are not yet known and publicly disclosed, they are called zero-day vulnerability,
ZDV. These vulnerabilities are not very common nor are they widely available. In
order to be able to exploit these faults in the computers of the opposing countries,
it is necessary to develop or obtain specific software for this, called exploits, and if
they were developed for ZDV, they are called ZDE. In addition, if any country that
claims to acquire cyber military capacities might not be able to obtain and use ZDE,
this claim is impossible to achieve.
Cyber warfare is a very interesting alternative, but given these constraints and
limitations, it may not be a very viable alternative in certain situations and for certain
countries. A cyber warfare, by itself, may not be enough to defeat an “enemy” or
to subjugate it even if it is only economically. In addition, not all conditions and
required resources to establish a cyber army are easy to obtain and maintain.
Acknowledgements I thank the Portuguese National Defense Institute as well as the ISEC/IPC
for the support and facilities granted.
This paper is based on the work Final Report of National Defense Course, done by the author in
the Portuguese National Defense Institute.
References
1. Carr, J.: Inside Cyber Warfare, 2nd edn. O’Reilly Media Inc., Sebastopol, CA, USA (2012)
2. Kostopoulos, G.K.: Cyberspace and Cybersecurity. Boca Raton, FL, CRC Press—Taylor &
Francis Group (2013)
3. Barbosa, J.: Pequenas potências militares convencionais, Grandes potências militares cibernéti-
cas—Abordagem da utilização de meios informáticos na defesa/ataque militar moderno. Final
Report of National Defense Course, Portuguese National Defense Institute, Lisbon, Portugal
(2018)
4. Clark, R.A., Knake, R.K.: Cyber War: The Next Threat to National Security and What to do
About. HarperCollins Publishers Inc, New York (2010)
5. Halpin, E., Trevorrow, P., Webb, D., Wright, S. (eds.): Cyberwar, Netwar and the Revolution in
Military. Palgrave Macmillan, NY, USA (2006)
6. Ranger, S.: What is Cyberwar? Everything you Need to Know About the Frightening Future of
Digital Conflict. https://www.zdnet.com/article/cyberwar-a-guide-to-the-frightening-future-of-
online-conflict/ (2018), Last accessed in November 2018
7. Singer, P.W., Friedman, A.: Cybersecurity and Cyberwar—What Everyone Needs to Know. NY,
Oxford University Press (2014)
Memory Auditing for Detection of
Compromised Switches in
Software-Defined Networks Using
Trusted Execution Environment
Filipe Augusto da Luz Lemos, Rubens Alexandre de Faria,

Paulo Jose Abatti, Mauro Sergio Pereira Fonseca
and Keiko Veronica Ono Fonseca
Abstract Current solutions for detecting compromised switches in software-defined

network (SDN) usually rely on the monitoring of the network traffic or conformance
of the packets traversing through them and the rules defined by the controllers.
Although satisfying, those solutions cannot detect a switch that has been compro-
mised if it is not acting maliciously at the moment of the traffic monitoring as sleeper
agents, which can pose as a national security risk when defense networks are the
targets. An architecture capable of auditing the memory of switches in software-
defined networks is proposed as a solution to detect compromised switches even
when they are not acting maliciously and only leave micro-traces of its activities.
This auditing should be able to verify the conformity between what is in the memory
and the flow rules set by the controllers without overusing the system resources. A
trusted execution environment is also proposed to improve the security of the auditing
processes.
Keywords Software-defined networks · Memory auditing · Compromised

switches · Trusted execution environment · Sleeper agents
F. Augusto da Luz Lemos (B) · R. Alexandre de Faria · P. Jose Abatti ·

M. S. Pereira Fonseca · K. V. Ono Fonseca
Federal University of Technology - Parana, Curitiba, Brazil
e-mail: flemos@alunos.utfpr.edu.br
R. Alexandre de Faria
e-mail: rubens@utfpr.edu.br
P. Jose Abatti
e-mail: abatti@utfpr.edu.br
M. S. Pereira Fonseca
e-mail: maurofonseca@utfpr.edu.br
K. V. Ono Fonseca
e-mail: keiko@utfpr.edu.br

https://doi.org/10.1007/978-981-13-9155-2_7
78 F. Augusto da Luz Lemos et al.
1 Introduction
Software-defined network (SDN) is a network architecture that proposes the separa-

tion of the control plane from the data plane [1]. The separation of the planes is made
by software and introduces the centralizing figure of the controller [2], as exemplified
in Fig. 1. The decoupling amplifies the awareness of the controller over the network
topology under its supervision, providing a better control over the network traffic,
for example, [2]. This greater awareness of the network has been successfully used
to improve network security against attacks, such as in DDoS attacks [3].
The separation of planes has brought improvements to the network security; how-
ever, event detection of possible security incidents mostly relies on the monitoring
of the data flows, the compliance between network flow rules and the network traffic
[4]. This fact raises a problem of detecting a compromised network equipment: the
detection requires it to be behaving maliciously at the moment of the monitoring [5].
Figure 2 exemplifies that case: the switch has an invalid flow entry, but the network
data monitor only verifies the data flowing through it and would probably not detect
the problem. Fitting to this case are the “sleeper agents” which are compromised
network devices that only start to function maliciously when activated by their true
controller [6]. Examples of hidden malware and malicious hardware installed with-
out the user knowledge can be found in cases like tampering of servers that were
sold to major Americans Telecommunication Companies [7], posing an issue to the
national security.
In this paper, we propose a solution to detect compromised SDN switches by
auditing its memory to identify if the contents of the switch memory comply with
the rules set by the network controller or if any unauthorized rule was added or
removed from it, what we define as micro-traces. As a further improvement of our
proposal, a trusted execution environment (TEE), in specific the Intel Software Guard
Extensions (Intel SGX) [8], will be discussed as the execution environment of the
aforementioned auditing process.
Fig. 1 Simplified example of a software-defined network architecture

Memory Auditing for Detection of Compromised Switches … 79
Fig. 2 Example of a sleeper

agent: an invalid entry is not
detected by the network data
monitor
The remainder of this paper is organized as follows: Sect. 2 presents some assump-
tions for the software-defined network architecture described in this paper, Sect. 3
presents our proposal, Sect. 4 discusses the major points of the proposed architecture,
and Sect. 5 presents the conclusions.
2 Assumptions on the Targeted SDN Architecture
The SDN architecture targeted in this paper relies under two assumptions regarding
its switches: the size of the switch-audited memory is relatively small and its memory
content has little variance varying accordingly to the OpenFlow Protocol rules [9].
The concept behind this is that the auditing process needs to have the lowest impact
possible on the system, so it does not deprive the controller and the switches from
the system resources.
2.1 Audit Memory Size
The memory size of the switches refers to the space needed to store the flow table
that is set by the network controller. The table size varies with the number of flow
entries, and each flow table entry content is defined under the OpenFlow Switch
Specification [9] as having seven different fields, as shown in Table 1. That is a linear
correlation between the memory size and the number of flow table entries.
Table 1 Seven different fields in a flow table entry according to the OpenFlow Switch Specification
1.5.1 [9]
Match fields Priority Counters Instructions Timeouts Cookie Flags
2.2 Entries Contents
As mentioned previously, the seven different fields of a flow table entry vary accord-
ingly with the OpenFlow Protocol and are well documented under the Switch Spec-
ification [9], thus limiting the possible outcomes of the memory. Although this may
look like a lot at a first glance, an entry always follows the same rules even on
SDN controllers deployed by different developers, therefore limiting the number of
combinations of values for the entries.
3 Proposed Architecture
As stated in the previous section, the auditing process is concerned about the limi-
tation of the system resources during its execution. Therefore, an online (real-time
execution) verification of each entry in the flow table of a targeted switch would
require processing and network resources that may be not available at the switch.
Figure 3 presents a simplified flowchart of the envisioned auditing process. The figure
shows that the auditing entity collects information from the network controller (NC)
and makes a dump of the switch memory, so it can later calculate the hash values
and compare them. Figure 4 exemplifies how the proposed architecture leverages the
usual monitoring approach against sleeper agents: the auditing process can audit the
memory, identify if there is an invalid entry, and generate a warning to the network
controller, so it can take the necessary measures.
A malicious attacker could intercept the memory dump data and change its con-
tents to comply with valid entries as expected by the NC. An approach to overcome
this attack is to exchange only encrypted data between the NC and the switch. How-
ever, if the encryption process itself were not protected, the auditing cannot be trusted.
By using trusted execution environments (TEEs), additional protection can be pro-
vided aiming at improving data security, as discussed in the next section.
3.1 Trusted Execution Environments
The introduction of the trusted execution environments (TEEs), like the Intel SGX
[8], aims to prevent the auditing to become a new security threat. TEEs are pro-
tected areas of the processor that allow code to be executed inside of it from being
Fig. 3 Example of flowchart of the auditing process
Fig. 4 Under the proposed

architecture, the auditing
process is able to detect a
problem even when the
switch is not acting
maliciously
Fig. 5 Representation of the

execution of a code inside a
secure enclave
disclosed or modified [8]. As shown in Fig. 5, an untrusted code calls a function to

be executed inside the protected enclave (ecall). Another important point shown in
Fig. 5 is the ability of a trusted code to call an outside function (ocall), for example,
the printf function. Although the Intel SGX has presented some vulnerability issues
[10], Intel’s TEE is still relevant when applied to secure processes like encryption,
attestation, auditing, and others. From the steps shown in Fig. 3, the auditing process
will utilize TEEs to encrypt the communication with the network controller, to pro-
tect the processes of preparing the retrieved information, to calculate the hash values,
and to compare them. This usage of the enclave can impact the performance of the
auditing process, and its use needs to be limited to only the necessary as shown in
previous research [11].
3.2 Performance Evaluation
The two main metrics for software-defined networks [12] are the throughput, how
much information is sent per second, and latency, the time between a request and
its response. Those two metrics combined with the analysis of the system resource
usage (memory, processor, etc.) will be used to evaluate the process performance
and impact on the system. Specifically to the Intel SGX TEE, it will be important to
evaluate the time spent on e/ocalls and the processing time spent inside the enclaves.
Previous research [13] has shown the high impact that e/ocalls causes to the system.
4 Discussion
The proposed architecture aimed to enforce the auditing of running switches should
consider security by design, that is, from its conception, a separation of trusted
and untrusted software must be identified in order to decide which/what should run
Fig. 6 Representation of a scenario where: 1 shows the network working properly; 2 shows the
hijacking of the switch; 3 the auditing process warning the network controller; 4 the network
controller isolating the malicious switch and diverts the host traffic to a new and trusted switch
inside trusted execution environments. Similarly as developed in previous researches

on auditing of software [11], it was not possible to guarantee that the auditing pro-
cess would not affect the system performance without a well-defined system where
a benchmark could be performed. Other questions, such as what would be the peri-
odicity of the memory auditing should be answered from demand performance tests
of the implemented system. For performance tests, a few scenarios were envisioned,
including single and/or multiple switches starting with and without invalid entries;
single and/or multiple switches having an entry being changed randomly; and a switch
being hijacked (not responding to its true network controller). The last scenario is
exemplified in Fig. 6, where after a switch has been hijacked, the audit process warns
the network controller that acts to isolate the compromised switch and divert the host
traffic to a new and trusted switch.
5 Conclusion
At this stage, the proposed architecture is theoretically adequate to cater the prob-
lem of identifying compromised switches in software-defined networks while not
demanding much of the system resources, but, as a work in progress, the results of
this work are still under development and analysis. The process is able to identify
the conformity of the flow tables using a simple hash value method. The proof of
concept under development utilizes a simplified SDN architecture running on the
Intel SGX. As future study, we propose the implementation of an architecture able
to audit the switch execution environment and to evaluate the impact of securing the
controller-switch communication with Intel SGX to further ensure software-defined
network security.
Acknowledgements This research work explores possible applications for TEE and secure and
scalable cloud applications as part of the EU-BR SecureCloud project. The project has been receiving
funds granted from the 3rd EU-BR Coordinated Call (Brazilian Ministry of Science, Technology and
Innovation, MCTIC/RNP, BR grant agreements 2550, 2549, 2553, 2552 and 2568) and European
Union Horizon 2020 programme—EU Grant Agreement 690111). The project is also supported
by the Swiss State Secretariat for Education, Research and Innovation (SERI). This research work
also explores possible security solutions for forensics applications under the CAPES Pró-Forenses
Project 025/2014.
References
1. Nunes, B.A.A., Mendonca, M., Nguyen, X., Obraczka, K., Turletti, T.: A survey of software-
defined networking: past, present, and future of programmable networks. IEEE Commun. Surv.
Tutor. 16(3), 1617–1634, (2014). (Third Quarter).https://doi.org/10.1109/SURV.2014.012214.
00180
2. Nadeau, T.D., Gray, W.K., SDN - Software Defined Networks: O’Reilly (2013). ISBN:
1449342426
3. Van Trung, P., Huong, T.T., Van Tuyen, D., Duc, D.M., Thanh, N.H., Marshall, A.: A multi-
criteria-based DDoS-attack prevention solution using software defined networking. In: 2015
International Conference on Advanced Technologies for Communications (ATC), pp. 308-313.
Ho Chi Minh City (2015). https://doi.org/10.1109/ATC.2015.7388340
4. Zhou, H., et al.: SDN-RDCD: a real-time and reliable method for detecting compromised SDN
devices. IEEE/ACM Trans. Netw. 26(5), 2048–2061 (2018). https://doi.org/10.1109/TNET.
2018.2859483
5. Dabbagh, M., Hamdaoui, B., Guizani, M., Rayes, A.: Software-defined networking security:
pros and cons. IEEE Commun. Mag. 53(6), 73–79 (2015). https://doi.org/10.1109/MCOM.
2015.7120048
6. Price, D.: A guide to cyber intelligence. J. US Intell. Stud. 21(1) (2014–2015)
7. Robterson, J., Riley, M.: The big hack: how China used a tiny chip to infiltrate U.S. companies.
Bloomberg Businessweek, 4 Oct 2018
8. Intel Corporation, Intel Software Guard Extensions (Intel SGX), Website, Accessed December
14 2018
9. Open Networking Foundation, OpenFlow Switch Specification, Version 1.5.1 (Protocol version
0x06), 26 March 2015
10. Newman, L.H.: Spectre-Like Flaw Undermines Intel Processors’ Most Secure Element. Wired,
14 Aug 2018
11. Pereira, L., et al.: Using Intel SGX to enforce auditing of running software in insecure envi-
ronments. In: The 10th IEEE International Conference on Cloud Computing Technology and
Science (2018)
12. Gelberger, A., Yemini, N., Giladi, R.: Performance analysis of software-defined networking
(SDN). In: IEEE 21st International Symposium on Modelling, Analysis and Simulation of
Computer and Telecommunication Systems, San Francisco, CA 2013, pp. 389–393 (2013).
https://doi.org/10.1109/MASCOTS.2013.58
13. Costa, R.S., Pigatto, D.F., Fonseca, K.V.O., Rosa, M.O.: Securing Video on Demand Content
with SGX: A Decryption Performance Evaluation in Client-Side, Simposio Brasileiro em
Segurança da Informação e de Sistemas Computacionais (SBSeg), [S.l.], pp. 127–140 (2018)
Mobile Communication Systems:
Evolution and Security
Teresa Guarda, Maria Fernanda Augusto, Isabel Lopes, José Avelino Victor,
Álvaro Rocha and Lilian Molina
Abstract With the evolution and popularization of mobile communication systems,

especially for access to data services, it becomes increasingly important to have a
closer look at security systems. From analog systems to the fourth generation, many
changes and evolutions occurred. This makes possible to question whether mobile
networks have sufficiently secure systems, and what vulnerabilities we should be
aware of. Due to the widespread risk of attacks on mobile communications, telecom-
munication companies are increasingly emphasizing telecommunication security
T. Guarda (B) · M. F. Augusto · L. Molina

Universidad Estatal Península de Santa Elena—UPSE, La Libertad, Ecuador
e-mail: tguarda@gmail.com
M. F. Augusto
e-mail: mfg.augusto@gmail.com
L. Molina
e-mail: lmolina@upse.edu.ec
T. Guarda
Universidad de Las Fuerzas Armadas-ESPE, Sangolqui, Quito, Ecuador
T. Guarda · I. Lopes
Algoritmi Centre, Minho University, Guimarães, Portugal
M. F. Augusto
BITrum-Research Group, Calle San Lorenzo 2, 24007 León, Spain
I. Lopes
UNIAG (Applied Management Research Unit), Polytechnic Institute of Bragança, Bragança,
Portugal
e-mail: isalopes@ipb.pt
J. A. Victor
Instituto Politécnico da Maia, Maia, Portugal
e-mail: javemor@ismai.com
Instituto Universitário da Maia, Maia, Portugal
Á. Rocha
Universidade de Coimbra, Coimbra, Portugal
e-mail: amrocha@dei.uc.pt

https://doi.org/10.1007/978-981-13-9155-2_8
88 T. Guarda et al.
technologies specifically in fourth generation (4G) technology, in order to ensure

the main security functions to provide an efficient service to its customers.
Keywords Telecommunications · Security · G technologies · Mobile

communications
1 Introduction
Networks that provide mobility services have evolved rapidly, going far beyond
their original purpose of providing mobile voice services. With 4G/LTE (Long-Term
Evolution), which are fully IP-based, it is possible to support data transmission at
high speeds, in the order of Mbps, in addition to traditional voice services. This
structure is critical to building new business models based on mobile services such
as the Internet of things, cloud computing, video content distribution, application
providers, and more [1].
With 4G technology, new security issues have emerged, gaining great relevance
over LTE/4G networks because they are based on the IP protocol. As a result, new
threats to a mobile service provider can be identified as a breach of confidentiality,
breach of integrity, availability reduction, and fraud [1, 2].
Threats are becoming more important as mobile subscribers are having a more
and more powerful smartphone because the 4G architecture considers the use of
external networks not controlled by operators as new forms of access, and threats
are spreading more and more in IP networks.
In this scenario, mobile operators must prepare for challenges by building a robust
and flexible security architecture, while at the same time ensuring at least interoper-
ability with legacy systems (2G/3G/4G networks), a smooth transition to IPv6 and
support for virtualizing infrastructure [3].
2 Background
The first generation (1G) was a technology widely used in the 1980s by the Advanced
Mobile Phone System (AMPS) network and was made up of analog systems, with
only voice transmission possible. It was gradually replaced by the digital signal (2G)
[4].
Second generation (2D) became popular in the 1990s when the mobile phone
began to be more used by the population, and the signal changed from analog to a
digital signal with the most widely used technologies Code Division Multiple Access
(CDMA) and Global System for Mobile Communications (GSM).
CDMA is a method of accessing communication systems that can be used for
both mobile telephony and GPS. GSM—it is a method used to send and receive
Short Message Service (SMS) [3, 4].
Mobile Communication Systems: Evolution and Security 89
Between the second generation (2G) and the third generation (3G), a number of
developments have taken place: Deploy the General Packet Radio Service (GPRS)
wireless data transmission system for telephones that send and receive short mes-
sages; the 3G pioneer standard, Enhanced Data rates for GSM Evolution (EDGE), is
being used; set the High-Speed Circuit-Switched Data (HSCSD) as a specification
for transferring data over GSM networks with digital voice signals and channels; and
the Wireless Application Protocol (WAP), that is, the basic mobile Internet system
for any mobile device, allowing easy access of content to original pages [1, 4].
As there was evolution in the technology of communication of wireless networks
and evolution of the devices that until then only allowed to make analog connection,
there was the advance in the digital technology that allowed not only the voice call,
but also the communication of data through images, audio, and video. With the
new form of wireless network communication, large companies in the global mobile
device landscape began to vie for a growing market of users looking for devices
capable of communicating over a more advanced wireless network, services ranging
from sending messages to sending image, audio, or video and technology called 3G
made it all possible.
3G technology emerged in the 2000s, when mobile phones finally became smart-
phones with their own operating systems, Wi-Fi Internet, better quality cameras,
Bluetooth signal, and more. 3G technology has come to enable smartphones to func-
tion properly, as they have better capabilities. Highlights include data transmission
over long distances in mobile environments. 3G mobile technology is capable of
supporting a greater number of customers and data, especially in urban areas, by
increasing data rates [4].
The fourth generation started in 2010, being entirely based on IP, it can be under-
stood as a system and network at the same time, being much more than the mere
extension of 3G technology. In this way, it is possible to converge cable, wireless, and
computer networks with electronic devices, exceeding access speeds of 100 Mbps
and reaching up to 1 Gbps [3] (Fig. 1). It is anticipated that mobile networks of 5G
will be 10–20 times faster than those of previous generation (4G).
The world does not stop, and 5G technology is already taking its initial steps.
Mobile devices and computers are increasingly interconnected, transmitting more
data, and soon they will need a new network to handle all that speed and information.
3 4G Technology
It is not a defined technology or standard, but a collection of technologies and pro-

tocols that allow maximum performance. This convergence of technologies arises
from the need to group different standards in use, in order to delimit the scope of
operation of each of them, and integrate all communication possibilities in a single
device, in a transparent way to the user.
4G is the fourth generation Internet connection technology; at the moment, it is
the new technology present in the majority of mobile phones worldwide. Being a
90 T. Guarda et al.
Fig. 1 G technologies speed: from 1G to 4G
system based on packets containing fewer network elements, among its main features
are providing high performance in terms of higher connection speed and decreased
latency when accessing the connection compared to its predecessor technologies [5].
This technology works with the TCP/IP architecture and its set of protocols, and
because of the fact that it works with the TCP/IP protocol group, this technology has
greater vulnerabilities in its security compared to its previous technologies [4].
In 4G technology, it is a real challenge to ensure the security of communications,
and one of its main factors to take into account is that wireless technology does
not have fixed connection points and must focus on the mobility and flexibility
requirements of the 4G technology [6].
The 4G technology was initially designed to support a bandwidth with large
capacity, are designed based on flat architectures with fewer network elements, low
latency, and fast convergence for adapt to new wireless network technologies.
The 4G architecture was developed taking into account the principles of security
from its inception and design based on the following security features: security of
access to the network; security of the network domain; security of the user’s domain;
and security of the domain of the application [1].
4G technology is based on the use of IP technology in the core network and
access in order to support all services. In comparison with the old technologies, the
4G technology has a smaller number of elements in its implementation, basically as
all the technologies of the mobile communications have three networks where they
must follow until finally reaching the service to the station: core network; transport
network; and access network [6].
In the core network, we have the main controllers MME and SGW, where the
security parameters of the different 4G nodes are configured, the transport network
is responsible for transporting the encapsulated service that leaves the core network
and it transports it to the access network through the different existing networks using
fiber optic in most cases, the access network is responsible for bringing the service
finally to the 4G nodes, the main technologies of the access network are optical fiber
and microwave [7].
4G technology is designed with strong cryptography techniques, mutual authen-
tication between the elements of an LTE network, these techniques are embedded in
its architecture, and industry regulators have evaluated and identified new security
vulnerabilities that can be mitigated moment of network implementation.
The main purpose of implementing the LTE standard was to provide a high-
performance radiofrequency access that allows high transmission and reception
speeds in mobile devices and allows coexistence with other systems [5, 6].
The LTE standard has enough flexibility to allow vendors to combine these differ-
ent modules into a single device or multiple devices, for example, separating MME
and SGW into different devices.
4 Vulnerabilities and Threats
In just over a decade, there has been an extraordinary growth associated with mobile
telephony, both in terms of number of users and technology, which went from the
first to the fourth generation. Due to this phenomenon, the mobile devices have
changed their status, becoming fewer and fewer phones and computers, and therefore
gradually becoming more susceptible to various forms of attacks.
An unprotected mobile device associated with an insecure network can result in
an invasion of individual privacy [8], allowing for facts such as tampering with data
or even theft, the intruder will be able to listen to your calls, read your text messages,
track your activity on the Internet and not know your geographical location.
A network to be minimally safe should follow some basic principles of security:
confidentiality; integrity; availability; and authenticity [9]. Confidentiality ensures
that access to information is linked only to legitimate entities, that is, those authorized
by the owner of the information. In the case of integrity, all manipulated informa-
tion maintains all original characteristics established by the owner of the informa-
tion, including changes, birth, maintenance, and end of the information. Availability
ensures that the information is always available for legitimate use, by those users
authorized by the owner of the information. With regard to authenticity, it is guar-
anteed that the information was produced, modified, or discarded by a particular
person, body, entity, or system in order to validate its origin.
Among the most common types of attacks in mobile networks are: unauthorized
access; denial of service; channel jamming; eavesdropping; session hijacking; man-
in-the-middle attack; and message forgery [10–12] (Table 1).
Due to the dynamism of the technology of a mobile network, there is no way
to offer any kind of definitive security certainty. However, there are many methods
to minimize or in several cases eliminate these attempts to break down the security
mechanisms of mobile networks [13].
92 T. Guarda et al.
Table 1 Most common types of attacks in mobile networks

Attack type Description
Unauthorized access If an authentication method is not properly applied or is
misconfigured, then an attacker can gain free access to the network
and use its services even though it is not authorized to do so
Denial of service Is caused by the excessive sending of data on the network, more
than it can support, leaving users without the network resources
available
Channel jamming It is a technique used by the invaders whose objective is to destroy
or degrade the signal of the air interface and thus to disable the
access of the legitimate users of that network leaving them
exposed to other networks
Eavesdropping If traffic in the air interface is not heavily encrypted, an attacker
could spy on or intercept important data or confidential phone calls
Session hijacking A malicious user can hijack an already established session and act
as a legitimate user of a particular base station
Man-in-the-middle attack An attacker can be between a cell phone and a network access
point to intercept messages
Message forgery If the communication channel is not enough secure, an attacker
can change the content of the messages in both directions without
the actual recipients realizing it
The fourth generation of mobile communication systems had the development

principle to be fully IP, that is, all elements, except the terminal, should communicate
through purely IP interfaces [3]. As a result, various authentication, integrity, and
confidentiality techniques existing and used in IP networks are used in 4G, such as
IPSec, Certificate Management Protocol (CMP) and Transport Layer Security (TLS).
A fourth generation network uses less physical elements, equipment, but each with
more functions. It is also planned to use cells to cover small areas, such as homes
or offices, called Femto Cells (HeNB-House and NB). In the LTE, the encryption is
applied to the user and control data, and the integrity protection is applied only to
the control data, through the Packet Data Convergence Protocol (PDCP) [3, 14].
Among the main vulnerabilities of 4G, we can mention the vulnerabilities due
to the joint operation with 2G and 3G networks and attacks from the IP interfaces;
as well as 3G networks where many users use smartphones, and attacks can exploit
operating system vulnerabilities through malicious code hidden in applications, text
messages, or multimedia messages [7, 14].
Breaches discovered by researchers in LTE networks, which can lead to problems
of privacy of user data. According to the researchers at Ruhr-Universität Bochum
and the University of New York, there are three types of attacks that can happen due
to the security flaw found in the 4G LTE data link layer. The first two attacks are
passive, meaning the attacker cannot interfere with the attacking network. In these
two cases, they are responsible for doing an identity mapping and for a method that
can perform fingerprints of Web sites. The third, in turn, is an active attack, which
the researchers nicknamed aLTeR.
ALTeR uses a separate device by which the cybercriminals can simulate the orig-
inal network and trick the victim so that they unknowingly reveal information. This
method, however, is quite complex since the attacker first redirects network connec-
tions by performing DNS spoofing and then exploits a specific fault in LTE.
Apparently, the attack can happen to anyone, but because it is a complex method
of exploiting vulnerability, it is likely that such invasions will occur with people of
interest, i.e., politicians, journalists, and the like. The researchers further conclude
that while the demand to perform these actions is exhaustive, current technology
allows cybercriminals to evolve and even perform them outside of LTE; even those
that support authenticated cryptography [2].
5 Conclusions
First generation networks have almost no security features, gaining second generation
attention from developers, creating systems of authentication and confidentiality,
which were previously non-existent.
In turn, 3G networks have brought integrity techniques as well as advances in
existing security requirements, such as increasing the size of 128-bit security keys.
Fourth generation networks, predominantly IP-based, add security techniques until
then used only in the network core.
In 4G, all the interfaces except the radio are IP, and this facilitates and makes
flexible the installation and maintenance of the networks. Authentication and key
exchange systems have been improved compared to systems applied to 2G and 3G
networks using HMAC-SHA-256 switches. Encryption uses authentication vectors
that are incompatible with those used in 3G networks and to maintain user com-
patibility, the USIM must check the separation bit. The algorithms used are AES
and SNOW 3G, which use 128-bit keys capable of processing 256-bit keys in an
upcoming release.
The fifth generation (5G) is already in the test phase, with promises to be imple-
mented by the middle of 2020, to revolutionize its services, uniting 3G technologies
such as voice, data, videos; and services available in 4G such as digital broadband,
IP communication, HD transmission, dynamism of information exchange, among
others.
5G technology will better serve the radio spectrum and enable more devices to
access the mobile Internet at the same time, promising faster data download and
upload speeds, broader coverage, and more stable connections.
The purpose of 5G is not to be a new technology, but to unify the best in 3G and 4G
technology. In addition, it intends to improve the hardware part such as the devices
themselves that will serve as base stations for connection between the devices, thus,
there will be no loss of signal or interruption of services, thus managing to maintain
the quality of connection.
94 T. Guarda et al.
References
1. Dahlman, E., Parkvall, S., Skold, J., Beming, P.: 3G evolution: HSPA and LTE for mobile
broadband. Academic Press (2010)
2. Rupprecht, D., Kohls, K., Holz, T., Pöpper, C.: Breaking LTE on layer two. In: IEEE Symposium
on Security & Privacy, pp. 2–16 (2019)
3. Rost, P., Banchs, A., Berberana, I., Breitbach, M., Doll, M., Droste, H., Sayadi, B.: Mobile
network architecture evolution toward 5G. IEEE Commun. Mag. 54(5), 84–91 (2016)
4. Gupta, P.: Evolvement of mobile generations: 1G to 5G. Int. J. Technol. Res. Eng. 1, 152–157
(2013)
5. Ratasuk, R., Prasad, A., Li, Z., Ghosh, A., Uusitalo, M.A.: Recent advancements in M2M
communications in 4G networks and evolution towards 5G. In: 18th International Conference
on Intelligence in Next Generation Networks (ICIN) (2015)
6. Dahlman, E., Parkvall, S., Skold, J.: 4G, LTE-advanced Pro and the road to 5G. Academic
Press (2016)
7. Sun, K., Kim, Y.: Gap analysis for adapting the distributed mobility management model in
4G/5G mobile networks. In: IEEE Conference on Network Softwarization (NetSoft) (2017)
8. Garba, A.B., Armarego, J., Murray, D., Kenworthy, W.: Review of the information security and
privacy challenges in bring your own device (BYOD) environments. J. Inf. Priv. Secur. 11(1),
38–54 (2015)
9. Li, S., Tryfonas, T., Li, H.: The internet of things: a security point of view. Internet Res. 26(2),
337–359 (2016)
10. Maxim, M., Pollino, D.: Wireless Security. McGraw-Hill/Osborne (2002)
11. Yang, H., Ricciato, F., Lu, S., Zhang, L.: Securing a wireless world. Proc. IEEE 94(2), 442–454
(2006)
12. Welch, D., Lathrop, S.: Wireless security threat taxonomy. In: Information Assurance Work-
shop, pp. 78–83 (2003)
13. Liyanage, M., Abro, A.B., Ylianttila, M., Gurtov, A.: Opportunities and challenges of software-
defined mobile networks in network security. IEEE Secur. Priv. 14(4), 34–44 (2016)
14. Mavoungou, S., Kaddoum, G., Taha, M., Matar, G.: Survey on threats and attacks on mo-bile
networks. IEEE Access 4, 4543–4572 (2016)
Part II
Computer Networks, Mobility and
Pervasive Systems
Evaluating Trail Formation in
Collaborative UAV Networks with Lethal
Threats
Nícolas Pereira Borges, Cinara G. Ghedini

and Carlos Henrique Costa Ribeiro
Abstract Aerial vehicles are widely used for missions in hostile environments.
Some of these missions can benefit from the use of unmanned aerial vehicles (UAVs)
working collaboratively to achieve a common goal. In such scenarios, the use of
stealth approaches is essential for mission success, because it allows missions to
be conducted in the proximity of threat radars and also reduces the probability of a
UAV being spotted by enemies by enemies. In this direction, we extend a model that
allows autonomous UAVs to dynamically and collaboratively reduce their exposure to
threats by adopting a stealth policy in which they turn off their communication radars
once a threat is detected. The model also encompasses strategies for UAVs leaving
the threat regions as quickly as possible and estimating the future position of others
UAVs, thus trying to regroup to formation. The model was evaluated in simulations
using different hostility and weapon range distributions. The results demonstrated
that the proposed model is able to significantly improve the number of UAVs that
are able to complete the mission without being hit by any threat. In addition, the
proposed approach for calculating the goal position when a UAV is exposed to a
threat proves to be a key factor for missions performed in hostile environments as its
significantly reduces the fraction of UAVs hit by threats.
Keywords Unmanned aerial vehicles · Trail formation · Leave threat region ·

Hitting probability
N. P. Borges (B) · C. G. Ghedini · C. H. C. Ribeiro

Computer Science Division, Aeronautics Institute of Technology, São José dos Campos, SP,
Brazil
e-mail: nicolas@ita.br
C. G. Ghedini
e-mail: cinarag@gmail.com
C. H. C. Ribeiro
e-mail: carlos@ita.br

https://doi.org/10.1007/978-981-13-9155-2_9
98 N. P. Borges et al.
1 Introduction
Advanced air defense radar systems significantly reduced the traditional electronic
warfare effectiveness. Unmanned aerial vehicles (UAVs) are currently used as a re-
placement for traditional aircraft on potentially dangerous missions where threats
have to be detected and avoided. In addition, considering that autonomous UAVs are
able to make decisions faster than non-autonomous counterparts [1], it is possible
to connect different types of autonomous UAVs to dynamically and collaboratively
reduce their exposure to threats. When autonomous UAVs make decisions collabo-
ratively, they compose a collaborative UAV network (CN) [2].
Since the UAVs act collaboratively, the features of formation flying become a
key factor for the success of the mission. Formation flying is a technique that is
implemented by more than one aircraft. The aircraft keeps its positions against a
leader and maintain the formation shape during the flight. Different formations can
be used with different objectives, such as maximizing coverage area for surveillance
or rescue missions. Regarding offensive missions in hostile scenarios, one of the
most used formation flyings is the trail. This formation minimizes the exposure of
UAVs, thus improving their stealthiness [2]. Moreover, this formation also reduces
the fuel usage during the flight [3].
In this work, we focus on a stealth approach for trail formation proposed in
[2]. Their model performance was evaluated based on the minimization of the CN
exposure to threats, disregarding how the formation would behave in scenarios where
the UAVs can be hit by the enemies. In this sense, we extend this model to allow the
UAVs, which are exposed to enemy radars, to evade the threat regions, thus reducing
their overall time exposure to enemies and improving their chance to regroup with
the other UAVs. Additionally, considering that the trail formation is an effective
formation for attack missions and that UAVs may be hit during missions, we analyze
the impact of lethal threats on the mission accomplishment by verifying the fraction
of UAVs hit during the flight. Regarding lethal threats, we consider that our scenario
contains ground-based air defense systems [4].
The remaining of this paper is organized as follows. Section 2 presents the model
proposed in [2]. Section 3 details the proposed approach for UAVs to leave threat
regions. Section 4 discusses the hitting probability of a UAV. Experiments are pre-
sented in Sect. 5 and the results are discussed in Sect. 6. Finally, Sect. 7 concludes
the paper and outlines future research steps.
2 Collaborative UAV Networks
In this section, we describe the general aspects of collaborative UAV networks using
the model proposed in [2] as a reference. Thus, our scenarios are composed of UAVs,
all flying at the same height, and threats. For convenience, we will refer to this model
as reference model.
Evaluating Trail Formation in Collaborative UAV Networks … 99
Fig. 1 An instance of the

proposed scenario
Threats are represented by circles, defined by their center position (physical radar
position) and detection range. Considering that all enemy radars monitor critical
locations, we extended the scenario of the reference model to make the threats lethal,
that is, whenever a radar detects a UAV it automatically triggers the weapon systems
that aim to destroy it. Notice that this does not imply that the UAV will necessarily
be hit by the enemy. The likelihood of a UAV being hit by the weapon is addressed
further.
As in the reference model, UAVs flight from a start to a goal position using a
collaborative approach to detect and avoid threats through a stealth policy. Similarly
to [2], we adopt a simulation model over a time interval t = [t0 . . . tmax ]. However,
considering that the proposed scenario is composed of lethal threats, the simulation
for a single UAV ends when it reaches the goal position, is destroyed by the enemy
or due to the maximum simulation time. If the simulation of all UAVs ends before
tmax , the simulation also ends. A typical scenario consisting of UAVs performing
collaborative tasks in hostile environments is presented in Fig. 1 with threats depicted
by red circles.
2.1 Communication Model
Considering that the communication between UAVs is commonly relying on high-

frequency, very high-frequency or ultra-high frequency radars, it is assumed that each
UAV is equipped with one of these devices [5]. In this sense, a UAV can communicate
with others that are within its communication radius, establishing a network that can
be modeled as an undirected graph G. Regarding the graph, each UAV is a vertex
and each communication link between two UAVs is an edge.
2.2 Threat Detection
UAVs should be able to detect threats in flight. Therefore, as in the preliminary

model, it is assumed that each UAV is equipped with a passive receiver intended to
detect and identify enemy radar signals, named radar warning receiver (RWR). In
addition to threat detection, it can also estimate the position of the radar [6]. Our
model assumes that a threat is considered detected by a UAV if its range overlaps
the RWR range of the UAV. Accordingly, if a UAV detects the presence of a threat,
the threat position and detection range are precisely known.
2.3 Stealth Policy
In military missions, surveillance radars are expected to monitor strategic spots. Thus,
they are often paired with passive radars to enhance their detection capabilities. In
this regard, when a UAV is exposed to a surveillance radar, its communication with
neighbors can also be exposed to a passive radar, which may compromise the entire
mission. In such a scenario, for stealth purposes, the UAV shuts off its communication
radar while its RWR identifies a threat and connects it again when no threat is detected
by the RWR. This approach, called stealth dissemination, was based on the Try-and-
Bounce communication protocol described in [7]. Notice that the RWR is a passive
receiver, so it does not have to be disabled.
2.4 Formation Flying
According to [8], in a typical formation flying there are the roles of leader and
wingmen. In the context of the trail formation, the wingmen try to follow the trajectory
of the leader, taking the other aircraft as a reference to keep its own position in the
formation. The leader defines the trajectory to the goal.
For generating the trail formation, firstly, it is computed the path of all UAVs to
the goal position. The UAV that is chosen to be the leader of a connected component
of the network is the UAV that has the shortest path to the goal. The other UAVs’
ranking is generated based on the length of their path to the leader: The closer a UAV
is to the leader, the lower is its ranking index, i.e., the higher is its ranking position.1
The formation ranking, however, cannot be static because the connection between
two UAVs may be lost. This can occur because both the distance between the UAVs
is greater than its communication range or a UAV shuts off its communication radar.
Hence, it is essential for the formation flying to detect these situations so that UAVs
can perform actions to reestablish the formation as quickly as possible [8].
1 The rank of the leader is 1.

Another scenario where it is necessary to update the formation flying is when

there are two connected components close enough to be merged. For updating a
formation flying, the number of connected UAVs within a formation ranked as 1,
i.e., the number of leaders, is verified. This is necessary because there can only be
one leader in trail formation. In case of more than one leader, the one with the shortest
path to the goal is designated as the leader. We choose as leader the UAV with the
shortest path to the goal instead of the UAV that is actually closest to the goal because
the scenario contains threats that must be avoided, and it is not possible to ensure
that the UAV that is closest to the goal will reach the goal earlier [2]. In the context of
this work, the formation flying is responsible for defining the goal position of each
UAV. Once the goal position is defined, the next step consists in computing the path
planning.
2.5 Path Planning
There are several path planning algorithms for UAVs and, among those that consider
obstacle or threat avoidance, one of the most frequently referenced in the literature
is the Rapidly Exploring Random Tree (RRT) [9].
This algorithm is based on the construction of a tree of feasible trajectories through
randomly generated intermediate points. The branches are created as long as they do
not pass through any threat regions. One limitation of the RRT for UAV applications,
especially for fixed-wings, is that there is no restriction on invalid movements to be
performed. In this direction, it is possible to adapt the algorithm for only generating
branches in the tree if the angle between the new intermediate goal point and the
nearest node in the tree is in a specific range.
Notice that the output of the RRT is a path composed of consecutive lines. These
lines can be processed to reduce the length of the resulting path. Since the path of
all UAVs is computed and reduced, they must fly according to a control law, referred
here as displacement model [2].
2.6 Displacement Model
Given the path of each UAV, defined by the path planning, it is possible to define
the control law responsible for implementing the UAVs movements. Consider that
each UAV is modeled as a single integrator system and its velocity can be directly
controlled, namely:
ṗi = u i , (1)
where u i ∈ Rm is a control input that must be set so that a global objective can be
achieved.
The objective of this control law is to drive the UAV i to follow a path avoiding
threat regions. Thus, consider the path Pathγ (i) of the UAV i, which is located at
the position pi at a moment t. By using interpolation, it is possible to estimate the
position xγi in which the UAV should be at a moment t + 1, by following Pathγ (i).
Thus, the control law that drives the UAV i to follow the path is defined by:
x i − pi
u i = xγi − p β (t) , (2)
γ i
where β (t) ∈ R is the linear velocity of the UAV.

For performance purposes, the path planning algorithm does not have to be exe-
cuted at each iteration t. For instance, considering a scenario where the UAV i does
not shut down its communication radar and that the straight segment that passes
through pi and xγi does not intercept any threat regions, instead of running RRT
again it is possible to update the previously computed path. The case in which the
UAV is inside a threat region is analogous. However, when the UAV turns on the
communication radar again the path needs to be recalculated.
3 Leaving the Threat Region
The stealth policy of disabling the communications radar, described in the preliminary
model, considers that a UAV in a threat region cannot communicate with other UAVs.
Besides, it should still act independently of the network and outline a path to leave
the threat area toward the goal position. The reference model considers that a UAV
that has entered a threat region will trace a path to the goal position and will fly until
it leaves the threat region. We extend this approach, considering different strategies,
described as follows.
Regarding strategies to leave the threat region, one of the most intuitive is to
leave it as quickly as possible. However, the collaboration between UAVs is crucial
for stealthiness [2]. Therefore, it is also important providing UAVs with means that
allow them to trace a path that improves their chance of regrouping with the others
UAVs in the network.
The procedure consists in generating candidate points that allow the UAV that is
exposed to a threat to leave the threat region as quickly as possible. For this, it is
assumed that a UAV has two candidate points to use as a reference, named here as Pi
and Pi . Given that, the linear regression of the last position of the UAV L Pr oj and its
current position Pi are computed. The candidate points Pi and Pi are the first points
that are outside the threat region, considering the line L Per p which is perpendicular
to L Pr oj and pass through the center of the threat.
The procedure for computing a goal for the UAV in a threat region assumes
that each UAV knows its own position Pi and its neighbor positions. Thus, the next
candidate point Pi results from the estimation of future positions of connected UAVs.
Considering that all wingmen follow the leader in the trail formation, it is possible
Fig. 2 Candidate points for computing path planning to leave threat region
to estimate the future position of UAVs by computing the path of a UAV’s successor
in formation. The UAV that is exposed to a threat computes the path assuming its
successor’s position in the formation as its current position. Moreover, it is assumed
that its successor knows about the threat. The candidate point Pi is the point in the
successor path that is closest to Pi or Pi . Figure 2 illustrates a scenario in which this
procedure is applied.
Give the three candidate points, the last step consists in selecting which one will
be the goal position to leave the threat area. Considering that the collaboration is
crucial for CN, in case that the length of the path between the current UAV position
Pi and Pi is lower than a threshold, Pi will be selected as the goal. Otherwise, the
goal will be the candidate point (Pi or Pi ) which results in the shortest path length
from Pi .
4 Hitting Probability
For computing the probability of a UAV being hit by a threat, several factors must be
taken into account. For instance, the position and characteristics of both the UAV and
the weapon, as well as the weather conditions must be considered. A straightforward
way to evaluate the hitting probability of a UAV is addressed in [10]. However, it is
based only on computed path. We adapted it for computing the probability of a UAV
being hit considering only its current position Pi , as follows:

0 if dw > rweapon
Pweapon (Pi ) = (3)
Dw2 /dw2 otherwise.
where Dw is the weapon attack range, dw is the distance between Pi and the weapon
and rweapon is the hitting radius of the weapon.
Notice that as the UAV approaches to a weapon its probability of being hit in-
creases. We assume that the position of the weapon and the surveillance radar is the
same. Besides, for validation purposes, we consider that a UAV that is inside a threat
region can be hit by it, i.e., rweapon is the same value of the detection range of the
corresponding surveillance radar. In addition, once a UAV is hit it cannot proceed
the mission, i.e., the UAV is removed from the CN.
5 Experimental Setup
The proposed model was evaluated through simulations, using MATLAB® R2018a.
The UAVs fly from a start to a goal point in different scenarios. Each scenario
is composed of a start and a goal position—same position for all the scenarios—
approximately 17 km apart, as well as threats and UAVs, and last at most 3,000 s,
with a 1 second time step. Three sets of scenarios were considered: low, medium and
high hostility, with 2, 8 and 16 threats, respectively. The location of each threat in
each scenario was randomly generated.
Each UAV possesses a communication and RWR ranges of 200 m, while threats
have 700 m of detection range and hitting radius. The model performance was evalu-
ated considering networks with 10 UAVs. The initial network topology is always the
same, since the variation of the threat displacement is more relevant to the scenario
than the UAVs start positioning. Moreover, it is assumed that the trail formation starts
established.
For model validation, we evaluate the fraction of UAVs that have been hit during
the simulation. This metric assumes values in the range [0, 1]. The closer the value
is to 1, the higher the fraction of UAV hit during the mission.
With regard to the limited weapon attack range distribution (Eq. 3), let N t be
the total number of threats, [ri, r f ] be the possible values interval for the weapon
range, and [ri , r f ] be a subset of [ri, r f ]. Moreover, let f d 2 be the fraction of
threats that will have weapon range in the interval [ri , r f ]. For instance, setting
N t = 20, ri = 100, r f = 600, ri = 100, r f = 250 and f d = 0.8, means that 16
threats will have weapon range between 100 and 250 ((80% of 20) and 4 threats
will have weapon range between 250 and 600 (20% of 20)). For properly evaluating
the impact of different weapon range distributions on the model performance, we
predefined weapon ranges in three intervals: (100, 250] (ψ = 1), (250, 400] (ψ = 2)
and (400, 550] (ψ = 3). The results were confronted with a uniform weapon range
distribution (ψ = 0).
2 The value of f d adopted in this work was 0.8.

Simulations were carried out for 100 scenarios with 10 UAVs, for each hostil-
ity level and weapon range distribution. The results are then the average of these
scenarios. Regarding the parameters of the CN, simulations were performed con-
sidering the stealth approach active. For convenience, we adopted ρ = 1 when this
approach is active and ρ = 0 when it is inactive. For ρ = 0, the UAVs fly directly
to the goal position without the threat detection and avoidance policy. Moreover, we
performed simulations with and without the approach for leaving threat regions. For
convenience, we adopted φ = 1 when this approach is active and φ = 0, when it is
inactive. Thus, we investigate 3 combinations of parameters: (ρ = 0, φ = 0), (ρ = 0,
φ = 1) and (ρ = 1, φ = 1). Notice that the combination of parameters (ρ = 1, φ = 0)
corresponds to the reference model, while the combination (ρ = 1, φ = 1) simulates
the extension of this model.
6 Results and Discussion
This section presents and discusses the results of the experiments. The reference
model results show that the exposure of UAVs reduces when the stealth policy is
active [2]. Once we introduced lethal threats in the scenario, the fraction of UAVs
that have been hit during the mission significantly reduces when the stealth policy is
active, as shown in Fig. 3. This occurs because the exposure of UAVs to threats also
reduces, as described in the preliminary model.
Fraction of UAVs hit ( =0) Fraction of UAVs hit ( =1)

Fraction of UAVs hit
1 1
=0 =0 =0 =0
0.8 =1 =0 0.8 =1 =0
=1 =1 =1 =1
0.6 0.6
0.4 0.4
0.2 0.2
0 0
Low Medium High Low Medium High
Hostility level of scenarios Hostility level of scenarios
(a) Results with ψ = 0 (b) Results with ψ = 1
Fraction of UAVs hit ( =2) Fraction of UAVs hit ( =3)

1 1
=0 =0 =0 =0
0.8 =1 =0 0.8 =1 =0
=1 =1 =1 =1
0.6 0.6
0.4 0.4
0.2 0.2
0 0
Low Medium High Low Medium High
Hostility level of scenarios Hostility level of scenarios
(c) Results with ψ = 2 (d) Results with ψ = 3
Fig. 3 Fraction of UAVs hit, varying the weapon range distribution

Notice that as the hostility level of the scenario increases, the fraction of UAVs
hit also increases, especially when the stealth policy is inactive (ρ = 0). In addition,
for this case, the weapon range distribution does not result in significant variation of
UAVs hit. The first reason for this relies on the fact that the fraction of UAVs hit, for
the medium and high hostility level, is already close to 1. In the case of low hostility
level, the variation was insignificant because the UAVs can be exposed, at most, to
only 2 threats during the mission.
When the stealth policy is active (ρ = 1), the weapon range distribution signifi-
cantly affects the hitting probability of UAVs. The fraction of UAVs increases when
the values of ψ were larger. For the case in which the weapon range distribution was
uniform (ψ = 0), the fraction of UAVs hit was closer to the case in which weapon
range distribution was intermediary, i.e., ψ = 2.
The approach for leaving threat areas proposed in this paper proved to be a key
factor regarding the hitting probability of UAVs. The fraction of UAVs hit by threats
was always lower when this approach was active (ρ = 1 and φ = 1). The importance
of this approach becomes more evident as long as the values of weapon range dis-
tribution and hostility level of the scenario increases. For instance, considering the
high hostility level and ψ = 3, the fraction of UAVs hit decreases in approximately
300%, when this approach is active, in comparison with the approach ρ = 1 and
φ = 0. This occurs for three reasons. The first is because the overall time of UAVs
exposure to threats reduces. The second is due to the fact that our approach makes
the UAV avoid the center of the threat, where the hitting probability is larger. The
last one is because it increases the probability of the UAV that turned off the radar re-
connecting to the network, increasing the stealth level of the network; consequently,
reducing the number of UAV exposed.
In this paper, we presented an extended model to improve stealth in collaborative

UAV networks under local communication constraints, using trail formation flying.
We proposed an approach to evade threat regions, when a UAV is exposed to an
enemy radar. Moreover, considering that the trail formation is an effective formation
for attack missions, and that UAVs may be hit during missions, we also conducted
an analysis of the impact of hit UAVs on the CN.
The approach was validated through simulations, varying the hostility of the sce-
nario and the weapon attack range of threats. The fraction of UAVs hit decreases
when the stealth procedure is active. Also, it was shown the importance of the ap-
proach for leaving threat regions, when a UAV is exposed to a threat. The proposed
approach provides promising results, considering the fraction of UAVs hit by threats
was significantly lower when this approach was active.
The proposed model does not consider the maximum operation time of the UAVs
to perform a mission. Thus, the path length calculated by the UAVs may be im-
practical for real applications. In this context, we propose as future work modeling
the maximum operation time of the UAVs during the computation of the path plan.
Moreover, we are assuming that the threats position and ranges are precisely esti-
mated. Evaluating the impact of the model for the cases which these are not correctly
estimated can be exploited. Finally, 3D modeling of the scenario will allow exploring
other stealth approaches considering control of yaw and pitch angles [10].
Acknowledgements The authors thank CAPES (Social Demands Program) and FAPESP (proc.
no. 2017/02055-8) for the financial support to carry out this research.
References
1. Bekmezci, I., Sahingoz, O.K., Temel, Ş.: Flying ad-hoc networks (FANETs): a survey. Ad Hoc
Netw. 11(3), 1254–1270 (2013)
2. Borges, N.P., Ghedini, C.G., Ribeiro, C.H.C.: A local communication model for improving
stealth in collaborative UAV networks. An. do Comput. Beach 249–258 (2017)
3. Cetin, O., Yilmaz, G.: Real-time autonomous UAV formation flight with collision and obstacle
avoidance in unknown environment. J. Intell. Robot. Syst. 84(1–4), 415–433 (2016)
4. Wållberg, K.: Method of fire control for gun-based anti-aircraft defence (2017). US Patent
9,625,236
5. Adamy, D.: EW 102: A Second Course in Electronic Warfare. Artech House, Boston (2004)
6. Gee, W.A.: Radar warning receiver (RWR) time-coincident pulse data extraction and process-
ing. In: 2012 IEEE Radar Conference, pp. 0752–0757 (2012). https://doi.org/10.1109/RADAR.
2012.6212238
7. Turgut, D., Turgut, B., Bölöni, L.: Stealthy dissemination in intruder tracking sensor networks.
In: 2009 IEEE 34th Conference on Local Computer Networks, pp. 22–29 (2009). https://doi.
org/10.1109/LCN.2009.5355150
8. Giulietti, F., Pollini, L., Innocenti, M.: Autonomous formation flight. IEEE Control Syst. 20(6),
34–44 (2000). https://doi.org/10.1109/37.887447
9. LaValle, S.M.: Rapidly-exploring random trees: a new tool for path planning (1998)
10. He, P., Dai, S.: Stealth coverage multi-path corridors planning for UAV fleet. In: Proceed-
ings of the 2013 International Conference on Mechatronic Sciences, Electric Engineering and
Computer (MEC), pp. 2922–2926. IEEE (2013)
Visual Analytics for the Reduction of Air
Pollution on Real-Time Data Derived
from WSN
Dorys Quiroz, Byron Guanochanga, Walter Fuertes, Diego Benítez,

Jenny Torres, Freddy Tapia and Theofilos Toulkkeridis
Abstract Nowadays, the contaminated and poor air quality that is inhaled by the
city population in industrialized cities around the world has led to one of the main
causes of premature death due to respiratory diseases. Therefore, the improvement
of air quality becomes a priority. In this sense, the current study aimed to design and
implement a visual analytics tool, in order to process large data sets, which have been
generated by wireless sensor networks (WSN), which measured different environ-
mental pollutants in real time. Hereby, the phases of the CRISP-DM methodology
have been applied as a reference to guide the process. In the data preparation phase,
programs have been implemented using Python. Then, the results have been stored in
collections within a MongoDB database. Furthermore, for the modeling and visual
exploration of the data, the Tableau tool has been used. The evaluation of the results
allowed to demonstrate certain behavior of air pollutants around the city, such as the
D. Quiroz · B. Guanochanga · W. Fuertes (B) · F. Tapia · T. Toulkkeridis

Departamento de Ciencias de la Computación, Universidad de las Fuerzas Armadas ESPE,
171-5-231-B Sangolquí, Ecuador
e-mail: wmfuertes@espe.edu.ec
D. Quiroz
e-mail: dsquiroz@espe.edu.ec
B. Guanochanga
e-mail: beguanochanga@espe.edu.ec
F. Tapia
e-mail: fmtapia@usfq.edu.ec
T. Toulkkeridis
e-mail: ttoulkeridis@espe.edu.ec
D. Benítez
Colegio de Ciencias e Ingenierías El Politécnico, Universidad San Francisco de Quito USFQ,
Campus Cumbayá, Casilla Postal, 17-1200-841 Quito, Ecuador
e-mail: dbenitez@usfq.edu.ec
J. Torres
Departamento de Informática y Ciencias de la Computación, Escuela Politécnica Nacional,
17-01-2759, Quito, Ecuador
e-mail: jenny.torres@epn.edu.ec

https://doi.org/10.1007/978-981-13-9155-2_10
110 D. Quiroz et al.
increased air pollution levels during daylight hours. Similarly, we discovered that the
presence of particulate material PM10 and PM2.5 is directly related to the increase
of the Air Quality Index for the city of Quito (IQCA). This leads to the conclusion
that our analysis may be useful as a support tool in the decision-making of public
policies for the reduction of air pollution.
Keywords Air pollution · Data mining · Wireless sensor networks · Visual

analytics
1 Introduction
Air pollution is caused by the presence of chemical substances or toxic compounds

derived mainly by a variety of anthropogenic activities that trigger adverse effects
on human well-being and the environment [1]. The presence of these pollutants
decreases air quality and harms the health and quality of life of people [2]. Due
to this, it is necessary to provide the corresponding agencies with tools that facili-
tate finding solutions that improve air quality. Therefore, both the industry and the
scientific community have been developing several solutions using wireless sensor
networks (WSN) in order to reduce the aforementioned negative effects. However,
most of these implementations have focused on the design of the electronic circuits,
the assembly of electronic sensor boards, the implementation of software bridges to
store data and, finally, the development of interfaces, which allow to know the mea-
surements. Nonetheless, all these efforts become unproductive, if the final users or
those responsible for this issues are not provided with tools that benefit the descriptive
analysis, in real time, depending on what the WSN determines.
The aim of the current study was to analyze the information derived from a WSN,
which we have assembled in order to determine air pollution. Hereby we used data
mining (DM) and visual analytics techniques. The large volumes of data obtained
from the WSN are stored in real time in a non-relational database. Sensor networks
were deployed in four sectors of the cities of Quito and Sangolquí, within the province
of Pichincha in north-central Ecuador. The CRISP-DM methodology has been used
in order to guide the DM process. In the data preparation phase, programs were
implemented in Python. Subsequently, the results have been stored in collections in
a new database. For the modeling and visual exploration of the data, we used the
Tableau tool.
The main contribution of this study is the generation of a visual analytical tool
that benefits the descriptive analysis to decision-making support and as an important
reference for the generation of public policies within the control of air pollution.
The results obtained demonstrate that the presence of particulate material PM10 and
PM2.5 increases the IQCA determination. Furthermore, during daylight rush hours,
with the highest density of vehicular traffic, we observed a significant increase in the
level of air pollution. This certainly has an impact on people’s health leading directly
to an increase of respiratory diseases such as asthma, bronchitis [2], among others.
Visual Analytics for the Reduction of Air Pollution … 111
The remainder of the article is structured as follows: Sect. 2 presents related works.
Section 3 shows an insight of the conducted research process, while in Sect. 4 we
explain the results obtained and the validation of this study. Section 5 finalizes with
the obtained conclusions.
2 Related Work
With the implementation of WSNs that monitor the concentration of air pollutants
in real time, hundreds of thousands of data have been available that will continue
to increase over time. Therefore, the design of visualization tools that facilitate the
interactive analysis of these data becomes indispensable. Thus, Liao in [3] developed
a visual analytics system for monitoring air quality data called AirVIS for China.
This system has the availability of allowing three views: GIS for spatial analysis of
data, scatter plots for temporal analysis, and a parallel coordinated view used for
multidimensional analysis. Similarly, [4] presents a system that uses visual analysis
for the study of air pollution in Honk Kong. The authors in such work propose sev-
eral visualizations that allow tasks such as the detection of correlations of variables,
comparative data, identification of trends, and predictions about climate and air qual-
ity. In [5], a semi-supervised learning approach was used to generate visualizations
that allow researchers to model the spatial correlation of the air quality of different
locations according to meteorological data, traffic flow, human mobility, structure of
road networks, and points of interest.
In [6], authors designed a system for visual analysis of air pollution data. They
added cluster concepts, transforming the data into 2D plots views, allowing them
to relate geographic areas, and identifying patterns among them. They also used
the Voronoi diagram to allow users to iterate with the spatial clusters. Currently,
they provide access to a large quantity of multidimensional spatiotemporal data and
greater precision generated by environmental monitoring sensors. This has generated
the need to create visualizations available to users with a more comprehensive and
interactive approach.
To facilitate the analysis in [7], a detailed study focused on the use of visual meth-
ods based on maps, calendars, and views of trends to identify lines and patterns was
presented. Within this same context in [8], the identification of correlations between
different regions where environmental pollution is determined was also presented.
In order to allow such approach, meteorological data (wind and its direction) were
used to calculate the lead/lag correlation. The Pearson correlation coefficient was
used to determine the pollution ratio between the different regions monitored along
different time spaces.
The attempt toward the implementation of smart cities based on (IoT) has been
described in [9], where an ecosystem of management and analysis of data was con-
structed, which ranges from processing, integration, analysis up to visualization.
Also, in [10], an analysis of the air pollutants in Bahrain was conducted using the
robust and nonparametric Theil–Sen method in order to study temporary quantitative
variations of PM10 and PM2.5 particles, NO2 , SO2 , and O3 , as well as trends and
effects of temperature, humidity, and wind.
The correlation analysis between different parameters of air quality determina-
tion is the center of the study reported in [11], which demonstrates how to analyze
distributed data in spatial temporal contexts. Additionally, the authors proposed a
visualization tool using a Time Correlation Partition Tree (TCP). Similarly, in [12],
authors presented an interesting approach in visual analytics in order to detect pat-
terns in numerical time series, specifically environmental sciences. Furthermore, a
guide for the correct design of large data visualizations has been analyzed in [13].
This is a fundamental guide for the visual management of different types of data.
Finally, in [14], large stored volumes of fixed data derived in real time from our
WSN have been analyzed using business intelligence (BI) and data mining (DM)
techniques.
All previous studies described in this section demonstrate the importance of pro-
viding managers with some descriptive analysis tools for decision-making support.
Therefore, our project has been designed to provide visual analytics as an input in
the improvement of public policies in this context.
3 Research Design
This section describes the process conducted based on CRISP-DM, which has been
chosen to provide a standardized description of the life cycle of a data analysis
project, which is analogous to the way used in software engineering with software
development life cycle models [15]. The CRISP-DM model covers the project phases,
their respective tasks, and the relationships between these tasks. Additionally, CRISP-
DM considers the data analysis process as a professional project, thus establishing
a much broader context that influences the development of the models. Descriptions
of the phases executed are described below.
3.1 Business Understanding
The air pollution study is important because of the possible negative effects on
peoples health [1]. Therefore, the World Health Organization (WHO) has designed
several air quality guidelines, which set the guidelines on the values of concentration
of pollutants and their permissible limits [16].
The Metropolitan District of Quito, from the Ministry of the Environment, has
implemented the Environmental Monitoring Network, establishing the Index of Air
Quality for the city of Quito (IQCA), which is based on the Ecuadorian Standard of
Air Quality [17, 18]. The established ranges to determine pollution levels are shown
in Table 1. Based on these guidelines, the information to be used was identified and
it was defined how this problem will be addressed through data mining.
Table 1 Index of Air Quality for the city of Quito (IQCA) [17]
Ranges Condition Condition from the health point of view
0–50 Desirable Air quality is considered satisfactory and environmental pollution has
little or no risk to health
50–100 Acceptable Air quality is acceptable. However, there may be minor effects on
health for individuals highly sensitive to environmental pollution
100–200 Caution Not healthy for individuals (chronic and convalescent)
200–300 Alarm Not healthy for the majority of the population
300–400 Alert Not healthy for the majority of the population and dangerous for
sensitive individuals
400–500 Emergency Dangerous for the entire population
Fig. 1 Example of a message (frame) in JSON format
3.2 Data Understanding
The data collection has been conducted automatically. The sensor nodes send the
determined information to each gateway node. The latter then sends the data to the
application server in the cloud, where they are stored in a non-relational database such
as MongoDB. Each message sent from the sensor nodes is composed by the concen-
tration measurements of each gas, as well as for the particulate material concentra-
tions, geographical location, time of measurement, temperature, pressure, humidity,
and identifier of each sensor node. Figure 1, illustrates the JSON format of the frame
that each sensor node sends.
3.3 Data Preparation
In this phase, we accomplished an extraction, transformation and loading process

(ETL) that included the cleaning, transformation, and storage of the data. The fol-
lowing tasks are specified below: (1) Based on the guidelines on air quality for the
city of Quito, the pollutant concentration of gases: CO, SO2 , NO2 , and particulate
material: PM2.5 and PM10, have been used in units of micro-grams per cubic meter
(ug/m3). (2) The information has been transformed to one half per minute. (3) In
the cleaning process, concentration values less than or equal to zero have been elim-
inated. (4) Ranges have been verified, and erroneous, meaningless, and duplicate
values have been identified and corrected. (5) Fields with the air quality index for
Fig. 2 Schematic architecture of the descriptive model
each gas have been added. Then, the total air quality index has been obtained. For
these tasks, several programs for data cleaning, transformation and subsequent stor-
age in several collections in MongoDB, have been implemented in Python.
3.4 Modeling
Figure 2 illustrates the architecture of the descriptive model of the data analysis
process conducted on air pollution data and meteorological variables. The model
has been composed of three individual layers: (1) The data source layer. This is
the information from the WSN, which has been stored in a central database. In the
same way, the information derived from external sources that have been used in the
preparation process. (2) The storage layer, composed of the base resulting from the
ETL process, where the data has been purified, transformed, treated, and then stored
in different collections of a database. Finally, (3) the visualization layer, for which the
Tableau tool has been applied in order to visualize and understand the data through
the generation of dashboards and interactive graphics, correlation and trend among
the studied variables.
3.5 Visual Modeling
With the execution of the tasks indicated in the data preparation, the obtained infor-
mation was stored in collections in a database. This allowed the task of data analytics
through the preparation of visualizations in order to study the behavior of pollutants
measured in the sensed sectors. In order to provide important information about these
data, a view model has been designed that follows the scheme shown in Fig. 3. The
Fig. 3 Elements and interrelations of visual modeling
used design criteria were the bar graphs that facilitate the comparison between the
different values of the pollutants. The visualization of the behavior of the pollutants
has been designed using time series in which trend lines were included. The iden-
tification of patterns is fundamental within this analysis; therefore, we have used
dispersion graphs. Additionally, we have included correlation matrices in order to
detect direct or inverse linear relationships between the different parameters; thus,
we calculated the Pearson correlation.
Furthermore, dashboards have been designed using the Tableau visualization tool.
In each of them, the behavior of air pollutants in the monitored sectors has been
reflected. Figure 4, for example, illustrates the first dashboard, where the behavior
of gases in each of the concentrators is presented based on their averaged index.
Specifically, it is the general visualization per concentrator and pollutant. From this
panel, users may select a concentrator and/or a contaminant to see the details in the
two remaining panels.
The second dashboard, illustrated in Fig. 5, presents the categories in which the
determinations have been framed. These categories have been those defined in the
regulations on the IQCA. This consist on IQCA by concentrator: It demonstrates the
index that has been calculated from the IQCA for each concentrator and the category
according to the Quito standard.
The identification of patterns and correlations is one of the most important points
in visual analysis. Therefore, one visualization was created, which indicates the
Pearson correlation values between the IQCA and the presence of pollutants (see
Fig. 6).
4 Evaluation of Results
The dashboard of Fig. 4, allows to analyze the behavior of the pollutants in the
Alameda Concentrator. As it can be seen, the bar graph allows to identify that PM2.5
has the highest concentration. The dashboard of Fig. 5 shows the panel with the heat
Fig. 4 General view of determinations per concentrator and pollutant
Fig. 5 The weekly behavior of the IQCA
maps, which allows, for example, to identify that the last week of October in the
Villa Flora Concentrator the IQCA went from DESIRABLE to ACCEPTABLE.
During the months of July, August, and September, in the sensors of Villa Flora
and ESPE nodes, a high influence of NO2 on the IQCA points out, while in Alameda
node a low correlation has been observed even being inverse due to its negative
coefficient. While, for San Diego node, it lacks of a clear conclusion, since it does
not provide of any significant correlation.
Fig. 6 Correlation of the IQCA versus contaminants
Nonetheless, the concentration of CO and SO2 in the concentrators of Villaflora

and ESPE has a high correlation with the IQCA in the first months of the moni-
toring. Since, for November this correlation is stabilized although its value is not
as high as PM10 and PM2.5, which has a certain degree of influence. For the La
Alameda concentrator, the CO does not have a noteworthy increase over time, while
SO2 marks a growing trend. At the San Diego concentrator, the data available for
November indicate that the correlation of both gases with the IQCA is insignificant.
All these analyses lead to the conclusion that PM10 and PM2.5 particulate matter
are the pollutants that mostly influence the increase of the IQCA values. Of all the
correlations analyzed, those of PM10 and PM2.5 have been the most significant and
are evidenced in the four concentrators. This most likely means, that with any PM10
increase, the presence of PM2.5 also increases.
Another significant correlation that was visualized in the first three months of
monitoring in Villa Flora and ESPE concentrators is that of NO2 and SO2 . However,
as of October, this correlation drops substantially. However, in Alameda concentrator,
the given correlation is inverse. That is, the greater the presence of NO2 , the lower
the presence of SO2 and vice versa. While in San Diego concentrator, the correlation
for the month of November behaves as in Villa Flora and ESPE nodes, although
there are no data for the month of December with which this correlation could have
been corroborated. A high correlation between CO and SO2 occurs only in the ESPE
concentrator, while in the other concentrators it remains to be insignificant.
5 Conclusions
This study has been conducted with the purpose of analyzing the determined data of
air pollutants collected by a WSN in real time. The CRISP-DM methodology was
followed for the DM process. In the preparation phase, an ETL process has been
operated, in which data were cleaned, treated, and transformed, the results of this
process were subsequently stored in collections within a non-relational database.
For the analysis and visualization of the obtained data, we applied the Tableau tool.
Views were generated, allowing pollutant behavior analysis as well as the correlation
and patterns analysis of pollutants at every minute in the sensed sectors. Therefore,
the descriptive analysis may be used as a significant decision-making support tool.
This study may be used as an aid for the generation of public policies that certainly
will better address environmental problems.
Acknowledgements The authors would like to thank the financial support of the Ecuadorian Cor-
poration for the Development of Research and the Academy (RED CEDIA) in the development of
this study, within the Project Grant CEPRA-XI-2017-13.
References
1. Kampa, M., Castanas, E.: Human health effects of air pollution. Environ. Pollut. 151(2), 362–
367 (2008)
2. WHO: Ambient air pollution: A global assessment of exposure and burden of disease. World
Health Organization, Geneva, Switzerland (2016)
3. Liao, Z., Peng, Y., Li, Y., Liang, X., Zhao, Y.: A web-based visual analytics system for air quality
monitoring data. In: 2014 22nd International Conference on Geoinformatics (GeoInformatics),
pp. 1–6. IEEE (2014)
4. Qu, H., Chan, W.Y., Xu, A., Chung, K.L., Lau, K.H., Guo, P.: Visual analysis of the air pollution
problem in hong kong. IEEE Trans. Vis. Comput. Graph. 13(6), 1408–1415 (2007)
5. Zheng, Y., Liu, F., Hsieh, H.: U-air: When urban air quality inference meets big data. In:
Proceedings of the 19th ACM SIGKDD International Conference on Knowledge Discovery
and Data Mining, pp. 1436–1444. ACM (2013)
6. Zhou, Z., Ye, Z., Liu, Y., Liu, F., Tao, Y., Su, W.: Visual analytics for spatial clusters of air-
quality data. IEEE Comput. Graph. Appl. 37(5), 98–105 (2017)
7. Du, Y., Ma, C., Wu, C., Xu, X., Guo, Y., Zhou, Y., Li, J.: A visual analytics approach for
station-based air quality data. Sensors 17(1), 30 (2016)
8. Du, Y., Abish, M., Zhou, L., Zhou, Y.: A correlation visual analytics system for air quality.
Chin. J. Electron. 27(5), 920–926 (2018)
9. Ahlers, D., Kraemer, F., Braten, A., Liu, X., Anthonisen, F., Driscoll, P., Krogstie, J.: Analysis
and visualization of urban emission measurements in smart cities. In: Proceedings of the 21st
International Conference on Extending Database Technology (EDBT) (2018)
10. Jassim, M.S., Coskuner, G., Munir, S.: Temporal analysis of air pollution and its relationship
with meteorological parameters in bahrain, 2006–2012. Arab. J. Geosci. 11(3), 62 (2018)
11. Fangzhou Guo Tianlong Gu, W.C.F.W.Q.W.L.S., Qu, H.: Visual exploration of air quality data
with a time-correlation partitioning tree based on information theory. ACM Trans. Interact.
Intell. Syst. 1(1) (2018). https://doi.org/10.1145/3182187.
12. Sips, M., Kothur, P., Unger, A., Hege, H.C., Dransch, D.: A visual analytics approach to
multiscale exploration of environmental time series. IEEE Trans. Vis. Comput. Graph. 12,
2899–2907 (2012)
13. Keim, D., Andrienko, G., Fekete, J.D., Görg, C., Kohlhammer, J., Melançon, G.: Visual analyt-
ics: Definition, process, and challenges. In: Information visualization, pp. 154–175. Springer,
Berlin (2008)
14. Fuertes, W., Alyssa, C., Torres, J., Benítez, D.S., Tapia, F., Toulkeridis, T.: Data analytics on
real-time air pollution monitoring system derived from a wireless sensor network (2019)
15. Marbán, Ó., Mariscal, G., Segovia, J.: A data mining and knowledge discovery process model.
In: Data Mining and Knowledge Discovery in Real Life Applications. InTech (2009)
16. WHO.: Air Quality Guidelines: Global Update 2005. Particulate Matter, Ozone, Nitrogen
Dioxide and Sulfur Dioxide. World Health Organization (2006)
17. Secretaría del Ambiente: índice Quiteño de la Calidad del Aire (2004)
18. del Ambiente, Secretaría: Informe de la Calidad del Aire del Distrito Metropolitano de Quito
2017. Tech. rep, Secretaría del Ambiente Alcaldía (2018)
Part III
Defense Engineering
Toward the Development of Surveillance
and Reconnaissance Capacity
in Ecuador: Geolocation System
for Ground Targets Based
on an Electro-Optical Sensor
Zurita C. Marco Antonio, Aguilar C. Wilbert Geovany

and Enríquez C. Victor Xavier
Abstract This paper presents the progress of a research work that seeks to deter-
mine the geographic coordinates of an object on the land of interest by applying
the Denavit–Hartemberg methodology to the surveillance and recognition system
developed in the CID FAE SEO-D1; this together with mathematical cartography
will provide information in support of military operations. The results shown in the
simulation present a good approximation of the geographic coordinates of interest. In
light of the above, it is suggested that this kind of study provides a low-cost surveil-
lance and reconnaissance alternative to existing advanced systems such as satellites
for countries that do not have this type of technology, as is the case of Ecuador.
Keywords Robot’s articulation · Kinematic chain · Denavit–Hartemberg

Algorithm · Mathematical cartography · Space projection · Gimbal
1 Introduction
The Armed Forces of Ecuador, as part of their doctrine, must guarantee Ecuadorian
society the defense of sovereignty and territorial integrity and, in addition, support
the integral security of the state [1]. This doctrine also establishes a relationship
with the socioeconomic development of the population, a relationship that within its
process and conception has had changes due to the appearance of organizations that
have caused problems within the country. Therefore, the Ecuadorian state has seen
the need to prioritize the border sector, specifically the northern border where illegal
activities have been developing, which constitute serious threats to the security of
the state [2].
Z. C. Marco Antonio (B) · A. C. Wilbert Geovany

University of the Armed Forces-ESPE, Sangolqui, Ecuador
e-mail: marco.zurich@hotmail.com
E. C. Victor Xavier
Ecuadorian Air Force Research and Development Center, Ambato, Ecuador
https://doi.org/10.1007/978-981-13-9155-2_11
124 Z. C. Marco Antonio et al.
On the other hand, the strategic planning of the defense is intended to guarantee
the security and defense of the state, conceiving from a comprehensive approach a
correct decision making in everything related to sovereignty and territorial integrity;
generating policies aimed at contributing to the safety and welfare of its citizens,
through surveillance and control activities such as land, river and air patrol operations.
Therefore, the Research and Development Center of the Ecuadorian Air Force
CID FAE is an entity dedicated to the development of projects oriented to security
and defense. Among the emblematic projects developed in this research center, there
is the surveillance and reconnaissance system. This system arises from the need to
capture images in real time for surveillance and reconnaissance applications [3].
The surveillance and reconnaissance system developed at the CID FAE consists
of a ground station, the gimbal system and a camera. These sub-systems allow the
operation of the gimbal in flight along with the camera and obtaining real-time
information of images and parameters such as height, gimbal geographic location,
pitch, yaw and roll degrees.
The known geographical location (latitude and longitude) corresponds to the gim-
bal mounted on an aircraft in flight, but we cannot say that these coordinates are those
corresponding to the surrounding objects on land. That is to say that we can visualize
images on the ground without knowing their geographical location.
To solve this problem, it is necessary to know the orientation and position of the
lens of the gimbal’s camera in the space; this way we would determine the direction
where the camera points and consequently know the location of surrounding objects
on land.
In robotics, it is a common need to know the orientation and location of the final
effector of a robot (a tool normally) from the knowledge of the values taken by the
articulations of the robot. This is called the direct kinematic problem [4] (Fig. 1).
There are different approaches to solve the direct kinematic problem, but for the
geolocation system developed the Denavit–Hartemberg methodology was used.
Denavit and Hartenberg proposed in 1955 a matrix method to systematically
establish a coordinate system which we will call Si linked to each link i of an articu-
lated chain, and the kinematic equations of the whole chain can then be determined
[5] which are the result of the multiplication of the homogeneous transformation
matrices.
The homogeneous transformation matrix is symbolized by i−1 Ai and represents
the relative position and orientation of the system Si with respect to the system Si−1 .
That is to say that 0 A1 describes the position and orientation of the solidary reference
system to the first link with respect to the solidary reference system to the base.
Fig. 1 Direct kinematic

problem. Source Robotics
fundamentals [4]
Toward the Development of Surveillance and Reconnaissance … 125
i−1
Ai = T(z, θi )T(0, 0, di )T(ai , 0, 0)T(x, αi ) (1)
⎡ ⎤⎡ ⎤⎡ ⎤⎡ ⎤
cos(θi ) − sin(θi ) 0 0 1 0 0 0 100 ai 1 0 0 0
⎢ sin(θi ) cos(θi ) 0 0⎥ ⎢0 1 0 0⎥ ⎢0 1 0 0⎥ ⎢ 0 cos(αi ) − sin(αi ) 0⎥
i−1
Ai = ⎢
⎣ 0
⎥⎢ ⎥⎢ ⎥⎢ ⎥ (2)
0 1 0 ⎦⎣ 0 0 1 di ⎦⎣ 0 0 1 0 ⎦⎣ 0 sin(αi ) cos(αi ) 0⎦
0 0 0 1 0 0 0 1 000 1 0 0 0 1
To represent the position and orientation of the solidary reference system to the final
effector S6 with respect to the solidary reference system to the base S0 , it will only
be necessary to multiply the homogeneous transformation matrices of all successive
links of the robot.
⎡ ⎤
nx ox ax px
⎢ n oy ay py ⎥
T = 0 A6 = 0 A1 1 A2 2 A3 3 A4 4 A5 5 A6 = ⎢
⎣ nz
y ⎥ (3)
oz az pz ⎦
0 0 0 1
Since the Denavit–Hartemberg algorithm gives us the position and orientation of the
final effector in values relatives of the plane X, Y, Z, it is necessary the application
of mathematical cartography for the representation of values of latitude (φ) and
longitude (λ) of the globe to their respective values passed to a plane and vice versa.
Cartography is the science that deals with the drawing of geographical charts.
Geographical charts, also called maps, are representations on a plane of all or part of
the earth’s surface. In order to make this representation, a certain projection of the
aforementioned reference land surface on a plane will be used [6].
To obtain the cartographic representations of a terrestrial zone, it is necessary to
know the formulas as that relate the values in the plane with the values of latitude
and longitude.
x = f 1 (φ, λ) (4)
y = f 2 (φ, λ) (5)
To these, formulas must exist a biunivocal correspondence for its inverse case.
φ = F1 (x, y) (6)
λ = F2 (x, y) (7)
For the calculation of these values, we used the (UTM) Universal Transversal Mer-
cator representation, because it is the most widely used and has less deformations
in its representation to the plane. Whose formulas and calculations are detailed in
“Bolletino de Geodesia e Science Affini” number 1 presented by Alberto Cotticia
and Lucia-no Surace (Fig. 2).
Fig. 2 UTM zones in a plane. Source Antonio R. Franco [10]
Fig. 3 UTM zones for Ecuador. Source Geo-ingeniería [11]

Ecuador is inside four zones (15, 16, 17 and 18), so the formulas applied will have
this condition (Fig. 3).
2 Design of the Geolocation System
The geolocation system is composed of elements that have the electro-optical sensor
such as the altimeter and inertial measurement unit (IMU), the gimbal, the ground
station with its HMI screen and algorithms implemented by software. In this paper,
we will focus on the algorithms to be implemented by software.
First, it is necessary to establish the characteristics of the gimbal on which the
Denavit–Hartenberg algorithm will be applied.
Since the gimbal is mounted on an aircraft, this will have variations in height,
pitch, yaw and roll product of the maneuvers executed in flight; these four variables
will be considered as values equivalent to those taken by the joints of a robot (four
degrees of freedom). These four degrees of freedom are added to two degrees of
freedom corresponding to two servomotors that integrate the gimbal (Fig. 4).
With these conditions, we proceed to put all systems associated with each joint
in accordance with Denavit–Hartemberg (Table 1).
Fig. 4 Gimbal characteristics. Source Author

Table 1 Systems associated with each articulation

S0 (Altimeter) S1 (Yaw) S2 (Pitch)
S3 (Roll) S4 (Servomotor 1) S5 (Servomotor 2)
S6 (Camera lens)
Table 2 Denavit–Hartemberg Joins θzi−1 dzi−1 axi αxi

parameter table
0A 1 0 h1 0 0
1
1A
2 2 q1 0 0 −90
2A
3 3 q2 − 90 0 0 −90
3A
4 4 q3 − 90 0 −d0 −90
4A
5 5 q4 − 90 −d1 0 −90
5A
6 6 q5 0 +d2 0
Once the systems associated with the joints have been established, we can obtain
the values for the homogeneous transformation matrices (Table 2).
T = 0 A6
⎡ ⎤⎡ ⎤⎡ ⎤
10 0 0 cos(q1 ) 0 − sin(q1 ) 0 cos(q2 − 90) 0 − sin(q2 − 90) 0
⎢0 1 0 ⎥ ⎢ 0⎥ ⎢ 0⎥
=⎣⎢ 0 ⎥⎢ sin(q1 ) 0 cos(q1 ) ⎥⎢ sin(q2 − 90) 0 cos(q2 − 90) ⎥ (8)
00 1 h 1 ⎦⎣ 0 −1 0 0 ⎦⎣ 0 −1 0 0⎦
00 0 1 0 0 0 1 0 0 0 1
⎡ ⎤⎡ ⎤
cos(q3 − 90) 0 − sin(q3 − 90) −d0 cos(q3 − 90) cos(q4 − 90) 0 − sin(q4 − 90) 0
⎢ ⎥⎢ ⎥
⎢ sin(q3 − 90) 0 cos(q3 − 90) −d0 sin(q3 − 90) ⎥⎢ sin(q4 − 90) 0 cos(q4 − 90) 0 ⎥
⎢ ⎥⎢ ⎥
⎣ 0 −1 0 0 ⎦⎣ 0 −1 0 −d1 ⎦
0 0 0 1 0 0 0 1
⎡ ⎤
cos(q5 ) − sin(q5 ) 0 d2 cos(q5 )
⎢ ⎥
⎢ sin(q5 ) cos(q5 ) 0 d2 sin(q5 ) ⎥
⎢ ⎥
⎣ 0 0 1 0 ⎦
0 0 0 1
The T matrix will allow obtaining the orientation and position of the camera lens
S6 with respect to the S0 system.
In this way, a straight line in the x-axis (camera lens) of the system S6 can be
projected into space using two points that are part of the x-axis (camera lens). The
equation of the line in space with the continuous form is given by [8]:
x − x0 y − y0 z − z0
= = (9)
v1 v2 v3
v1 = x1 − x0 (10)
v2 = y1 − y0 (11)
v3 = z 1 − z 0 (12)
From this projected line, the x and y values of the plane z = 0 are determined.
Which we will call x0utm and y0utm , these are the differential values in X and Y with
reference to the origin system S0 (Fig. 5).
Fig. 5 UTM target points.

Source Author
As already mentioned, the surveillance and reconnaissance system already provide

the geographic coordinates of the gimbal in real time. Then, it will be necessary to
pass these coordinates to UTM values through the formulas presented in “Bolletino
de Geodesia e Science Affini” to obtain the xutm and xutm values [12].
xtarget = xutm + x0utm (13)
ytarget = yutm + y0utm (14)
If it is required to obtain in geographic coordinates, it will be necessary to carry

out the conversion of UTM coordinates to geographic coordinates. These values will
be appropriate if the coordinates of both the gimbal and the ground target are within
the same UTM zone.
3 Simulation and Test
In order to carry out the simulations, the conditions in a certain period of time in
which the gimbal is in flight are established. For this, it is imposed that the gimbal
is up to 1000 m of height, in the coordinates (1.285309, −78.834736) and whose
orientation is the result of the variation of q4 and q5 in 45°.
With these parameters, the orientation given by the transformation T matrix will
be as follows (Fig. 6).
The values of the desire parameters are obtained by programing in MATLAB,
and the formulas are presented or mentioned in this paper (Table 3).
In order to appreciate the results, Google maps [9] are used, where the gimbals
and target will be visualized (Fig. 7).
Fig. 6 UTM target points simulation. Source Author

Table 3 Parameter results Parameter Value

table
Xutm 7.409466249039110e + 05
Yutm 1.421692420845669e + 05
xoutm 7.070360705084230e + 02
youtm 7.070360705084230e + 02
xtarget 7.416536609744194e + 05
ytarget 1.428762781550754e + 05
Target coordinates (1.291695931165799, −
78.828379873412587)
Fig. 7 UTM target points Google maps. Source Google maps [9]
As we can see, the geolocation system of targets on land shows a good approxi-
mation of the geographical coordinates of the target pointed by the camera.
In order hand to test the system, preliminary experiments were carried out in
controlled conditions and in locations with heights of no more than 13 m (Fig. 8).
Fig. 8 Geolocation test. Source Author
4 Discussion
This paper presents the results of the simulation of the designed geolocalization
system. This is a tool that can give information in real time of the geographic coordi-
nates of the surroundings, specifically of the objects visualized by the camera of the
gimbal. This information can be very useful as an input for the planning of military
ground operations.
The results of the preliminary tests of Table 4 show good results with a tolerable
error; this error was product in greater proportion by the sensitivity that presented
the servomotors of the gimbal which by their physical characteristics and the time
of operation presented a resolution of 1.5° if this resolution improved the error will
consequently reduce.
It is important to emphasize that the system has the limitation of the UTM zone,
since the formulas of the transformation of geographic coordinates to UTM and vice
versa use as input to the central coordinate of the corresponding UTM zone, which
is different for each zone.
With the development of the geolocation system, a boost will be given to the
development of operational capabilities in the Air Force by means of new tech-
nologies. The use of this technology will be attached to the nation’s security and
Table 4 Preliminary experiments table

Test height (m) Objective range Error x Error y Range variation
(m) (m) (m) (m)
1 5 22.5 1.4468 0.88633 1.6967
2 5 37.4 3.5672 4.3113 5.5958
3 13 37.3 4.7877 0.55752 4.8201
4 13 43.2 4.6758 1.2211 4.8326
5 10 46.6 1.7792 2.7675 3.2901
6 10 1.1 0.3338 0.2215 0.4007
defense activities. In addition, this technology can be extrapolated for use in civilian
applications.
References
1. COMACO: ModeloEducativo de las Fuerzas Armadas del Ecuador. Dirección Ge-neral de

Educación y DoctrinaMilitar (2016)
2. Gómez López, A.O.: Políticaspúblicas de seguridad de Colombia y de Ecuador: una visión-
comparada. Fronteras: rupturas y convergencias, 63–75 (2013)
3. Rafael, M.O.: Informe De Avance De Componentes Del Avión Uav - Sistema Electrooptico
Seo D1. Centro de Investigación y Desarrollo de la FAE, Ambato (2014)
4. Barrientos, A.: Fundamentos de robótica. McGraw-Hill, New York (1997)
5. Baturone, A.O.: ROBOTICA manipuladores y robots móviles. Mar-combo, Barcelona (2001)
6. Gamboa, J.M.: Fundamentos para CartografíaNáutica. JM ediciones, Cadiz (2009)
7. Maass, S.F., Pérez, M.E.: Principio Básicos de Cartografía y Cartografía au-tomatizada. Uni-
versidad Autónoma del Estado de México, Tolupa (2003)
8. Kindle, J.H.: Teoria y Problemas de GeometriaAnlítica Plana y del Espacio. McGraw-Hill,
México (1987)
9. Google. Google maps (Internet)
10. Barcelona, P.O.:. Astronomíaesférica y mecánica celeste (noviembre de 2018)
11. Geo-ingeniería, R.P.: Coordenadas UTM en google earth (9 de Diciembre de 2018)
12. Cotticia, A., Surace, L.: Bolletino Di Geodesia E Science Affini
Fuzzy Logic for Speed Control in Object
Tracking Inside a Restricted Area Using
a Drone
Richard Navas Jácome, Harley Lovato Huertas, Patricia Constante Procel

and Andrés Gordón Garcés
Abstract This article presents the autonomous speed and positioning control for a
drone using fuzzy logic to track a target within a restricted area. Classical controllers
present a problem, as in general they only support one input and one output and a
system model is always required. For this application, it is necessary to analyze two
inputs, the position in “x” and the position in “y” of an object that will be recognized
by the drone through artificial vision. The goal is to control the speed at which the
drone moves according to the position of the object detected by the machine vision
within a restricted area, resulting in a faster or slower movement that will improve
the tracking of a moving target by delivering real-time object monitoring information
to the user in order to take some action based on this information.
Keywords Fuzzy control · Drone · Artificial intelligence
1 Introduction
Fuzzy logic is a mathematical language that represents non-quantifiable values. In

classical logic, an element belongs or does not belong to a set, whereas in fuzzy
logic, each element has degrees of membership [1]. This type of control is typical of
Artificial Intelligence (AI), where attempts are made to emulate human behavior. The
design of the fuzzy controller uses the knowledge and experience of the designer;
one of the advantages of this control is that it does not require a mathematical model
of the process to be controlled [2].
In recent years, unmanned aerial vehicles, and more specifically drones, have been
used for the development of multiple and varied applications, either because of their
versatility, relative ease of flight control, or because of the ability to obtain informa-
R. N. Jácome · H. L. Huertas · P. C. Procel (B) · A. G. Garcés

Universidad de las Fuerzas Armadas ESPE, Sangolquí, Ecuador
e-mail: pnconstante@espe.edu.ec
A. G. Garcés
e-mail: amgordon@espe.edu.ec

https://doi.org/10.1007/978-981-13-9155-2_12
136 R. N. Jácome et al.
tion from the environment through cameras, sensors, etc. [3]. At Zayed University,
they analyzed the potential applications of drones in environmental monitoring and
urban space management, as well as their potential risks [4]. In the USA, researchers
are looking for ways to supply basic necessities [5] and improve logistics in the
event of natural disasters [6]. In agriculture, combining the benefits of working with
unmanned aerial vehicles and introducing concepts of AI and the Internet of Things
(IoT), so-called precision farming is sought after [7]. This shows that drones are an
ideal tool to provide solutions to multiple problems in today’s world. An application
for security and surveillance is the tracking of targets by means of several robots
with sensors on board [8], and this evidences that the use of robots that provide data
in real time is an advantage in the monitoring of restricted areas. Early detection of
an intruder invading private or restricted property ensures that defense mechanisms
take more efficient action to safeguard the integrity of the complex and the security
of elements of force, occupying a network of technologies that help maintain order.
This research proposes a fuzzy control system to improve the positioning of a
drone in a satisfactory manner. There are classic control systems well known as
the PID which requires precise mathematical models, used by Korean engineers
to develop an intelligent tracking drone [9]. Hybrid controllers such as fuzzy PID
control have also been used to avoid obstacles in enclosed environments, proposed
by the Universidad de La Frontera (Chile) [10]. However, fuzzy logic control was
chosen for its favorable performance in systems where it is intended to position
itself in a point of interest, a field in which this research is focused; for example,
the autonomous landing system of an unmanned aerial vehicle based on artificial
vision, which, through the integrated camera, detects the landing point and positions
it through a fuzzy control [11]. The implementation of the fuzzy control was carried
out using a Python programming algorithm without the use of libraries. To form the
fuzzy logic sets, the work of researchers at the University of Surabaya [12] was taken
into account, showing how they built the sets for each of the axes of a drone.
2 Definition of the Control System
Fuzzy logic allows ordinary language to be used as a variable for a computer. Knowl-
edge is interpreted as a collection of fuzzy constraints on a group of variables. This
logic is widely used in decision support systems. Fuzzy logic allows a computer to
reason in linguistic terms and rules in a manner similar to that of human beings [13].
A fundamental part of the fuzzy logic is the selected membership functions. A
fuzzy logic set is an ordered pair of an “x” element belonging to a U universe and its
degree of membership, as shown at (Fig. 1). The functions of membership, which give
the degree of membership, can be defined in a functional way, which has different
forms such as trapezoidal, triangular, bell, etc. [14].
The type of controller chosen was the Mamdani [15], its architecture, which has
the following structure [16]:
Fuzzy Logic for Speed Control in Object … 137
Fig. 1 Fuzzy set
• Determination of membership functions that will perform the fuzzy scaling.

• Establishment of rules based on conditional proposals.
• Selection of belonging functions for deburring and the method for finding the out-
put value, which is the centroid of the geometric figure formed by the assemblies;
it is used (1).

μ(x) · x μ(x) · x
(1)
μ(x) μ(x)
3 Development
3.1 Membership Functions for the Fuzzy Process
To form fuzzy sets and membership functions, it is necessary to know what input
data and the range of values they can take. The system detects by artificial vision
the position of the object’s center of interest so the data entry will be the location of
pixels in the X- and Y-axes of the image; and the size in pixels of the input image is
640 in the X-axis and 480 in the Y-axis as shown in (Fig. 2a).
(a) image size (b) Reference plane
Fig. 2 Image acquisition

(a) Fuzzy sets on the reference plane (b) Linguistic variables
Fig. 3 Reference image
The purpose of the application is to move the drone to the location of the object
in question; therefore, it must be located in the center of the input image, being this
the set point. Thus, we have the image divided into four quadrants, sectioning the
X-axis into two parts of 320 pixels each and the Y-axis into 240 pixels as shown in
(Fig. 2b).
Within the image, three regions are defined to determine in which area the object
is located. “Close” corresponds to the area near the center of the image, i.e., when
the distance between the drone and the object is relatively small. “Normal” refers
to the middle zone between the center of the image and the peripheral zone of the
image, and in this region, the distance will be intermediate. And “Far” corresponds
to the farthest area located at the edges of the image, and at these points, the drone is
at a long distance from the target, as shown at (Fig. 3a). The set point, located in the
center of the image, divides the X- and Y-axes into two equal parts. The X-axis has
its components right and left, and the Y-axis has its components above and down, as
shown at (Fig. 3b).
Once the input parameters and the way in which the data obtained is sorted are
defined, the fuzzy sets are formed. As can be seen in (Fig. 4), both the X- and Y-axes
are similar, so the number of sets will be the same, with a variation in the range of
values. On the X-axis, the range of values is from −320 to 320 pixels forming five
sets: “FarL,” region far to the left side, “NormalL” intermediate region to the left
side, “Near” region near the center of the image, “NormalR” intermediate region
to the right side and “FarR” distant region to the right side, as shown in (Fig. 4a).
Fig. 4 a Fuzzy sets on the “x” axis. b Fuzzy sets on the “y” axis
Similarly, on the Y-axis, the range of values ranges from −240 to 240 pixels, making
five sets: “FarL,” far region on the left side, “NormalL” intermediate region on the
left side, “Near” region near the centre of the image, “NormalR” intermediate region
on the right side and “FarR” far region on the right side, as shown in (Fig. 4b).
Once the sets have been formed, they are mathematically defined for later coding in
programming language. The mathematical equation of the lines forming the polygon
of each set and the range of values they are in are described in Tables 1 and 2.
3.2 Settings the Rules
With the fuzzy process data dissemination, the fuzzy controller rules for speed vari-
ation are developed, where the considerations are as follows: For “far,” the drone
will be at its maximum “fast” speed, and as it approaches the object it will pass at
“normal” distance so that its speed will be “medium” and when it reaches a location
near it “near,” its speed will be “slow” with respect to the center point. The rules are
shown in Table 3.
Table 1 Equation for X pixel variation

Name Equation

Near x
N ear = 80 + 1, −80 < x < 0
−x
80 + 1, 0 < x < 80
⎧
Normal right ⎪
⎪ x
− 0.5, 40 < x < 120
⎨ 80
N or mal R = 1, 120 < x < 200
⎪
⎪
⎩ −x
+ 6, 200 < x < 240
40
Normal left N or mal L =

⎧
⎪ −x
⎨ 80 − 0.5, −40 < x ← 120
⎪
1, −120 < x ← 200
⎪
⎪
⎩ x + 6, −200 < x ← 240
40

Far right x
− 5, 200 < x < 240
Far R = 40
1, 240 < x < 320

Far left −x
− 5, −200 < x ← 240
Far L = 40
1. − 240 < x ← 320
Table 2 Equation for Y pixel variation

Name Equation

Near y
N ear = 60 + 1. − 60 < x < 0
−y
60 + 1.0 < x < 60
⎧
Normal above ⎪
⎪ y
− 0.5.30 < x < 90
⎨ 60
N or mal R = 1.90 < x < 150
⎪
⎪
⎩ −y
+ 6.150 < x < 180
30
Normal down N or mal L =

⎧
⎪ −y
⎨ 60 − 0.5, −30 < x ← 90
⎪
1. − 90 < x ← 150
⎪
⎪
⎩ y
+ 6. − 150 < x ← 180
30

Far above y
− 5.150 < x < 180
Far R = 30
1, 180 < x < 240

Far down −y
Far L = 30 − 5, −150 < x ← 180
1, −180 < x ← 240
Table 3 FAM matrix Axis X

Far L Normal Near Normal Far R
L R
Axis Y Far L F F F F F
Normal F M M M F
L
Near F M S M F
Normal F M M M F
D
Far D F F F F F
The speed parameters: F = fast, M = medium, S = slow
3.3 Membership Functions for Defuzzification
For the deburring of the processed data, the membership functions for the fuzzy
control are formed where the speed given for the drone is from 0 m/s to a maximum
of 15 m/s, considering the three parameters for “slow,” “medium” and “quick,” as
shown in (Fig. 5). Finally, the previously formed sets are mathematically defined for
later implementation in programming language (Table 4).
Fig. 5 Fuzzy output sets
Table 4 For the output speed Name Equation

Low speed
Slow = − u7X + 17 , 0 < V < 7

Medium speed 5 ∗ u X + 3.3 < V < 8
Medium =
−5 ∗ u X + 13.8 < V < 13

High speed 4 ∗ u X + 8.8 < V < 12
Quick =
u X = 112 < V < 20
4 Implementation
The implementation of the code was done in Python, since in this language it is
possible to find the functions that allow control the air vehicle. The programming of
the image acquisition, addressing and fuzzy controller for the speed was based on
the structure shown in (Fig. 6).
The drone used for the implementation is a 3DR SOLO (Fig. 7) with a PIX-
HAWK 2.1 controller, to encode the board is used Python programming language
and as Linux development environment. The controller board has its own libraries to
send instructions to the unmanned aerial vehicle and position it at the point of inter-
est; these libraries are used to send the fuzzy controller result to the drone’s flight
parameters. The communication protocol used is mavlink wifi, which establishes a
stable connection with the PC control medium.
Fig. 6 Structure of the system

Fig. 7 Drone 3DR Solo
(a) RGB Image (b) Object detected
Fig. 8 Image processing
For image processing, OpenCV was used to develop a BGR file with the parame-
ters to highlight the blue object between the ranges (B = 100, G = 65, R = 75) and
(B = 130, G = 255, R = 255) placed in a filter layer as in (Fig. 8). In this case, the
blue object was used to represent the unauthorized object inside the restricted area,
but it could be customized in order to detect different objects, such as vehicles or
people, using advanced artificial vision algorithms which are not presented in this
paper.
5 Analysis of Results
5.1 Comparison Between Matlab Results and Implemented

Code
Then, it is analyzed whether the result of the fuzzy control is consistent with the
requirements, i.e., whether the velocity value grows and grows according to the
position of the detected object. The speed values are then compared with the values
obtained through Matlab to check if the implemented code responds correctly.
Table 5 Comparison Test Speed (m/s)

between Matlab results and
implemented code Matlab Code
1 2.333 0.04375
2 6.88 5.9885
3 8 8
4 12.2 13.49
5 14.25 14.25
Test 1: The object is in one position (1.0) and the speed is 0.04375, tending to
zero. Therefore, the result is acceptable as in this position the drone should remain
static.
Test 2: The object is in a position (−69.43) one and the speed is 5.9885. The
two coordinates are at extreme values of the near and normal sets, so it follows that
the speed must be within the range of slow and average speeds. The speed value is
acceptable.
Test 3: The object is in a position (−123.75) one and the speed is 8. The two
coordinates are in values belonging to the normal position sets, i.e., the speed of the
drone is average. The result is acceptable for this test.
Test 4: The object is in one position (−276,166) one and the speed is 13.49. The
values of the X, Y coordinates are between the far and far ranges of normal and far,
and speed is expected to be between the mid and fast ranges. The speed delivered in
this test meets the requirements.
Test 5: The object is in one position (−290.199) one and the speed is 14.25. Both
coordinate values are in the far range, and the speed is the maximum the algorithm
can provide. The result is optimal in this case.
Table 5 presents the results where it is compared with the Matlab software where
a similarity between the values obtained is evident without this altering the func-
tionality of the implemented controller. As you can see, the results vary slightly,
although the growth trend is similar. The probable cause is that Matlab uses a dif-
ferent method to debrister the final result than the one used for code implementation
(Discrete centroid method).
5.2 Final Performance Tests
Several tests were carried out, three of which are presented at (Fig. 9), which shows
an increase in speed as a function of the distance from the drone to the object. The
results are shown in Table 6.
Fig. 9 Implementation and testing
Table 6 Final result in each Function Position Speed (m/s) Parameters

case
A Near (−9, 4) 1 Slow
B Normal (145, 50) 8 Medium
C Far (220, 50) 12.75 Quick
6 Conclusions
A fuzzy controller with two inputs and one output (MISO) has been easily imple-
mented on a drone, which provides speed data based on the distance of the object
that has been recognized. Real-time images of 640 × 480 pixels have been acquired,
resulting in triangular-shaped membership functions defined for the distance in “x”
and “y” to be able to associate when an object intended to track is near or far.
Fuzzy logic control was implemented in Python programming language because
it is a high-level hybrid language, and it is the development native language of the
drone. The principal advantage of this fuzzy logic controller is that it could be applied
to any type of UAV because the system model of the UAV is not required for control,
and it could be developed in any other programming languages depending on the
UAV manufacturer.
Importance of the presented research work is that the speed control and positioning
system can be used and developed within applications mostly in the area of security
and defense, such as, for example, non-identified people and objects tracking inside
secure areas.
Acknowledgements Thanks to the University of the Armed Forces ESPE for the support provided
to this research work.
References
1. Ramírez, N.V., Laguna Estrada, M.: La lógica borrosa: conjuntos borrosos, razonamiento aprox-
imado y control borroso (2012). http://pistaseducativas.itc.mx/wp-content/uploads/2013/01/6-
RAMIREZ-PE-100–55-65.pdf. Accessed 20 November 2017
2. Camastra, F., Ciaramella, A., Giovannelli, V., Lener, M.: A fuzzy decision system for geneti-
cally modified plant environmental risk assessment using Mamdani inference, ScienceDirect,
1710–1716 (2015)
3. Olivares Mendez, M., Mejias, L.: See-and-avoid quadcopter using fuzzy control. In: IEEE
World Congress on Computational Intelligence, p. 1 (2012)
4. Gallacher, D.: Drone applications for environmental management in, Science Target, p. 1 (2016)
5. Scott, J.E., Scott, C.H.: Drone delivery models for healthcare. In: 50th Hawaii International
Conference on System Sciences, p. 1 (2017)
6. Restas, A.: Drone applications for supporting disaster, SciRes, p. 1 (2015)
7. Saha, A.K., Saha, J., Ray, R.. Sircar, S.: IOT-based drone for improvement of crop quality.
IEEE, p. 1 (2018)
8. Hausman, K., Müller, J., Hariharan, A., Ayanian, N., Sukhatme, G.S.: Cooperative control for
target tracking with onboard sensing. Exp. Robot. 879–892 (2015)
9. Yang, W.-S., Chun, M.-H., Jang, G.-W.: A study on smart drone using quadcopter and object
tracking techniques. IEEE, p. 1 (2018)
10. Gatica, N., Muñoz, C., Sellado, P.: Real fuzzy PID control of the UAV AR.Drone 2.0 for, IEEE
(2017)
11. Olivares Mendez, M., Kannan, S., Voos, H.: Vision based fuzzy control autonomous landing
with UAVs: From, IEEE, p. 1 (2015)
12. Indrawati, V., Prayitno, A., Utomo, G.: Comparison of two fuzzy logic controller schemes.
IEEE (2015)
13. González Morcillo, C.: Lógica Difusa. técnicas de Softcomputing. http://www.esi.uclm.es/
www/cglez/downloads/docencia/2011_Softcomputing/LogicaDifusa.pdf
14. Santos, M., Suescun, E.M.: Aplicación de la lógica difusa en el ambito de las energías renovables
(2012). https://dialnet.unirioja.es/descarga/articulo/5085360
15. Ören, A., Koçyığıt, Y.: Landing sequencing modelling with fuzzy logic: opportunistic approach
for unmanned aerial systems. In: 2016 International Conference on Unmanned Aircraft Systems
(ICUAS), pp. 943–948, (2016)
16. Lovato, A.V., Oliveira, J.C.M.: Airplane level changes using fuzzy control. In: International
Conference on Fuzzy Systems (2010)
Part IV
Health Informatics in Military Applications
Micro-controlled EOG Device for Track
and Control Military Applications
Nayana L. M. Viana, José Ailton L. Barbosa Junior

and Francisco A. Brito-Filho
Abstract This paper presents a system with tracking and remote control capabilities
for military applications. An electrooculogram and a micro-controlled system with
wireless communication were developed as controller based on the eye movement.
The system can be used as remote control for military purposes and also for soldier
monitoring. A proof of concept to act and be tested as mouse pointer was implemented
with reduced circuitry and PIC microcontroller in order to achieve low cost and
low profile platform. Experimental results are shown to validate a prototype of the
proposed system.
Keywords Electrooculogram · Tracking · Remote control
1 Introduction
Biopotentials are signals that result from electrophysiological activities that occur
inside the cells. Those activities can represent many physical behavior that human
being does along the day, for example, to walk, to pick an object and to blink the
eyes [1, 2].
Those biopotentials can be measured and applied to track some specific behavior
or to treat diseases and to control objects [2].
N. L. M. Viana · J. A. L. Barbosa Junior · F. A. Brito-Filho (B)

Federal University of Semiarid Region—UFERSA, Caraubas, RN 59780-000, Brazil
e-mail: francisco.brito@ufersa.edu.br
N. L. M. Viana
e-mail: nayanaleticia@ufrn.edu.br
J. A. L. Barbosa Junior
e-mail: ailton.junior@ufersa.edu.br
N. L. M. Viana
Federal University of Rio Grande do Norte—UFRN, Natal, RN 59078-900, Brazil
J. A. L. Barbosa Junior
Federal University of Ceara—UFC, Fortaleza, CE 60020-181, Brazil

https://doi.org/10.1007/978-981-13-9155-2_13
150 N. L. M. Viana et al.
There are four biosignals covering all activities performed by humans, that are
electrocardiogram (ECG), electroencephalogram (EEG), electromyography (EMG)
and electrooculogram (EOG), which are the potential arising from muscle contrac-
tions of the heart, neural activities (or brain activities), activity from fiber or group of
muscle fibers and difference of potential between the cornea and the retina, respec-
tively [1–4].
Rising advances in biopotential acquisition, processing and applications are allow-
ing low profile devices, with more battery life and lower cost, motivating new markets
to use them like consumer and defense [4, 5]. Acquisition and control of EOG sig-
nals appear in this scenario as an alternative to track soldier behavior, to monitor his
health from remote places or to use as a remote control signal for military devices
[6–11].
This work presents a device that can be applied for military purposes both in track
or monitor field as in control applications. As a proof of concept of device idea,
a mouse pointer micro-controlled by eye movement is developed and validated by
experimental tests.
2 Proposed System
The EOG signals are pulses that vary proportional to the degree of eye movement.
As higher the degree of eye movement, higher is the EOG signal amplitude [1, 2].
Figure 1 shows typical EOG signals.
To validate the proposed system idea, to track eye movement and to control objects
by the same biopotential, a mouse pointer device micro-controlled by the EOG signal
was implemented. The EOG system block diagram is shown in Fig. 2.
Fig. 1 Typical EOG signals

Micro-controlled EOG Device for Track and Control … 151
Fig. 2 EOG system block diagram
The implemented system can be divided into two subsystems: the acquisition
system, that contains the electrodes, instrumentation amplifier and filters; and the
processing and control system that is performed by the PIC18F microcontroller. Also,
a wireless transmission system can be added in order to send the control signals to
remote devices. Below are presented those subsystems in more detail.
2.1 Acquisition System
The acquisition system is basically composed of electrodes, instrumentation amplifier

and filters.
AgCl Electrodes and Conductive Gel. The biopotential acquisition is done by
means of electrodes that perform the transduction from the ionic currents in electrical
signals [1–5]. To capture the EOG signal is used silver chloride surface electrodes
(AgCl) and to optimize capture of that signal is used a conductive gel. The latter is
used to increase the conductivity between the electrode and the patient’s skin.
The electrodes are placed as shown in Fig. 3. They are placed below and above
the eye to capture vertical movement and at left and at right of the eye to capture
horizontal movement. The electrode placed on the forehead is the reference electrode
and must be connected to GND (ground) circuit.
Instrumentation Amplifier. As seen before, the EOG signals have low magnitudes.
So, they need to be amplified in order to be analyzed by measurement instruments
and used as input signal to the PIC microcontroller. Is necessary that their amplitudes
are larger and therefore the signal must pass through amplifiers.
The instrumentation amplifier is the most indicated when need to get a high
amplification with minimum noise. Since the EOG biosignal magnitude is on the
order of low-frequency noise, the amplifier needs a high Common Mode Rejection
Ratio (CMRR), high gain, low drift and high input impedance that is the case of the
instrumentation amplifier shown in Fig. 4.
In order to get a low profile device was chosen the use of two integrated circuits
(IC) to implement the two instrumentation amplifiers (one for acquisition of the
horizontal eye movement and other for the vertical eye movement), since each IC
has four operational amplifiers. Three operational amplifiers were used to implement
the instrumentation amplifier and the other was used to implement the output filter
that will be shown in next subsection. Each IC has inputs for signals that come from
electrodes and output to PIC microcontroller.
Fig. 3 Electrodes
positioning
Fig. 4 Instrumentation Rf3

amplifier schematic
Cf2
EL+ Rf2
Rf3
Vout
Cf2 Rf3
EL-
Rf3 Rf2
Filter. In order to get a better pulse acquired from eye movement and suitable to
be read in PIC inputs, an active low-pass filter is implemented at the output of the
instrumentation amplifier in order to eliminate the high frequency noise. The filter
schematic can be seen in Fig. 5.
At filter output is a signal ready to be read at PIC microcontroller input and
processed.
Cf2
Rf1 Rf2
Vin
Cf1 Vout
Rf3
Rf2
Fig. 5 Filter schematic
Start
A/D > 512? A/D > 512?

N N
Y Y
X-- X++ Y-- Y++
Fig. 6 Microcontroller algorithm flowchart
2.2 Processing and Control System
The processing and control system can have a transmitter after filter output, that
wirelessly transmits the EOG signals to a receiver coupled to the PIC microcontroller,
in order to process and control remote devices or to monitor the EOG signals. The use
of radiofrequency signals can allow remote control applications for long distances
for example in battlefield context.
The microcontroller used was the PIC18F. Its function is to acquire the analog
signal from the EOG and convert it into a digital signal using one of its analog-
to-digital converter (A/D) inputs. Then the signal is processed using an algorithm
developed exclusively for this application which its flowchart can be seen in Fig. 6.
Two PIC A/D inputs are used to acquire EOG signals represented in the flowchart
as X and Y inputs, or horizontal and vertical inputs, respectively. Since PIC A/D is
10-bit resolution so the microcontroller can read 1024 different levels representing
EOG voltages from 0 to 5 V. Before EOG signal enters in PIC A/D input, it passes
through a 2.5 V offset in order to test the average value 512. As can be seen in Fig. 6,
the mouse movement is incremented or decremented in two dimensions if the inputs
are below or above of the offset.
The post-processed signals are used to control the device. Depending on the device
to be controlled, the signals can be sent wirelessly or by serial connection, stored to
be monitored in real time by another device.
3 Results
In order to validate the proposed system, an EOG micro-controlled mouse pointer

was developed and the system was divided into previously described two stages for
acquisition and for processing and control purposes. As part of the next tests, wireless
transmitter-receiver interface is been provided.
Figures 7 and 8 can be seen two prototype boards for the two subsystems designed
for the system proof of concept. Since the boards were built independently from each
other, it can be used in stand-alone mode and can be stacked in order to get low profile.
In order to check the acquired signals, some volunteers were used to test the
system. Figures 9 and 10 show the measured EOG signals acquired, respectively,
from the horizontal and vertical eye movement.
Fig. 7 Prototype upper view

Fig. 8 Prototype side view
Fig. 9 Measured EOG horizontal signal

Fig. 10 Measured vertical EOG signal
In Fig. 9, the pulses higher than offset indicate that the volunteer eye movement
was to the left side and the pulses less than offset represent the eye movement to the
right side. In the same way, Fig. 10 represents the volunteer eye movement down
when pulses are higher than offset and up when pulses are less than offset which is
in agree with the fluxogram previously presented.
Transmission and reception prototypes in the same modular profile are under
construction.
4 Conclusions
This paper presented a system proposal for track and remote control based on the eye
movement and can be used for military purposes both monitor or control applications.
In order to validate the system device, a micro-controlled mouse pointer was
implemented and tested. The results showed the pulses acquired from the horizontal
and vertical movements and the prototype proved a low profile, as low as 5 × 3 cm,
and low cost platform to be used.
The prototype tested was effective attending the initial proposal, with the functions
of track eye movement for each direction: right, left, up and down, and also the blink,
that for military control applications could be a confirm signal.
Some improvements are been done as to adapt acquisition system to special glasses
without the necessity of disposable electrodes and cables, to be applied in a more
real battlefield context, and also, with receive-transmit capabilities.
References
1. Bronzino, J.D.: The Biomedical Engineering Handboook, 2nd edn. CRC Press, Boca Raton
(2000)
2. Malmivuo, J., Plonsey, R.: Bioelectromagnetism: Principles and Applications of Bioelectric
and Biomagnetic Fields. Oxford University Press, USA (1995)
3. Yazicioglu, R.F., Hoof, C.V., Puers, R.: Biopotential Readout Circuits for Portable Acquisition
Systems. Springer Science & Business Media (2009)
4. Tseng, Y., Ho, Y., Kao, S., Su, C.: A 0.09 µW low power front-end biopotential amplifier for
biosignal recording. In: IEEE Transactions on Biomedical Circuits and Systems, pp. 508–516.
IEEE (2012)
5. Ha, S., Kim, C., Chi, Y.M., Akinin, A., Maier, C., Ueno, A., Cauwenberghs, G.: Integrated cir-
cuits and electrode interfaces for noninvasive physiological monitoring. In: IEEE Transactions
on Biomedical Engineering, pp. 1522–1537. IEEE (2014)
6. Patil, N., Iyer, B.: Health monitoring and tracking system for soldiers using internet of
things (IoT). In: International Conference on Computing, Communication and Automation
(ICCCA2017), pp. 1347–1352, IEEE, Greater Noida, India (2017)
7. Bisio, I., Delfino, A., Lavagetto, F., Sciarrone, A.: Enabling IoT for in-home rehabilitation:
accelerometer signals classification methods for activity and movement recognition. IEEE
Internet Things J., 1–11. IEEE (2016)
8. Rossi, M., Rizzi, A., Lorenzelli, L., Brunell, D.: Remote rehabilitation monitoring with an IoT-
enabled embedded system for precise progress tracking. In: IEEE International Conference on
Electronics, Circuits and Systems (ICECS), pp. 384–387. IEEE, Monte Carlo, Monaco (2016)
9. Jiang, M., Gia, T.N., Anzanpour, A., Rahmani, A.-M., Westerlund, T., Salantera, S., Liljeberg,
P., Tenhunen, H.: IoT-based remote facial expression monitoring system with sEMG signal.
In: IEEE Sensors Applications Symposium (SAS). IEEE, Catania, Italy (2016)
10. Reyes, C.R.P., Vaca, H.P., Calderón, M.P., Montoya, L., Aguilar, W.G.: MilNova: an approach
to the IoT solution based on model-driven engineering for the military health monitoring. In:
CHILEAN Conference on Electrical, Electronics Engineering, Information and Communica-
tion Technologies (CHILECON). IEEE, Pucon, Chile (2017)
11. Tavakoli, M., Turicchia, L., Sarpeshkar, R.: An ultra-low-power pulse oximeter implemented
with an energy-efficient transimpedance amplifier. IEEE Trans. Biomed. Circuit. Syst. IEEE
(2009)
Part V
Leadership and e-Leadership
Multilevel Military Leadership Model:
Correlation Between the Levels
and Styles of Military Leadership Using
MLQ in the Ecuadorian Armed Forces
Celio Humberto Puga Narváez, Alex Fernando Jimenez Vélez,

Rafael Caballero Fernández and Osvaldo Fosado Téllez
Abstract This article presents the advances of a research focused on validating

a Multilevel Military Leadership Model, which relates: leadership levels, levels of
military strategy, personal and institutional competencies that leaders must develop,
and the leadership styles of the Full Range Leadership Model (FRLM). The model
applied in a first approximation to the Ecuadorian Armed Forces uses the Multifactor
Leadership Questionnaire (MLQ) to determine the predominance of each leadership
style within each level of leadership, as well as its influence with variables of per-
formance and satisfaction at work.
Keywords Multifactor Leadership Questionnaire (MLQ) · Full Range Leadership

Model (FRLM) · Military leadership · Multilevel leadership · Transformational
leadership
1 Introduction
The increasing complexity and evolution of organizational science means that lead-
ership research should be prepared for a multilevel study approach, supported mainly
by the progress in the analysis of its data. This analysis allows the complex field of
leadership to develop specific theoretical-methodological advances at each level [1],
going from conceiving a type of individual leadership in the 1980s, organizational
in the 1990s and multilevel since 2000 [2].
C. H. P. Narváez (B)
Programa de Doctorado de Economía y Empresa, Universidad de Málaga, Málaga, Spain
e-mail: celiopn@hotmail.com
A. F. J. Vélez
Centro de Investigación y Desarrollo FAE, Ambato, Ecuador
R. C. Fernández
Universidad de Málaga, Málaga, Spain
O. F. Téllez
Universidad Técnica de Manabí, Portoviejo, Ecuador
https://doi.org/10.1007/978-981-13-9155-2_14
162 C. H. P. Narváez et al.
The multilevel approach appears in 1989 under the name of stratified systems
theory, depending on the time and work complexity. Subsequently, Hunt (1991,
1996) adapts the theory to a vertical leadership perspective at the different levels of
the organization, called the multi-organizational leadership model.
This model establishes that there are critical tasks that must be carried out by
the leaders in an organization, according to their level, thus they will have tasks
that change its complexity based on their level. The complexity of these tasks can
be measured in several ways, one of them will be the time taken to accomplish
every single task, activities or projects developed by an individual or team. The tasks
assigned to the lower levels will have a shorter time of execution than the tasks
performed at higher levels.
That is why the multilevel extended model indicates that the variety of tasks com-
plexity leads to a division of three levels; where the lowest level of the organization
is called direct or production, the level in the middle is called organizational and the
highest level as systems or strategic, being these levels highly related to each other.
Leadership at the direct or production level involves administrative or operational
procedures and maintenance of individual or collective skills; the organizational
comprises the integration of organizational elements focused on the compliance of
the objectives and its mission statement; and the leadership of systems implies the
development of the mission statement, articulation of objectives, strategies, princi-
ples, policies and planning of the organizational systems [3, 4].
In this sense, Reeves-Ellington (2015) presents another classification of leader-
ship: junior, middle and higher. The junior level leadership is in charge of taking
care of the lower levels of the organization; the medium has a key role in the com-
pany integration, its broader social environment and personal climate have strong
links throughout the organizational community; and the superior level with the main
responsibility of controlling strategic execution through the supervision of human
and budgetary resources, achieving a community and holistic result [5].
1.1 Military Leadership Levels
According to the aforementioned, it can be determined that the theory of multilevel

leadership is accepted in the organizational field, this multilevel classification being
extrapolated to the military leadership, specifically to a vertical institution such as
the Armed Forces.
Under this organizational structure, military leadership comprises three levels:
tactical, operational and strategic leadership; division that is correlated based on the
classification of the three levels of military strategy, widely known in the Armed
Forces field.
The tactical leadership, which is the lower level in the military organization, seeks
to develop the ability in the execution of actions in order to achieve the purposes
proposed with the available resources. At this level, there is more certainty and less
complexity than at higher levels.
Multilevel Military Leadership Model: Correlation … 163
Although the time that the military officer remains at this level depends on the
acquired skills, generally spoken, it is made up of military personnel who start their
military career until around the 12 years of service, being this reference the limits,
and considering that the military profession could last for approximately 35 years;
logically, these times vary according to the institution and analyzed country.
In mid-level operational leadership, the main objectives and resources are adapted;
where a link between the complex decisions, guidelines of the strategic level and the
tasks fulfilled by the tactical level is established.
At this level, a large amount of useful information is analyzed to advise the
superior level; therefore, the people at this level must have a clear understanding of
the different systems of the organization, which allows them to decide their correct
interaction. The operational level is made up of military professionals whose service
time can generally be 12–24 years of service.
The strategic leadership is the highest level, where the organizational goals and
objectives are established, and it is unfolding in an uncertain, changing, unknown
and complex environment, because the strategic leader is the direct link to organiza-
tions outside the military environment, with different characteristics. Therefore, the
leader at this level has a knowledge focused on the national, regional and global per-
spective, with a capacity for convincing and negotiating, which allows him to make
successful decisions, since the implications will affect the direction and evolution of
the institution.
The level of strategic leadership is made up of the entire institutional high com-
mand, with approximately 24–35 years of service.
Hence, leaders are trained to finally achieve their full leadership at a higher level. It
is important to point out that during the process, they acquire a series of institutional
and personal competencies.
1.2 Institutional Competencies
Among the competencies that must be developed by leaders for the growth of their
leadership, there are three broad categories: technical, interpersonal and conceptual
competencies; at the tactical level, specific skills or technical knowledge is required,
which decrease as the level of leadership increases, in contrast to the interpersonal
skills that are essential at all levels, as well as the conceptual competencies that grow
as the level of leadership increases, allowing to determine important aspects of the
organization.
In the case of technical competencies, these cover specific knowledge regarding
the assigned technical aspects, they are required at all levels, but in a different way,
at the tactical level consists generally in using, operating or maintaining a team or
system; they focus on solving specific problems, as well as carrying out specific
tasks and missions. At the operational level, the personnel can lead to establishing a
network of systems, and at the strategic level, they are useful for decision-making, by
facilitating the synchronization of information with a system of systems, or to design
a well-articulated structure between forces or military units, as well as to work on

solving major problems that would affect the entire organization.
If we talk about interpersonal competencies, they refer to the capacities needed
to interact effectively with other people; these competencies also allow the leader
to have emotional intelligence, self-awareness, confidence and patience to foster an
environment of empowerment, being fundamental at all three levels (tactical, oper-
ational and strategic). The personnel at the tactical level can deal with interpersonal
relations under the framework of subordination; on the other hand, the personnel
at the strategic level have more lateral relationships, where collaboration, integra-
tion, persuasion, perception and negotiation skills may be necessary, but generated
through mutual trust and respect.
Finally, in the case of conceptual competencies, these include the ability to think,
mainly related to logical analysis and inference, which is why they are more necessary
at the strategic level than at the other levels. At this level, the strategic leaders get
information from external sources and internal, which allows them to understand
through visualization and to project adequate planning.
The ability to search for relevant and priority information is fundamental, as well
as the correct decision-making, identifying key situations that visualize potential
problems and possible lower risk solutions. Namely, to achieve a prospective vision
that allows projection to the organization, with proactive reasoning based on present
actions to visualize future objectives [6].
1.3 Personal Competencies
The statistics mentioned that no leadership model will have all the elements that
make it effective. The effective leader develops skills in three areas: personal, team
and organizational.
Personal competency is considered as the basis for the growth of the leader, since
it refers to leadership over oneself, here self-knowledge and awareness of skills,
strengths, weaknesses and values are highlighted, before leading the rest of people;
it also works on development, self-growth and continuous learning, which are based
on feedback and image as a role model for other followers.
In the case of the team competency, it is based on personal leadership, plus the
sum of values and competencies of the people, this competency means heading a
team that works in a coordinated way that pursues a common goal, and also the
leader seeks to lead the entire team toward the same vision, within the scope of their
competencies, contributing with collaboration, creativity and innovation, developing
relationships of trust and solidarity among team members.
Finally, the organizational competency is the one that handles the general context,
the macro-tendencies of the different systems, basing its action in the strategic think-
ing, where the organizational leader must establish the vision of the organizational
future, values, objectives, strategies, as well as making decisions and developing
actions to try to take advantage of the changes that occur in the external environment
of their institution [7].
1.4 Leadership Styles (Full Range Leadership

Model—FRLM)
Leadership has been one of the most important variables in the consideration of the
leadership effectiveness perceived by employees and their performance and satisfac-
tion at work.
Bass and Avolio, important leadership researchers, based on charismatic leader-
ship and transformational leadership, establish a new leadership model called multi-
factorial leadership theory in which the types of transformational, transactional and
laissez-faire leadership are included. These leadership researchers also raise a ques-
tionnaire of multifactorial leadership (MLQ Multifactor Leadership Questionnaire)
to evaluate and measure these types of leaders with their characteristics and effects.
The transformational leadership is the one that motivates, transforms, inspires
employees to overcome personal interests, seeking optimal performance, for organi-
zational benefit. It comprises five factors: idealized influence (attributed), idealized
influence (behavior), inspiration or inspirational motivation, intellectual stimulation
and individualized consideration (Table 1).
The transactional leadership is characterized through the fulfillment of objectives
that implies a reward; financial or moral; as well as, negative results due to poor
performance have a negative action from the leader. This leadership is not identified
by presenting a vision of the future, but rather by focusing on the current organi-
zational situation. It comprises three factors: contingent rewards, management by
active exception and management by passive exception (Table 2).
Table 1 Transformational leadership and its components

Components Definition
Idealized influence (attributed) Leaders serve as a role model for their followers. They are
admired, respected and inspire confidence
Idealized influence (behavior) Besides being models to be imitated, admired, respected and
give confidence, the leaders have an idealized behavior
Inspirational motivation Leaders motivate followers by providing a sense of work and
an attractive vision of the organizational future
Intellectual stimulation Leaders stimulate innovation and creativity. They also try
new approaches and seek solutions for themselves
Individualized consideration Leaders give individualized attention to each follower for
their achievement and growth by being a coach or mentor
Source Prepared by the authors, based on [16]
Table 2 Transactional leadership and components

Components Definition
Contingent rewards The leaders make an agreement with their follower
regarding a reward (material) before a well-executed task.
On the contrary, there may also be a punishment
Management by active exception The leader specifies the rules, limits and punishments.
The leader focuses on correcting errors and then
corrective measures that should be taken
Management by passive exception The leader usually leaves things as they are and only
intervenes when the problems become serious
Source Prepared by the authors, based on [16]
The laissez-faire leadership is considered a lack or absence of leadership, in which

the leader does not decide, does not guide, avoids tasks or making any decision and
being involved in important matters.
2 Military Multilevel Leadership Model
In the wide range of leadership, there is a topic of permanent relevance and it has been
the military leadership. Wong and several authors review the existing literature in this
field, highlighting the opportunities and importance of research in this field, bringing
up Hunt’s Multilevel Leadership Model, mainly for the various facets involved in
that leadership [8].
In addition, we know that this military leadership is classified into three levels,
which are related to levels of military strategy, requiring the military to develop
different types of competencies. Among the main required developed competencies
are the personal and institutional competencies, in order to be properly prepared to
be able to ascend to the next level of leadership.
That is why institutional competencies, classified in technical, interpersonal and
conceptual, are required at all levels of military leadership, but in different ways at
each level.
In the case of the technical skills, they are very necessary at the tactical level, since
the tasks developed at this level are predominantly technical and specific. However,
these competencies decrease as the leader rises in the operational level, and his needs
are even lower at the strategic level, which is no longer purely technical.
On the other hand, interpersonal skills are fundamental at all levels of leader-
ship; however, they are not always the same as they are developed according to the
level, because at the tactical level the people deal with military personnel, while at
the strategic level is required to have strong networking abilities, coordination and
negotiation with other institutions.
Finally, highlight that conceptual competencies begin at the tactical level, which
are being necessary to understand the importance of specific tasks for the use of
means, until reaching the strategic level where the global concept that implies leading
the organization is essential.
However, we not only have institutional competencies, but also personal compe-
tencies that exist within the military leadership, which refer to the development of
the military’s personal leadership, being these: personal, team and organizational.
Personal competency is referred to self-knowledge and self-awareness in the per-
sonal formation as a leader, fundamental in the level of tactical leadership at the
beginning of the military career, as the time passes by, the military personnel needs
less of this competency, as they evolve in the leadership levels, until they become a
leader with solid personal training at the strategic level.
Regarding the competency of teams and people, it will always be in the evolution
of the leader since the composition of the team members varies, for example, at the
tactical level teams will be made up of technical personnel, while at a strategic level
teams of directors, commanders, leaders of different disciplines must be led.
In the case of organizational competencies, these allow the leader to have a vision
of their environment, develop strategies to achieve objectives in the unit they lead,
which at the tactical level will be a small structure while at the strategic level the
whole organization will be.
It is important to note that currently the military leaders of countries such as
Canada, the USA and Portugal have focused their efforts on developing transforma-
tional leadership in depth, since it is due to its own characteristics of motivation,
transformation and the vision it transmits superiors to subordinates, the levels of
effort and satisfaction with the objectives have been evidenced, when it comes to
guiding behaviors in military organizations [9, 10], and also in civil organizations
[11, 12].
Therefore, even when the laissez-faire leadership is present in the tactical leader,
because he is a leader in the initial stage of their formation; that transactional lead-
ership is found at the operational level in actions, that generate responses and effects
that could be whether negative or positive from their superiors, and that transforma-
tional leadership is present in the strategic leader, who leads the entire organization;
there is no determined interrelation between the levels of leadership and the three
styles of leadership, being fundamental to establish this relationship as an initial step
to develop a Multilevel Military Leadership Model whose scheme is presented in
Fig. 1, and its purpose is focused on promoting the formation of qualified leaders
committed to the military organization.
3 Materials and Methods
The present study proposes the application of the Multifactorial Leadership Ques-
tionnaire (MLQ) of Bass and Avolio [13, 14], whose empirical validity has been
recognized and used to obtain data that allow to determine the style of leadership
and the relationship with variables of performance and satisfaction at work. The
Fig. 1 Military Multilevel Leadership Model. Source Prepared by the authors, based on [6, 7]
MLQ has some versions, we will use one adapted to the needs of the Armed Forces
of Ecuador [15].
The first stage will be done through surveys, whose questions will be based on
measurement instruments determined by the MLQ. This survey will be applied to
military officers of the three branches of the Armed Forces of Ecuador (Army, Navy
and Air Force), via Google Forms due to the respondents that are in different cities
of the country.
The officers have been initially considered, since they lead the different levels
of the military organization (tactical, operational and strategic) and are the ones
that receive the direct influence of their leaders, influencing their subordinates or
followers through their leadership, throughout the different charges they occupy.
The people surveyed will be stratified, based on the three leadership levels, in order
to provide data for each level. For the distribution and application of the survey,
the authorization of the corresponding authorities will be requested, establishing
preparatory meetings with the focal points determined by each military unit.
For the tabulation phase, data analysis and treatment of results, comparisons will
be made with similar studies in the military field. In addition, it is intended to deter-
mine, based on the surveys and data, the predominant leadership style at each level of
leadership, be it transformational, transactional and laissez-faire, as well as its rela-
tionship with influence on variables such as performance and satisfaction at work.
The final results will be needed to proceed to the validation stage using multivariate
techniques, determining the modeled relationship between the leadership style and
the aforementioned variables.
Finally, the results obtained with the organizational authorities will be analyzed
and discussed, in order to later move on to a socialization stage and structuring of
institutional leadership manuals with regional projection, which seeks to highlight the
importance of the influence of the type of leadership of officers in the organizational
management variables of the Armed Forces under a multilevel military leadership
approach.
4 Conclusion
This article proposes, through a Multilevel Military Leadership Model, to establish

the correlation between the levels and leadership styles based on the FRLM, empha-
sizing the importance and topicality of the multilevel analysis in the topic of lead-
ership. In addition, it is intended to determine the relationship between institutional
(technical, interpersonal and conceptual) and personal (personal, team and organi-
zational) competencies that are necessary for a greater or lesser degree according to
the levels of military strategy.
As a result of this study, conducted in first instance to officers of the Armed Forces
of Ecuador, it is expected to establish an approach of a Multilevel Military Leader-
ship Model, where the existing correlation between the levels and styles of leadership
allows the institutional command of the Armed Forces of Ecuador make better deci-
sions. Besides, they will be able to direct their efforts to guide the preparation of
leaders according to the leadership required by the organization, mainly aimed at
transformational leadership, for its approach of empowerment, commitment, inspi-
ration and motivation of people.
As a future work, this study is planned to be executed in the region, such as Armed
Forces officers from other countries, seeking to refine the model and determine
improvements in the processes of military leadership training in the region.
References
1. Dionne, S.D., et al.: A 25-year perspective on levels of analysis in leadership research, Leadersh.
Q. 25(1), 6–35 (2014)
2. Batistič, S., Černe, M., Vogel, B.: Just how multi-level is leadership research? A document
co-citation analysis 1980–2013 on leadership constructs and outcomes, pp. 1–18 (2016)
3. Hunt y, J., Ropo, A.: Multi-level leadership: grounded theory and mainstream theory applied
to the case of general motors 6(3), 379–412 (1995)
4. Van Wart, M.: Leadership in Public Organizations, 2nd edn. Routledge, New York (2014)
5. Reeves-Ellington, R.: Enviroscapes: a multi-level contextual approach to organizational lead-
ership. Emerald Gr. Publ. Ltd., pp. 337–420 (2015)
6. Jacobs, T.O.: Strategic Leadership: The Competitive Edge (2002)

7. Air-University, USAF Leadership Doctrine, vol. II, p. 71 (2015)
8. Wong, L., Bliese, P., McGurk, D.: Military leadership: a context specific review 14(6) (2003)
9. Moreira, M.C.M.: Liderazgo Transformacional y Género en Organizaciones Militares (2010)
10. Ivey, G.W., Kline, T.J.B.: Transformational and active transactional leadership in the Canadian
military. Leadersh. Organ. Dev. J. 31(3), 246–262 (2010)
11. Segovia, A.: El liderazgo, la compensación variable, el empowerment psicológico y su impacto
en la efectividad del empleado: un enfoque de modelación mediante ecuaciones estructurales,
Universidad Autónoma de Nuevo León (2014)
12. Saleem, H.: The impact of leadership styles on job satisfaction and mediating role of perceived
organizational politics. Proc. Soc. Behav. Sci. 172, 563–569 (2015)
13. Avolio, B.J., Bass, B.M.: You can drag a horse to water but you can’ t make it drink unless it
is thirsty. J. Leadersh. Stud. 5(1), 17 (1998)
14. Avolio, B.J., Bass, B.M., Jung, D.I.: Re-examining the components of transformational and
transactional leadership using the multifactor leadership questionnaire. J. Occup. Organ. Psy-
chol. 72, 441–462 (1999)
15. Samanta, I.: Modern leadership types and outcomes: the case of Greek public sector. J. Contemp.
Manag. Issues 23(1), 173–191 (2018)
16. Ganga-Contreras, F., Navarrete, E., Alt, C., Alarcón, N.: Percepción de los estilos de liderazgo:
el caso de un campus universitario IV (1)(2016), 1–35 (2016)
E-leadership Using WhatsApp,
A Challenge for Navy Organizations:
An Empirical Study
Rolando P. Reyes Ch., Luis Recalde Herrera, Galo Andrade Daza,

Victor Gómez Bravo and Hugo Pérez Vaca
Abstract Leadership is a fundamental characteristic of naval–military strategy.

However, the development of information and communication technologies (ICTs)
has led to the leadership being complemented by an emerging paradigm called
e-leadership. Therefore, it is interesting to evaluate this paradigm in a naval–military
organization where they are currently using WhatsApp as a tool to issue and follow
up on dispositions. Considering the perspectives of technology, communication, and
trust proposed by Avolio (Organ. Dyn. 2003, [3]), He et al. (E-leadership strategy in
virtual organizations and virtual teams, 2008, [8]) and Liu et al. (Leadersh. Organ.
Dev. J. 39(7):826–843, 2018, [10]), a survey is applied to Lieutenant Commanders
and Lieutenants of the organization. The results evidence a questionable use of e-
leadership over WhatsApp. However, e-leadership fails to be effective due to the
conflicts that are complicated to mitigate even in the traditional leadership of the
organization.
Keywords Leadership · E-leadership · Naval · Conflict · WhatsApp
R. P. Reyes Ch. (B) · L. Recalde Herrera · H. Pérez Vaca

Departamento de Seguridad y Defensa, Universidad de las Fuerzas Armadas ESPE,
Sangolquí, Ecuador
e-mail: rpreyes1@espe.edu.ec
L. Recalde Herrera
e-mail: llrecalde@espe.edu.ec
H. Pérez Vaca
e-mail: hlperez@espe.edu.ec
R. P. Reyes Ch. · G. Andrade Daza · V. Gómez Bravo
Centro de Estudios Estratégicos Marítimos CEESMA, Guayaquil, Ecuador
e-mail: gandraded@armada.mil.ec
V. Gómez Bravo
e-mail: vgomezb@armada.mil.ec

https://doi.org/10.1007/978-981-13-9155-2_15
172 R. P. Reyes Ch. et al.
1 Introduction
Leadership is a topic widely tackled by experts in business administration and man-

agement. Its importance has been such that it has also been studied in other areas,
such as: social, political, naval–military, among others [6]. However, the great devel-
opment of information and communication technologies (ICTs), digital communi-
cation, and other factors has allowed the leadership to begin to evolve or possibly to
complement an emerging paradigm that can act on changing and volatile scenarios
that impose the technology. This emergent paradigm is called as e-leadership (or
electronic leadership) [3]. In this regard, several authors [2, 3, 10] consider that this
type of leadership is an emerging research topic that poses several challenges and
contributions, both for business and military organizations, given its main character-
istic of allowing interaction and communication between the leader and his followers
in a virtual environment via the Internet [1].
Therefore, studying this paradigm not only in business organizations could be
interesting. Possibly, in the military–naval field, its study is transcendental due to
its hierarchical structure, command authority, and other elements depend on the
use of WhatsApp1 as a means to issue and follow up on dispositions. However, to
study e-leadership, it is necessary to take into account the proposed perspectives of
technology, communication and trust proposed by Avolio [3], He et al. [8] and Liu et
al. [10]. With this, it is possible to achieve an empirical evaluation of the existence of
possible e-leadership within a naval-military organization in Ecuador. For which, we
applied a randomized online survey to 46 Commander Lieutenants and Lieutenants
with extensive experience in leading onboard units. The results show the existence
of a questionable use of WhatsApp to exercise e-leadership in the naval-military
organization. However, there are conflicts that also often appear in his leadership
as a naval–military organization, the same ones that impede the effectiveness of
e-leadership in a virtual environment.
The rest of the article is structured as follows: Sect. 2 presents the background
and related studies as well as a brief explanation of the perspectives of e-leadership.
Section 3 details the research methodology. The results are explained in Sect. 4. The
threats to validity in Sect. 5. Finally, conclusions, discussion, and future work are
presented in Sect. 6.
2 Background and Related Studies
E-leadership is an emerging research topic and for this reason, several studies [2, 3]
have determined its importance for the challenges and benefits it brings to organiza-
tions; creating a new way of leading in a virtual environment. In this regard, we find
some empirical studies of e-leadership, as shown below:
1 WhatsApp is a messaging application for smartphones, which sends and receives messages via the
Internet, complementing instant messaging services, short message service or multimedia messag-
ing system.
E-leadership Using WhatsApp, A Challenge for Navy Organizations: An Empirical … 173
• Liu et al. [10], conducted a qualitative study to study the communicative adoption
of the e-leader. A model is used on structural equations proposed by Van Wart et
al. [12]. The authors mention that the model adjusts adequately to e-leadership,
achieving a greater understanding of the performance of e-leaders.
• Fan et al. [7] conducted an experimental study to test the influence of the moti-
vational language of the e-leader in virtual teams. The experiment was conducted
on 107 university students in Taiwan. The results showed that the instructions
given by digital media have positive effects on creativity and generation of ideas.
Although, the positive stimulus depends on the empathic writing of the e-leader.
• He et al. [8] conducted a study to establish the difference between leadership and
e-leadership, establishing the effectiveness and challenges of e-leadership in the
virtual environment, (e.g., technology, communication, cultural differences, trust,
and logistics). However, there are limitations such as: lack of confidence, lack of
cohesion, level of relationship, among others.
The aforementioned empirical works show that empirically evaluating e-leadership
in organizations is important. However, the evaluation could be specific to each orga-
nization because of the way each one works. Therefore, these authors [3, 8, 10] have
proposed strategies based on three perspectives: technology, communication, and
trust, as a starting point to evaluate e-leadership, as explained below.
2.1 Perspectives of E-leadership
According to [3, 8, 10], everyone considers that technology, communication, and

trust are perspectives that allow evaluating e-leadership in an organization. These
perspectives are briefly explained below:
• Technology. It is an important perspective because e-leadership appears in a virtual

environment where it is accessed through technological tools (e.g., WhatsApp,
e-mail, and others). That is why it is necessary to establish the competencies, such
as: learning, experience, and skills in ICTs of both the leader and members [8].
• Communication. This perspective refers to the communication between members
of a virtual group. Normally, the lack of presence of the members of the team allows
tension between them or simply that mutual understanding does not come to exist,
resulting in misinterpreted communications that limit communication in the virtual
environment [10].
• Trust. It is the perspective more complex than the previous ones because its artic-
ulation and development are very difficult in a virtual environment. There is no
physical presence and other factors [3].
The perspectives mentioned help to determine the situation of e-leadership in an

institution. Therefore, through these perspectives, it is possible to evaluate the situa-
tion of e-leaders and their virtual teams. Especially in a naval–military organization
where its leadership is paramount due it is more rigorous in terms of compliance
and supervision of dispositions, especially when in recent years its members have
ventured into the use of WhatsApp as a means to issue and follow up on dispositions.
This is the reason why we have posed the following research question (RQ):
RQ: What challenges can a naval–military organization find in its e-leadership

through electronic messaging (WhatsApp) considering the perspectives of technol-
ogy, communication, and trust?
3 Research Methodology
Our RQ proposes an exploratory research given the limited studies regarding

e-leadership in naval–military organizations. Therefore, it is appropriate to conduct
a survey as a methodological strategy. We use the survey creation recommendations
of Kitchenham and Pfleeger [9], which are summarized in:
1. Design of the Survey. An on-line survey was designed considering the guidelines
for creating and conducting surveys proposed by [9, 11]. This allowed, on the one
hand, to produce descriptive statistics, and, on the other hand, to obtain important
information to answer the research question proposed for this study.
2. Construction of the survey. Based on the perspectives of e-leadership mentioned
by [7, 8, 10] in the Sect. 2, the survey is constructed with 18 obligatory questions:
4 general information questions, 6 technology questions, 4 communication ques-
tions, and 4 questions of trust. The survey can be answered in an average of 10 min
(depending on the feedback provided) which does not fatigue the respondents.
The questions are closed so as not to give rise to erroneous interpretations or
irrelevant answers. The detail of the obtaining of the questions of the survey can
be found in the link: https://bit.ly/2QPLqOV.
3. Evaluation instrument. It was done in two stages. First, a researcher with experi-
ence in the subject of leadership reviewed the survey and provided feedback on its
readability, understandability, and potential ambiguities. Your feedback allowed
to lead to the elimination of 3 mandatory questions. Secondly, piloting was car-
ried out with two members of the naval–military organization (not included as
authors), who helped to pilot the survey before being formally applied.
4. Population and sample size. The population is oriented to the members of a
naval–military organization of Ecuador in different hierarchical levels: Comman-
ders Lieutenants and Lieutenants of the last years who have proven experience
of naval–military leadership. The population is adjusted around 1360 members.
To obtain the sample size (η), the following formula was applied (1):
(Z 2 ∗ p ∗ q ∗ N )
η= (1)
N ∗ e2 + Z 2 ∗ p ∗ q
Using this formula,2 and with a confidence level of 95%, an approximate of

50 members is the sample size obtained. We consider that the sample used is
representative, since they are members that have experience in leading their units
on board with approximately 18 to 25 years in the institution.
5. Data collection. The data collection was carried out in October 2018. The online
survey was anonymous, the same one that was announced by e-mail. In total, 46
46
surveys were successfully received, achieving an effective ratio of 50 = 92%, a
ratio that is adequate for this type of study.
4 Results
The results obtained from the survey are very interesting and allow us to answer our
research question. Next, the characteristics of the respondents are mentioned and
then our research question is answered.
4.1 General Characteristics of the Respondents
As mentioned above, 46 members of the naval–military organization successfully

answered the survey. Of these, 47.06% of respondents possess the rank of Com-
manders Lieutenant and 52.94 % of those surveyed possess the rank of Lieutenant.
All the respondents have between 5 and 7 years in that grade. Of these respondents,
67.65% are weapons specialty, 8.82% are technical specialty, and 23.53% are spe-
cialty services and administrative. Of them, 41.18% have been in the naval–military
operations areas.
4.2 RQ: What Challenges Can a Naval-Military

Organization Find in Its E-leadership Through
Electronic Messaging (WhatsApp) Considering the
Perspectives of Technology, Communication, and Trust?
The results allow us to respond to our RQ. However, they will be explained in the
three perspectives of e-leadership, as mentioned below:
2 Instatistics, the sample size (η) is the number of subjects that make up the sample drawn from
a population and are necessary for the data obtained to be representative of the population. The
variable Z is a constant that depends on the level of confidence for this study is 95%. The variable
p is the proportion of individuals who possess the characteristic of study in the population. This
data is generally unknown and it is usually assumed that p = q = 0.5, which is the safest option. The
variable q is the proportion of individuals that do not possess this characteristic, that is, it is 1-p.
Technology Perspective. According to He et al. [8], to exercise e-leadership, it

is necessary to use the right technological tools (e.g., WhatsApp). Therefore, the
importance of evaluating the experience and skills in ICTs. Table 1 shows the results,
in which it is observed that 55.88% of respondents actually use WhatsApp to issue and
make dispositions as part of their e-leadership within the naval–military organization.
We believe that this value could increase (to 88.23% of respondents) if the 32.35% of
respondents who answered sometimes are included in the future. Also, this 55.88%
of respondents mentioned that they have installed WhatsApp in their cell phones
6–9 years ago. As far as the experience and technological ability is insured in the
naval-military organization.
To establish the effectiveness of the skill and experience using WhatsApp in e-
leadership, respondents were also asked if they would have created groups to their
e-leadership. 91.18% of respondents mentioned having created between 0 and 50
virtual groups since they installed the tool. This percentage is similar to the percentage
of respondents (88.23%) who are using WhatsApp to issue their dispositions. In this
regard, we also asked about the average number of subordinates participating in
their virtual groups, almost the majority of respondents (44.12%) mentioned that
their virtual groups are made up of 6–10 subordinates, although sometimes they tend
to exceed ten subordinates (29.41%). In summary, almost half of the respondents
could be e-leading a small crew virtually.
Communication Perspective. According to Liu et al. [10], the communication in
e-leadership is usually very complicated due to the virtual environment of ICTs does
not allow the communicative fluency that physical presence usually offers. In this
regard, Table 2 shows that 50% of respondents confront communication conflicts in
e-leadership when using WhatsApp. Conflicts exist because there is no confirmation
of receipt of virtual messages by the subordinates or simply, the virtual messages of
the e-leader are ignored. Some respondents (29.41%) mentioned that communication
conflicts are usually caused by the lack of specific rules of participation at the time
of creation of the virtual group or other similar causes. Also, they were asked if they
have tried in any way to minimize these conflicts. The 32.35 % and the 35.29% of
respondents mentioned that they use an empathic language in their messages and
that they try to give immediate feedback to their issued dispositions, respectively.
This allows us to deduce that respondents make a communicational effort to avoid
conflicts. We also asked respondents if they use other means of communication
such as telephone or person-to-person conversation when conflicts have not been
mitigated through the virtual environment. The 64.71% of respondents indicated
that they have done it that way. Something interesting that we were able to determine
was that several of the conflicts were due to the disclosure (received/sent) of exclusive
information and high priority of the virtual group by other technological means (e.g.,
groups WhatsApp, Facebook, among others.).
Trust Perspective. Trust according to Avolio [3] is a very important factor in the
virtual environment. Without a high degree of trust, relationships in the virtual envi-
ronment become extremely fragile and e-leadership is condemned to be eroded. In
this regard, Table 3 shows that 52.94% of respondents believe that the trust within
WhatsApp should be indistinct among the members of the virtual team. Other respon-
Table 1 Technology perspective in e-leadership

Questions Answers %
Do you use “WhatsApp” to create virtual groups Always 55.88
with your subordinates where you mention and
follow up on their dispositions?
Sometimes 32.35
Never 11.76
How many virtual groups with your subordinates 0–50 virtual groups 91.18
have you created since you installed “WhatsApp”
for the first time?
51–100 virtual groups 0.0
More than 100 virtual 0.0
groups
I do not remember 8.82
What is the number (average) of subordinates that 2–5 subordinates 20.59
usually tend to conform to their virtual “WhatsApp”
groups?
6–10 subordinates 44.12
More than 10 subordinates 29.41
How long have you installed the “WhatsApp” on 6 to 9 years 55.88
your cell phone as a form of electronic messaging?
3 to 5 years 29.41
0 to 2 years 8.82
Do you believe that “WhatsApp” is or could be Yes 47.06
effective as an administrative means to give
dispositions?
Maybe 14.71
No 38.24
Are you aware of the different levels of Yes 85.29
technological competence about the use of whatsapp?
Maybe 14.71
No 0.0
dents (29.41%) mentioned that there should be trust from subordinates to the e-leader.
This allows us to reflect on a possible distrust of the superior; and for this reason, the
e-leaders considered that their subordinates could not ethically handle the disposi-
tion or information belonging to the virtual group (e.g., forwarded to other groups,
etc.). Almost half of respondents (41.18%) mentioned this strange behavior of their
subordinates while 44.12% of respondents are not sure if their subordinates would
have handled or not ethically the information or dispositions within the virtual group.
We also ask respondents if they apply confidence-building strategies. Several of
them mentioned writing congratulatory messages or similar within the WhatsApp
Table 2 Perspective of communication in e-leadership

Questions Answers %
During your communication in No confirmation of receipt of messages 50.0
e-leadership for “WhatsApp”, have you (ignore messages by the subordinate)
had any kind of conflict with subordinates
due to?
Due to technological limitations 20.59
By the excessive use of messages to give a 20.59
disposition
For the misinterpretation of 20.59
communications
Due to the lack of empathy among 11.76
members of the virtual team due to the
lack of presence
Due to the lack of specific participation 29.41
rules when the virtual group was created
(superiors/subordinates)
Mistrust between virtual members 8.82
I have not had conflicts 29.41
If there were virtual conflicts, did you use Yes 64.71
any telephone or face-to-face conversation
to resolve these conflicts in your virtual
team?
No 29.41
Maybe 5.88
To achieve a more creative performance or An empathic language in the messages 32.35
to carry out an efficient control and sent to the virtual group
follow-up of your dispositions through
“WhatsApp”, have you used?
Use of an enriched written communication 2.94
(emoticons)
An immediate feedback by messages to 35.29
avoid doubts regarding the dispositions
issued
Do not consider any aspect related to 29.41
improving performance with my
subordinates
When you receive superior dispositions Only e-leader 44.12
with exclusive information and high
priority through “WhatsApp”, this
information arrives
To subordinates by other means: 58.82
Internet, Facebook, other WhatsApp
groups, etc.
Table 3 Trust perspective in e-leadership

Questions Answers %
What kind of trust do you think may exist Confidence of the virtual team towards the 29.41
within a virtual group with subordinates? e-leader
Confidence of the e-leader towards the 17.65
virtual group
Confidence among all members of the 52.94
virtual team
There is no distrust 17.65
During your leadership in a virtual group, Yes 41.18
have you noticed that information or
dispositions are not handled ethically
(forwarded to other groups)?
No 14.71
Maybe 44.12
How do you raise confidence and improve Use fluid and proactive communication 47.06
the performance of your virtual team in in person between members before
“WhatsApp”? using WhatsApp
Try to minimize individualism and 20.59
motivate the virtual team
“Congratulations” messages or similar 47.06
within the WhatsApp group
I do not apply motivation 20.59
Within your virtual group in “WhatsApp”, A control model (continuous monitoring 55.88
have you followed up on its provisions by messages to the virtual group)
using some strategy?
A method of trust (relies on compliance 17.65
and only awaits compliance response)
None 26.47
group (47.06%) or simply they maintain a fluent and proactive communication before
interacting with WhatsApp (47.06%). Likewise, respondents were asked if they fol-
low up on dispositions. The results show that 55.88% of respondents use a control
model based on the continuous monitoring of virtual messages while a 17.65% con-
siders leaving the subordinate the compliance.
5 Threats to Validity
The possible threats to the validity of this study are discussed according to the aspects
mentioned by Creswell [5], as indicated below:
• Conclusion Validity. It refers to the degree to which the conclusions obtained are
derived from the data [5]. Only possible threat is the lack of representativeness of
the chosen sample. This threat does not operate, since it was chosen respondents
who possess a high level of experience and leadership practice within the naval–
military organization.
• Internal Validity. This type of validity refers to the absence of other influential
variables in the study, except those that are being studied [5]. In the online survey
conducted, the appearance of other influential variables should only appear due
to errors in the formulation or understanding of the survey questions. To mitigate,
an introduction to the survey is provided to familiarize the respondent with its
purpose. On the other hand, as a way to guarantee the veracity of the data, we have
given anonymity to the responders.
• Construct Validity. This type of validity refers to the relationship between a
theory behind the observation [5]. For this, a study of the previous literature about
the e-leadership mentioned in the Sect. 2 was carried out. Likewise, the survey was
evaluated by an experienced researcher and piloted by two members (non-authors)
of the organization.
• External Validity. External validity refers to the generalization of results [5].
The results obtained are valid externally because the selected sample size is very
representative in the naval–military organization.
6 Conclusions, Discussion and Future Work
The results obtained for our RQ based on the technology, communication, and trust
perspectives proposed by [7, 8, 10] show that e-leadership using WhatsApp in a
naval–military organization maintains conflicts that are not easy to solve in the virtual
environment, it is even possible to say that they are difficult to solve in a traditional
leadership. The conflicts we refer to are those that appear between perspectives: (1)
technology–communication, (2) communication–trust, and (3) trust–technology.
Regarding the first case, conflicts appear when the dispositions sent by virtual
messages are ignored. Reason why, the e-leader tries to make communication efforts
to mitigate these conflicts such as the telephone or in person. The second case refers
to conflicts for distributing exclusive information and high priority of the virtual
group by other electronic means (e.g., Facebook, other virtual groups, etc). The latter
is very complicated to mitigate by the e-leader given the degree of viralización of
information (without restriction) that WhatsApp allows. And regarding the third case,
conflicts are generated by the little effective trust that virtual environments allow,
since it is shown that subordinates do not usually handle ethically the information
and dispositions in their virtual group.
Practically, the e-leadership on electronic messaging tools like WhatsApp are still
complicated to articulate. Several of these conflicts that we mentioned were detected
by their authors [7, 8, 10] but in other virtual environments such as: videoconference
or e-mail. Therefore, before a military–naval organization wishes to complement
its traditional leadership with e-leadership, it is necessary to develop an attitude
framework based on institutional values and principles. Possibly, modeling an ethical
e-leadership over electronic messaging (WhatsApp) is the answer to the conflicts

found; maybe using a proposal similar to the one mentioned by Bass [4] but oriented
to virtual environments.
Finally, we consider investigating that could be happening in other organizations
similar to the study as: land–military and air–military organizations. For this reason,
we set out as future work, establish the effect or effects that e-leadership on WhatsApp
could have on these types of organizations.
Acknowledgements This work was supported by the Academia de Guerra Naval - Armada del
Ecuador and Dirección General de Educación y Doctrina and sponsored by Universidad de las
Fuerzas Armadas ESPE.
References
1. Aggarwal, A.: E-leadership-a new and modern style of leadership. IJAME (2018)
2. Avolio, B.J., Kahai, S., Dodge, G.E.: E-leadership: implications for theory, research, and prac-
tice. Leadersh. Q. 11(4), 615–668 (2000)
3. Avolio, B.J., Kahai, S.S.: Adding the e to e-leadership: how it may impact your leadership.
Organ. Dyn. (2003)
4. Bass, L., Israel, M.S.: Modeling ethical leadership: being an ethical leader means modeling
principles of self-awareness, reflective practice, transparency, and ethical behavior. In: Devel-
oping Ethical Principles for School Leadership, pp. 29–49. Routledge (2018)
5. Creswell, J.W., Clark, V.L.P.: Designing and Conducting Mixed Methods Research. Sage Pub-
lications, Thousand Oaks (2017)
6. Esguerra, G.A., Contreras, F.: Liderazgo electrónico, un reto ineludible para las organizaciones
de hoy. Estud. Gerenc. 32(140), 262–268 (2016)
7. Fan, K.T., Chen, Y.H., Wang, C.W., Chen, M.: E-leadership effectiveness in virtual teams:
motivating language perspective. Ind. Manag. Data Syst. 114(3), 421–437 (2014)
8. He, R., et al.: E-leadership strategy in virtual organizations and virtual teams (2008)
9. Kitchenham, B.A., Pfleeger, S.L.: Personal opinion surveys. In: Guide to Advanced Empirical
Software Engineering, pp. 63–92. Springer, Berlin (2008)
10. Liu, C., Ready, D., Roman, A., Van Wart, M., Wang, X., McCarthy, A., Kim, S.: E-leadership:
an empirical study of organizational leaders virtual communication adoption. Leadersh. Organ.
Dev. J. 39(7), 826–843 (2018)
11. Punter, T., Ciolkowski, M., Freimut, B., John, I.: Conducting on-line surveys in software engi-
neering. In: Proceedings of 2003 International Symposium on Empirical Software Engineering,
ISESE 2003, pp. 80–88. IEEE (2003)
12. Van Wart, M., Roman, A., Wang, X., Liu, C.: Integrating ICT adoption issues into (e-) leadership
theory. Telemat. Inform. 34(5), 527–537 (2017)
Part VI
Planning, Economy and Logistics Applied
to Defense
Career Anchors for the Portuguese
Army’s Volunteers and Contract
Personnel: Using the Career Orientations
Inventory
Lúcio Agostinho Barreiros dos Santos and

Maria Manuela Martins Saraiva Sarmento Coelho
Abstract This study aims to identify and discuss the career anchors of the Por-
tuguese Army’s volunteers and contract personnel, in order to outline a “profes-
sional profile” that can help the organization improving its decision-making process,
in terms of management practices and working and living conditions. The study
addresses the internal dimension of the concept of career using a three-pronged
approach—regulatory, theoretical, and empirical—to find out whether career anchors
are determinants for the choice to join the military, as well as for the length of
time served. The study uses quantitative and qualitative strategies, with emphasis on
Schein’s Career Orientations Inventory (COI). The survey had a valid final sample
of 1821 members from the Portuguese Army. The study’s findings suggest a pro-
fessional profile based on a dominant career anchor, “Lifestyle”, which, combined
with complementary anchors (e.g. “Security/Stability”), and provide a qualitative
explanation for the participants’ choices.
Keywords Military volunteers and contract personnel · Career anchors · Career

Orientation Inventory · Recruitment · Retention
1 Introduction
This study examines the career orientations of the Portuguese Army’s volunteer and
contract personnel to discover how this knowledge could help solve and/or mitigate
the current constraints to the model’s sustainability.
About 15 years after Portugal formally suspended its compulsory military service
(CMS) in the wake of it having been declared unconstitutional their requirements
L. A. B. dos Santos (B) · M. M. M. S. S. Coelho

Research and Development Center, Military University Institute, Lisbon, Portugal
e-mail: labs0892@gmail.com
M. M. M. S. S. Coelho
e-mail: manuela.sarmento2@gmail.com
Lusíada University, Lisbon, Portugal
https://doi.org/10.1007/978-981-13-9155-2_16
186 L. A. B. dos Santos and M. M. M. S. S. Coelho
[1] in 2004, implementing a strictly voluntary service [2–4], the military’s ability
to recruit and maintain the personnel needs has been progressively declining. These
difficulties are currently on the political agenda and have been addressed by multiple
studies conducted by the Ministry of National Defence, by the branches of the Por-
tuguese Armed Forces (AAFF), and by several independent researchers (e.g. [5–12]).
In the light of this, the paper attempts to diagnose the causes of the main difficulties
and devise ways to mitigate their negative consequences.
Despite the studies conducted in Portugal during recent years (e.g. [13, 14]), it
was only recently that some authors [9–11, 13] began to examine the AAFF’s man-
agement practices and the living and working conditions of citizens, who join the
military as volunteers for a maximum period of seven years [2, 3]. However impor-
tant, this awareness of some intrinsic factors in the organization’s functioning does
not seem to be sufficient to guarantee the success of the military professionalization
model, as the current situation is critical and calls into question the model’s sustain-
ability. Therefore, it remains necessary to explore, from a theoretical and practical
perspective, the reasons that hinder recruitment and retention, especially in terms of
“internal career” orientation [15, 16], a concept that will be explained further on.
The first phase of the study is focused on the Army branch, and its objective
was to identify and discuss the career anchors of the Portuguese Army’s volunteers
and contract personnel. In order to outline a “standard professional profile” which
the AAFF can use as a guide to select the most appropriate management practices
and create suitable working and living conditions, helping members to achieve their
“internal career” orientations as well as encouraging citizens to join the military.
The research question is: “What are the main ‘internal career orientations’ of the
Portuguese Army’s volunteer and contract personnel, and how do they relate to the
main reasons that lead citizens to join and remain in the military?”.
As for the quantitative dimension of the research, an adapted version of Schein’s
Career Orientations Inventory (COI) [15] was used to identify the main “career
anchors” of the military volunteers and contractors” of the Portuguese Army. The
qualitative dimension of the empirical study can provide a better understanding of
members’ professional profile, as well as increase awareness regarding the adoption
of management practices that can be used to improve recruitment and retention.
The article is divided into five sections. After this introductory section, the second
section briefly presents the legal framework of military service and the theoretical
framework, examining the concept of career anchors as a determinant for members’
professional choices and the nature and type of link that exists between them and
organization. The third section outlines and explains the methodology, and the fourth
section presents and discusses the study’s main findings. The fifth and final section
presents conclusions based on the main findings and answers the general objective
and the research question, also providing suggestions for future research.
Career Anchors for the Portuguese Army’s Volunteers … 187
2 Legal and Theoretical Framework
2.1 Legal Framework
The legal framework, an essential part of this research, concerns the service model
for volunteer and contract military personnel, its conceptual and legal organization,
and its operationalization. The model is defined in the legislation as voluntary service
and contract service (VS/CS) and is governed by a set of specific law, which include:
the Military Service Law, the Military Service Law regulations, and the regulations
of the military Incentive Scheme [2–4].
The concepts of recruitment and retention are especially relevant for this legal
framework. Military recruitment can refer to both voluntary and contract schemes,
and consists of the set of operations carried out to capture human resources, that is,
citizens above the age of 18 who are willing to serve in the AAFF [2]. Retention
refers to the need to maintain contract personnel and their expertise in the AAFF,
to meet specific organizational goals and obtain an appropriate return on the organi-
zation’s investment on education and training [3, 16]. If the “expectation of service
completion” is not met, members could have to pay damages proportional to their
training period and its associated costs.
Currently, the military voluntary and contract service model of the Portuguese
AAFF provides only short-term careers: if members fulfil certain requirements after
one year of voluntary service, they can become “regular” contract personnel for a
period of six years [2, 3]. Under certain conditions, members may also opt for a
medium-term career by entering a “special contract scheme” that allows them to
extend their contractual link up to 18 years of service. Despite a specific law having
been recently approved to allow for this type of career mobility [17], this option has
not yet been regulated and is in the process of regulatory expansion. In addition to the
possibility of career mobility between regular contract schemes and the possibility
to transfer to a longer-term contract scheme, members can also apply to the career
staff of the AAFF [2, 18], which provides them a long-term career until retirement
age.
In 2000, after the enactment of the new Military Service Law [2], an incentive
scheme [4] for military personnel serving under VS/CS was approved, which proved
instrumental to capture and retain, staff, as well as to support their socio-professional
transition to the labour market. Due to fiscal difficulties, this law has been revised
and reduced in its scope and possibilities. The increasing difficulties recruiting and
retaining staff are compounded by changes in the sociocultural, demographic, and
generational context. Therefore, a new law was recently approved expanding the
incentive scheme for military volunteers and contract personnel, which now mentions
the special military contract scheme [19].
2.2 Theoretical Framework
The theoretical framework looks at two distinct and complementary perspectives

for the concepts of career, external career, and internal career, as suggested by Derr
[20] and Schein [15], in order to discuss both the organization’s job offer and mem-
bers’ responsibilities. A career’s “external” component refers to the organization’s
needs, its functioning and culture, its management practices and constraints, and the
professional development opportunities it makes available to its employees.
A career’s “internal” perspective has an intra-individual nature and is the result
of a construction based on values, goals, ambitions, motivations, and individual
skills, which are consolidated (by developing self-knowledge) and based on reflection
and professional experience. This construction, which Edgar Schein called ‘career
anchors’, reveals a rather stable behaviour pattern, when it comes to making impor-
tant career decisions based on self-knowledge [15, 21].
After an initial classification using five types, Schein divided his career anchors
into eight final types, which are associated with certain identity components (e.g.
[15, 21–23]): (i) Technical/Functional Competence (TF) refers to specific skills
and areas of expertise; (ii) General Management Competence (GM) refers to
cross-sectional management skills and a markedly vertical career matrix; (iii) Auton-
omy/Independence (AU) concerns personal work organization (when and how
to work) and is based on flexibility and individual freedom of action; (iv) Secu-
rity/Stability (SE) emphasizes job and financial security, job stability, and remain-
ing in the same organization and functional area; (v) Entrepreneurial Creativity
(EC) reflects an entrepreneurial profile that favours innovation and creativity in the
management of objectives and projects; (vi) Service/Dedication to a Cause (SV) is
linked to the usefulness of the work and the opportunity to help others and solve soci-
etal problems and represents an orientation for the usefulness of the work performed;
(vii) Pure Challenge (CH) refers to complex challenges and obstacles that are diffi-
cult to overcome and translates into accepting a high level of risk and seeking novelty
and variety in the activities performed; and (viii) Lifestyle (LS) represents an appro-
priate balance and harmony between professional and personal/family requirements,
between one’s career and private life.
Therefore, the concept of career anchors is a determinant for professional choices
and reflects the nature of the link between an individual and his or her organiza-
tion. Regarding that link, two additional theoretical notions were selected for review
and exploration in this study—organizational career and protean career—which are
often differentiated by the nature of the link between individual and organization.
Organizational careers represent a traditional approach and are conceptually close
to a markedly vertical matrix, are defined by law, are largely controlled by the orga-
nization, and provide job/employment stability and security. Protean careers have
different elements, which are often the opposite of those in organizational careers.
They allow for greater fluidity and spontaneity and are therefore less institutionalized
since individuals have the power to make decisions and control most of the variables,
which are subject to his or her “individual will”, allowing for greater functional and
inter-organizational mobility [24–29].
Despite Schein’s [21] suggestion that each person has one anchor that defines
their strengths, goals, motivations, and values, several studies have tested the eight
types and have obtained different systematizations, which indicates that, while a
dominant anchor may exist, individual career orientations are a dynamic process and
a professional profile is consolidated over time, resulting from a combination of up
to three anchors, according to Feldman and Bolino [30].
On the other hand, the construct’s empirical limitations have led to other studies
that attempt to validate and/or refine Schein’s construct by reducing the number of
clusters (e.g. [31]) or, in contrast, by adding more clusters (e.g. Security/Stability;
Entrepreneurial-Creativity), such as those conducted by Danziger, Rachman-Moore,
and Valency [32]. Other studies even suggest additional anchors to meet twenty-
first-century requirements such as “Employability” and “Spiritual Dimension” [25].
Despite these various updates and adaptations, the foundations of the original con-
struct proposed by Schein seem to have remained current due to the instrument’s
internal consistency. However, in order to account for sociocultural, economic, gen-
erational, and other contextual factors, researchers should use the COI in combination
with other instruments to ensure more robust results. This study combines Schein’s
eight career anchors with a qualitative dimension.
Another concept addressed in this study, which is fully applicable to military
contexts, is that of psychological contract, which refers to employee perceptions
about the organization’s obligation to reciprocate the work performed [33, 34]. These
perceptions are based on explicit or implicit promises and may result from more or
less formal agreements with the organization, or simply from the expectations of
individuals regarding the organizational context and the interactions with some of
its elements [35]. The type of relationship that is established between individual
and organization and the degree to which the expectations match reality can either
strengthen or weaken the psychological contract and can ultimately lead to its dis-
solution (breach). This could cause employees to become less involved in activities
outside their regular duties, leading to uncertainty and dissatisfaction and increasing
employee exits from the organization [36–38]. Some common reasons for a breach of
the psychological contract may stem from (1) inadequate management practices, (2)
lack of support from the organization and/or management regarding external issues,
and (3) lack of equity among employees [35].
A review of the sociocultural and professional characteristics of the Y and Z
generations was conducted to investigate the implications of their specific features
for recruitment and retention. Despite their relevance, due to space constraints, the
results of the review are not included in this article.
Based on the regulatory and theoretical review, a chart was prepared to illustrate
the guiding structure for the study (Fig. 1).
Fig. 1 Guiding structure for the study: summarizes the theoretical framework and the main concepts
reviewed
This guiding structure includes the following elements:

– The military VS/CS model and its main dimensions—recruitment, retention, and
reintegration—were drawn from the review of the legal framework and are a key
element of the research framework;
– The structuring elements of the study’s theoretical framework are the concepts of
internal career (individual perspective), which includes “career anchors” and
external career (organizational context), in which opportunities and constraints
overlap with individual career and life ambitions and goals;
– The link between the perceptions about internal career, external career, and “psy-
chological contract”, which complements the “formal contract”, may help explain
the main career trends among the Portuguese Army’s volunteers and contract
personnel;
– Furthermore, the career profiles may help explain members’ behaviour towards
the AAFF: the reasons that lead them to join, the reasons for staying until the end
of their contract, and the motivations behind early dropouts;
– Similarly, the career anchors and psychological contract are key elements in the
definition of the concept of career/professional success, which is also addressed
in the theoretical framework and is part of the study’s research object.
The organizational aspects correspond to the “external career” concept, in which
management policies and practices serve to align organizational and individual
objectives, expectations, opportunities, and values (“internal career”). In military
contexts, these policies and practices are relatively standardized, although there are
some differences according to professional groups and areas.
3 Methodology and Method
This section summarizes the methodological procedures and identifies the popula-
tion, the sample, the gather data instruments, and the analysis techniques.
3.1 Research Methodology
The research is based on a case study [39] that investigates the applicability and value
of career anchors to the Portuguese Army’s volunteer and contract personnel during
2018.
The study used a mixed research strategy (quantitative and qualitative) developed
simultaneously in the context of the proof and of the discovery. The findings were
obtained by combining the processes of induction and construction of meaning, in
which the researchers assumed an ontological and epistemological posture closely
related to constructivism and interpretivism [40–42].
3.2 Research Method: The Sample and Data Processing

Techniques
The study used documentary data based on the legislation that regulates the military
VS/CS model, as well as empirical data obtained from a questionnaire that consisted
of 40 close-ended questions and four open-ended questions. The target public of the
questionnaire was volunteers in contract regime of the Portuguese Army belonging
to several professional categories.
The questionnaire used in this study (closed-ended questions) is adapted from
Dutra and Albuquerque questionnaire [22]. However, this one was also adapted from
Schein [15, 21].
In order to apply our questionnaire to the Portuguese military context, we did
three procedures: firstly, an adaption to the military population, secondly we did a
validation with experts on military human resources, and thirdly, we did a pretest
with 11 Portuguese military in regime of contract volunteers.
The final version of the questionnaire was online and the link was sent by e-mail to
the Army, Navy, and Air Force, being available for filling from March until October
2018. The answers were automatically introduced in an Excel database.
The study’s qualitative dimension (the questionnaire’s open-ended questions) was

combined with the COI to provide a deeper understanding of the research object [43].
We analyse the respondents’ answers regarding (i) the main reasons that led them to
join the military; (ii) the reasons for remaining until the end of their contract; (iii) the
motivations that lead to early dropouts; and (iv) the concept of career or professional
success.
We got 2474 answered surveys, but only 1832 had all the questions answered;
however, some were invalid, what correspond to 1821 valid surveys.
Table 1 shows the distribution of the final sample by professional category, gender,
and type of service.
The sample is robust and can be considered representative of the volunteers and
contract personnel of the Portuguese Army, which represent 26.78% of the Army’s
staff on 31 July 2018 (1821 respondents in a total of 6801 military). This percentage
is also close to the values of the population under analysis when compared with the
sample segmentation.
Crossing the age level with the professional category reveals that most participants
in the sergeants and enlisted categories are between 21 and 26 years old (70.56% of
sergeants and 70.81% of enlisted personnel), while most participants in the officers
category are between 24 and 29 years old (75.00%).
The methodological procedure used to process the answers to the closed-ended
questions was a descriptive statistical analysis, namely averages, percentages, and
frequencies. In the near future, we will perform bivariate and multivariate statistics
such as principal components analysis, cluster analysis, and inferential analysis.
The qualitative component of the research (the answers to the questionnaire’s
open-ended questions) was processed through content analysis using a specialized
software, SPSS Text Analysis for Surveys (STAFS), version 4.0.1. The analysis led
to the addition of some emerging types to the initial types, which correspond to
Schein’s eight career anchors.
Table 1 Sample characterization by professional category, gender, and type of service

Professional Volunteer Contract Total Volunteer Contract Total female Volunteer Contract Total
category service service male service service service service
Officers Value 2 89 91 0 29 29 2 118 120
% 2.20% 97.80% 100.00% 0.00% 100.00% 100.00% 1.67% 98.33% 6.59%
Sergeants Value 4 144 148 0 32 32 4 176 180
% 2.70% 97.30% 1.00% 0.00% 100.00% 100.00% 2.22% 97.78% 9.88%
Enlisted Value 125 1248 1373 14 134 148 139 1382 1521
% 9.10% 90.90% 100.00% 9.46% 90.54% 100.00% 9.14% 90.86% 83.56%
Sum Value 131 1481 1612 14 195 209 145 1676 1821
4 Data Presentation and Discussion of Results
4.1 Quantitative Research
An overall analysis of the results of the quantitative component reveals that the eight
career anchors defined by Schein are present in the sample, although in different
proportions.
The mean values show that all three professional categories (officers, sergeants,
and enlisted) and both genders have a dominant career anchor (“Lifestyle”), as well
as secondary career anchors (“Stability/Security”, “Technical/Functional”, “Service
and Dedication to a Cause”, and “Pure Challenge”), which obtained relatively high
values (Tables 2 and 3).
This analysis can be fundamental for the definition of a standard professional
profile (career orientations) of the citizens currently serving in the Portuguese Army
as volunteers or under contract.
Table 2 Mean value of career anchors by military category

Types of career anchors Officers Sergeants Enlisted
Value Order Value Order Value Order
Technical/Functional competence (TF) 4.86 2nd 4.74 2nd 4.56 3rd
General management competence (GM) 3.58 8th 3.91 8th 3.63 8th
Autonomy/Independence (AU) 3.79 7th 4.02 7th 3.97 7th
Security/Stability (SE) 4.73 3rd 4.60 5th 4.58 2nd
Entrepreneurial creativity (EC) 3.85 6th 4.02 6th 4.08 6th
Service/Dedication to a cause (SV) 4.58 4th 4.61 4th 4.49 4th
Pure Challenge (CH) 4.52 5th 4.66 3rd 4.41 5th
Lifestyle (LS) 4.99 1st 5.12 1st 4.94 1st
Table 3 Mean value of career anchors by gender

Types of career anchors Male Female
Value Order Value Order
Technical/Functional competence (TF) 4.60 2nd 4.60 4th
General management competence (GM) 3.66 8th 3.61 8th
Autonomy/Independence (AU) 4.00 7th 3.69 7th
Security/Stability (SE) 4.59 3rd 4.67 2nd
Entrepreneurial creativity (EC) 4.10 6th 3.75 6th
Service/Dedication to a cause (SV) 4.49 4th 4.62 3rd
Pure Challenge (CH) 4.44 5th 4.53 5th
Lifestyle (LS) 4.96 1st 5.02 1st
The dominant career anchor obtained much higher values than the second most
important anchor, although this difference is less marked in the officer’s category. The
relative importance of secondary anchors varies according to professional category
and gender: “Technical/Functional Competence” is the second most important career
anchor for officers and sergeants, as well as for men, and “Security/Stability” is the
second career anchor for enlisted personnel and for women. For the latter, the anchor
“Service/Dedication to a Cause” is also important.
The least relevant career anchor for all categories and genders is “General Manage-
ment Competence”, followed by the anchor “Autonomy/Independence”. The career
anchors “Service/Dedication to a Cause” and “Pure Challenge”, which are charac-
teristic of the military environment, are moderately important for military volunteers
and contract personnel, and are slightly higher for the sergeants category and for
female members.
It is relevant to conclude that women score high in the main anchors, and on the
other hand, they score low in the less important anchors, which can indicate a more
developed self-concept in terms of professional orientation.
4.2 Qualitative Research
The results obtained from the analysis of the qualitative component of the ques-
tionnaire confirm that career anchors are generally present in all phases dur-
ing the service term in the AAFF—recruitment, retention, and transition to the
labour market—although some anchors (General Management Competence, Auton-
omy/Independence, and Entrepreneurial-Creativity) obtained insignificant values,
which confirms the general trend of the results of the quantitative analysis.
Figures 2, 3, 4, and 5 summarize the results of the content analysis of the four
open-ended questions.
Only results that obtained 15 or more occurrences were included in the type net-
works, regarding the initially defined types (which means that not all types associated
with the anchors are shown in the images) and emerging types.
The size of the nodes and the thickness of the line segments that link them rep-
resent, respectively, the importance of each type within each topic/question and the
strength of the relationship between the types.
Fig. 2 Reasons for joining

the military
Fig. 3 Reasons for remaining in the military
Fig. 4 Reasons for early separation
Fig. 5 Concept of career/professional success
First and Second Open-Ended Question: Reasons for Joining and Remaining
in the Military
Reasons for joining the military. The types of career anchors Pure Challenge (CH)
(505 occurrences) and Service/Dedication to a Cause (SV) (261 occurrences), as
well as the relationship between them, obtained more relevant values than the other
types, which confirms the results obtained by earlier studies done by AAFF in 2012,
2016, and 2017 [7–9]. Security/Stability (SE) is a key type for the process of joining
the military, showing a significant correlation with the emerging type Employability.
In military context, this last type represents the choice to remain in the organization
under VS/CS or to seek further qualifications and/or apply to the career staff.
Reasons for joining the military. The type of career anchor Security/Stability (SE)
(750 occurrences) is one of the most important reasons given by respondents for
remaining in the military and is associated with the types Pure Challenge (CH)
(240) and Service/Dedication to Cause (SV) (119), as well as with the type of
Employability (138). Work Conditions and VS/CS Incentive Scheme are emerging
types closely correlated to the quality of the work environment and the possibility to
benefit from the incentive scheme provided by law.
Third Open-Ended Question: Reasons for Early Separation
The reasons that lead members to separate from the military before completing the
minimum length of service defined by law, although they still wish to maintain their
“employee” status (Security/Stability) (650 occurrences), stem from dissatisfaction
regarding certain individual needs. When faced with Work Conditions (310), which
they find inadequate, members look for job and financial stability outside the AAFF,
where they believe they can achieve better life conditions (Lifestyle (LS) (270).
In most cases, the Employability type (130) refers to the possibility of joining the
military for the advantages, when applying to a position in the security forces or to
certain positions in the Portuguese civil administration (provided for by law). The
VS/CS Incentive Scheme could be especially relevant in these cases.
Fourth Open-Ended Question: Concept of Career/Professional “Success”
The values obtained in the types of career anchors Security/Stability (271 occur-
rences) and Lifestyle (103), as well as the correlation between them, show that the
respondents’ concept of professional success confirms the overall relevance of the
corresponding anchors in the quantitative component of the analysis. Here, the first
type correlates to all the others. The emerging type Achievement (161) is the second
most important type for the concept of success and correlates with each of the types
obtained from the responses to the fourth open-ended question. The other emerg-
ing types that explain respondents’ perceptions are Employability (57), Recognition
(78), Military Career (101), Career Development (69), and Successful (46).
5 Interconnection Between Quantitative and Qualitative

Results
This section presents the main conclusions of the methodological procedure and
provides an answer to the general objective and research question.
The study’s general objective was to identify and discuss the types of career
anchors of the Portuguese Army’s volunteers and contract personnel. The purpose is
to outline a “standard professional profile” which can be used as a guide to improve

the AAFF’s management practices and create suitable working and living conditions,
allowing individuals to achieve their “internal career” orientations and encourage
them to join the military.
The study used a mixed research strategy (quantitative and qualitative) developed
simultaneously in the context of the proof and in the context of the discovery. The
study used documentary and empirical data obtained from a questionnaire consisting
of close- and open-ended questions, which was delivered to the Portuguese Army’s
volunteers and contract personnel, resulting in a sample of 1821 respondents.
The results of the quantitative component of the study revealed a dominant
career anchor “Lifestyle”, as well as several secondary anchors, namely “Stabil-
ity/Security”, “Technical/Functional”, “Service and Dedication to a cause”, “Pure
Challenge”, which vary slightly according to some demographic variables.
The results of the qualitative component seem to indicate that anchors are present
during all phases of the service term in the AAFF (recruitment, retention, and socio-
professional transition to the external labour market), confirming the general hypo-
thetical perspective presented in the theoretical framework.
Crossing the identified anchors with the main reasons for joining and remaining
in the AAFF provided a better understanding of the standard professional profile of
VS/CS personnel and how it could influence the adoption of management practices
that can be used to improve recruitment and retention.
The answer to the research question “What are the main internal career orientations
of the Portuguese Army’s volunteer and contract personnel, and how do they relate to
the main reasons that lead citizens to join and remain in the military?” was obtained
from two factors:
– The definition of the standard “Professional Profile” of the Portuguese Army’s
VS/CS personnel and
– The organization’s role in helping its staff to meet their “Internal Career” needs.
6 Final Conclusions
This section discusses the study’s contributions, usefulness, and limitations, as well
as it makes recommendations and suggestions for future studies.
According to the “Professional Profile”, VS/CS Personnel, the Results Suggest
– Seek to balance their personal, family, and career/professional needs, which career
anchor is the “Lifestyle”.
– Prefer to work in specific areas of expertise and/or technical functions, where they
can use the qualified skills they acquired previously or developed in a military
context, which career anchor is “Technical/Functional Competence”.
– Seek stability, work benefits, and financial security through long-term employment
in an organization, preferably, and if possible, in a functional and geographic area
that allows them to be close to their family/close social group, which career anchor
is “Security/Stability”.
– Are (moderately) motivated by the symbolic, vocational, and operational aspects
of military service, by the utility of the work and its role in helping to solve societal
problems, which career anchor are “Pure Challenge, and Service/Dedication to a
Cause”.
– Find typical management and general management activities and responsibilities,
which career anchor is “General Management Competence”, somewhat unappeal-
ing.
Regarding the Organization’s Role, the Results Suggest

– That the AAFF should accept the standard professional profile of its VS/CS per-
sonnel and help them achieve their anchors.
– It is imperative to implement management practices that can operationalize the
professional profile obtained and provide suitable working and overall living con-
ditions for the AAFF’s volunteers and contract personnel.
– Due to some of the differences that emerged regarding respondents’ perceptions
about secondary career anchors, the skills of the AAFF’s volunteers and contract
personnel should be managed in a more flexible and personalized way, in order
to meet the organization’s objectives while simultaneously meeting members’
“internal career” needs.
The usefulness of this study is based on a deeper knowledge of the career ori-
entations of the Portuguese Army’s volunteers and contract personnel. This will
facilitate the adoption of management practices that are better suited to members’
professional profile, aligning organizational and individual objectives. The observed
“career anchors” could be also suitable to explain the reasons why someone either
decide or not to join the army.
Therefore, the suggestions for future research include the expansion the quan-
titative methodological procedure and a denser qualitative analysis, as well as the
replication of the research scope to the other branches of the Portuguese AAFF.
It would be very interesting to analyse what are the differences and the similarities
among the “career anchors” EU countries, when the current political agenda is about
a common Army and a common European defence.
References
1. Law No. 1/97, of 20 September: Approves the Fourth Revision of the Constitutional Law.
Assembly of the Republic. Diary of the Republic, Lisbon (1997)
2. Law No. 174/99 of 21 September: Approves the Military Service Law. Assembly of the Repub-
lic. Diary of the Republic, Lisbon. https://dre.pt/application/dir/pdf1sdip/1999/09/221A00/
65416550.pdf (1999)
3. Decree-Law No. 289/2000 of 14 November: Approves the Regulation of the Military Service
Law. Portuguese Government. Diary of the Republic, Lisbon. https://dre.pt/application/dir/
pdf1sdip/2000/11/263A00/64256438.pdf (2000)
4. Decree-Law No. 320-A/2000 of 15 December: Approves the Regulation of Incentives to Ser-

vice in RV/RC (amended by Decree-Law No. 118/2004 of 21 May and by Decree-Law No.
320/2007 of 27 September). Portuguese Government. Diary of the Republic, Lisbon. https://
dre.pt/application/dir/pdf1sdip/2000/12/288A01/00020011.pdf (2000)
5. Bragança, N., Santos, L.: Study 2 – Desafios do recrutamento para o futuro em relação aos
cidadãos para a prestação de serviço militar nos regimes de voluntariado e de contrato das
Forças Armadas portuguesas [Future challenges in the recruitment of volunteer and contract
service members for the Portuguese Armed Forces]. In: Santos, L., Sarmento, M., Fachada, C.
(Coord.). Prestação de serviço militar em regime de contrato nas Forças Armadas portuguesas.
Dificuldades e desafios num contexto em transformação. [Contract service in the Portuguese
Armed Forces. Difficulties and challenges of a changing context.] ARES Collection, vol. 22,
pp. 53–103. Military University Institute, Lisbon (2018)
6. Branco, A., Santos, N.: Study 3 – Predisposição para ingresso nas Forças Armadas: Um estudo
sobre ações de recrutamento dirigido. [Predisposition to join the Armed Forces: A study on
directed recruitment.] In: Santos, L., Sarmento, M., Fachada, C. (Coords.). Prestação de serviço
militar em Regime de Contrato nas Forças Armadas Portuguesas. Dificuldades e desafios num
contexto em transformação. [Contract service in the Portuguese Armed Forces. Difficulties and
challenges of a changing context.] ARES Collection, vol. 22, pp. 105–15. Military University
Institute, Lisbon (2018)
7. Centre of Applied Psychology of the Army: Summary analysis of Training Dropouts and
Personnel Retention. Portuguese Army, Lisbon (2016)
8. General Directorate of Military Personnel and Recruitment: Study on the reasons for VS/CS
personnel dropouts in the Armed Forces. Ministry of National Defence, Lisbon (2012)
9. General Directorate of National Defence Resources: Study on the Sociodemographic Charac-
teristics and Organizational Satisfaction of Volunteer and Contract personnel from the three
branches of the Armed Forces. Scientific coordinator – Professor Helena Carreiras. Ministry
of National Defence, Lisbon (2017). (Restricted document)
10. Rijo, F., Marreiros, J., Mairos, J., Paquete, O.: Employee Retention in the Military. J. Mil.
Sci. VI(1), 333–356. https://www.ium.pt/cisdi/index.php/pt/publicacoes/revista-de-ciencias-
militares (2018). Accessed 9 Oct 2018
11. Santos, L., Sarmento, M.: Study 1 – Evolução do modelo de prestação de serviço militar nas
forças armadas portuguesas em regime de contrato. Análise crítica e formas de intervenção.
[Evolution of the contract service model of the Portuguese armed forces. Critical Analysis and
intervention measures.] In: Santos, L., Sarmento, M., Fachada, C. (Coord.). Prestação de serviço
militar em regime de contrato nas Forças Armadas portuguesas. Dificuldades e desafios num
contexto em transformação. [Contract service in the Portuguese Armed Forces. Difficulties and
challenges of a changing context.] ARES Collection, vol. 22, pp. 19–51. Military University
Institute, Lisbon (2018)
12. Santos, L.: Reflections Arising from military service under contract in the portuguese armed
forces: functional and social and citizenship perspectives. J. Mil. Sci. (online) III(1), 331–362.
https://cidium.ium.pt/docs/artigos/Artigo_123.pdf (2015)
13. Batista, M., Ribeiro, R.: Study 7 – Carreira Militar – Motivações e expetativas dos militares
do regime de Contrato. [Military Career – Motivations and expectations of military contract
personnel.] In: Santos, L., Sarmento, M., Fachada, C. (Coords.). Prestação de serviço militar em
Regime de Contrato nas Forças Armadas Portuguesas. Dificuldades e desafios num contexto em
transformação. [Contract service in the Portuguese Armed Forces. Difficulties and challenges
of a changing context.] ARES Collection, vol. 22, pp. 289–316. Military University Institute,
Lisbon (2018)
14. Cunha, F., Costa, P.: Study 3 – Quadro de Incentivos à prestação do serviço militar como
estímulo ao recrutamento e à retenção. [The role of the military Incentive Scheme in promoting
recruitment and retention.] In: Santos, L., Sarmento, M., Fachada, C. (Coords.). Prestação
de serviço militar em Regime de Contrato nas Forças Armadas Portuguesas. Dificuldades e
desafios num contexto em transformação. [Contract service in the Portuguese Armed Forces.
Difficulties and challenges of a changing context.] ARES Collection, vol. 22, pp. 193–238.
Military University Institute, Lisbon (2018)
15. Schein, E.: Career Anchors: Discovering Your Real Values. Pfeiffer & Company, University
Associates, San Diego (1990)
16. RTO/NATO: Recruiting and Retention of Military Personnel. Technical Report of RTO
(Research Task Group), HFM-107, NATO Research & Technology Organization. https://www.
nato.int/issues/women_nato/Recruiting%20&%20Retention%20of%20Mil%20Personnel.pdf
(2007)
17. Decree-Law No. 75/2018 of 11 October: Second amendment to Decree-Law No. 130/2010 of
14 December, approving the special military contract scheme. Portuguese Government. Diary
of the Republic, Lisbon (2018)
18. Decree-Law No. 90/2015 of 29 May: Approves the Statute of the Armed Forces Military Person-
nel. Portuguese Government. Diary of the Republic, Lisbon http://www.emgfa.pt/documents/
cqw3zjnhvg4s.pdf (2015)
19. Decree-Law No.76/2018 of 11 October: Establishes the incentive scheme for former or active
duty AAFF personnel under contract (CS), special contract (SCS) or volunteers (VS). Por-
tuguese Government. Diary of the Republic, Lisbon. (2018)
20. Derr, C.: Five definitions of career success: implications for relationships. Appl. Psychol. 35(3),
415–435 (1986)
21. Schein, E.H.: Identidade Profissional: como ajustar suas inclinações e opções de trabalho.
Nobel, São Paulo (1996)
22. Dutra, J.S., Albuquerque, L.: Âncoras de Carreira – In: Career Anchors by Edgar H. Schein.
Adapted for Human Resource Management. FEA-USP, São Paulo (2009)
23. Rego, A., Cunha, M., Gomes, J., Cunha, R., Cabral-Cardoso, C., Marques, C.: Manual de
Gestão de Pessoas e do Capital Humano, 3rd edn. “Gestão” Collection. Edições Sílabo, Lda,
Lisbon (2015)
24. Arthur, M., Rousseau, D.: The Boundaryless Career: A New Employment Principle for a New
Organizational Era. University Press, New York, Oxford (1996)
25. Baruch, Y.: Managing Careers: Theory and Practice, pp. 79–80. Prentice Hall, FT, Harlow,
England (2004)
26. Chanlat, J.F.: Quais carreiras e para qual sociedade? RAE – Revista de Administração de
Empresas. São Paulo 35(6), 67–75 (1995)
27. Dutra, J.: Administração de Carreiras – Uma Proposta para Repensar a Gestão de Pessoas.
Atlas, São Paulo (2012)
28. Hall, D.: Careers In and Out of Organization. Sage, London (2002)
29. Martins, H.: Gestão de Carreiras na Era do Conhecimento. Qualitymark, Rio de Janeiro (2010)
30. Feldman, D.C., Bolino, M.C.: Careers within careers: reconceptualising the nature of career
anchors and their consequences. Hum. Resour. Manag. Rev. 6(2), 89–112 (1996)
31. Coetzee, M., Schreuder, D.: Using the career orientations inventory (coi) for measuring internal
career orientations in the South African organisational context. S. Afr. J. Ind. Psychol. 35(1),
99–111 (2009). https://doi.org/10.4102/sajip.v35i1.806
32. Danziger, N., Rachman-Moore, D., Valency, R.: The construct validity of Schein’s career
anchors orientation inventory. Career Dev. Int. 13(1), 7–9 (2008). https://doi.org/10.1108/
13620430810849506
33. Rousseau, D.M.: Psychological and Implied Contracts. Empl. Responsib. Rights J. 2(2),
121–139 (1989)
34. Rousseau, D.: Psychological Contracts in Organizations: Understanding Written and Unwritten
Agreements. Sage, Thousand Oaks (1995)
35. Conway, N., Briner, R.: Understanding Psychological Contracts at Work. Oxford University
Press, New York (2005)
36. Cable, D.A.J.: The Psychological contract: the development and validation of a managerial
measure. Ph.D. Thesis in Phylosophy. Hamilton, Waikato University, New Zeeland (2008)
37. Turnley, W.H., Feldman, D.C.: Re-examining the effects of psychological contract violations:
unmet expectations and job dissatisfaction as mediators. J. Organ. Behav. 25–42 (2000)
38. Van de Ven, C.: The psychological contract: a big deal! (Chapter 3E). In: Recruiting and
Retention of Military Personnel, Literature Reviews for Selected Recruitment and Retention
Topic 3-1 Areas, pp. 3E1-19. Research and Technology NATO, Paris. https://www.nato.int/
issues/women_nato/Recruiting%20&%20Retention%20of%20Mil%20Personnel.pdf (2007)
39. Yin, K.R.: Case Study Research – Design and Methods, 5th edn. Sage Publications, Los Angeles
(2014)
40. Bryman, A.: Social Research Methods, 4th edn. Oxford University Press, Oxford (2012)
41. Creswell, J.W.: Qualitative Inquiry & Research Design: Choosing Among Five Approaches,
3rd edn. Sage Publications, Los Angeles (2013)
42. Guerra, I.C.: Pesquisa Qualitativa e Análise de Conteúdo: Sentidos e Formas de Uso. Principia,
Parede (2006)
43. Denzin, N., Lincoln, Y. (eds.): Collecting and Interpreting Qualitative Materials, 4th edn. Sage
Publications, Los Angeles (2013)
Part VII
Simulation and Computer Vision in
Military Applications
Implementation of Dubin Curves-Based
RRT* Using an Aerial Image
for the Determination of Obstacles
and Path Planning to Avoid Them During
Displacement of the Mobile Robot
B. Daniel Tenezaca, Christian Canchignia, Wilbert Aguilar

and Dario Mendoza
Abstract The application of mobile robots in autonomous navigation has con-

tributed to the development of exploration tasks for the recognition of unknown
environments. There are different methodologies for obstacles avoidance imple-
mented in mobile robots; however, this research introduces a novel approach for
a path planning of an unmanned ground vehicle (UGV) using the camera of a drone
to get an aerial view that allows to recognize ground features through image process-
ing algorithms for detecting obstacles and target them in a determined environment.
After aerial recognition, a global planner with Rapidly-exploring Random Tree Star
(RRT*) algorithm is executed, Dubins curves are the method used in this case for non-
holonomic robots. The study also focuses on determining the compute time which is
affected by a growing number of iterations in the RRT*, the value of step size between
the tree’s nodes and finally the impact of a number of obstacles placed in the envi-
ronment. This project is the initial part of a larger research about a Collaborative
Aerial-Ground Robotic System.
Keywords RRT* · Path planning · Image processing
B. Daniel Tenezaca (B) · C. Canchignia · W. Aguilar · D. Mendoza

Universidad de las Fuerzas Armadas ESPE, Sangolqui, Ecuador
e-mail: hdtenezaca@espe.edu.ec
C. Canchignia
e-mail: cscanchignia@espe.edu.ec
W. Aguilar
e-mail: wgaguilar@espe.edu.ec
D. Mendoza
e-mail: djmendoza@espe.edu.ec

https://doi.org/10.1007/978-981-13-9155-2_17
206 B. Daniel Tenezaca et al.
1 Introduction
Unmanned aerial vehicles (UAVs) and mobile robots are used in different operations
according to their performances. UAVs have many limitations, for example, the
carrying capacity, flight time, etc. However, UAVs has a more ample range of vision
of a workspace and higher velocities [1]. On the contrary, mobile robots present many
advantages compared to UAVs. Mobile robots can carry heavier loads, have more
power units, and higher capacity for processing [2, 3]. The purpose is to combine
them to create a hybrid system taking into consideration the advantages of those
two robotic systems. In this proposal, the UAV is able to analyze a broad portion of
ground in specific workspace to provide safe displacement directions to the grounded
vehicle [4, 5].
The research presents a study to establish a global planner that represents the data
obtained from a top view generated by UAV [6]. First, the system develops an image
through processing algorithms based on artificial vision in order to determinate the
most relevant features inside the picture. In this case, the algorithm estimates the
position of the obstacles as well as the mobile vehicle’s destination.
LaValle [7] proposed RRT algorithm considered as an optimal technique to con-
struct trajectories in nonconvex high-dimensional spaces and used to solve path
planning problems. Commonly, an RRT is deficient to solve a problem related with
the planning; therefore, it is necessary to implement a path planning algorithm based
on that RRT* to expand the tree in free spaces [8]. The kinematic constraints of the
ground vehicle based on Ackerman steering makes RRT* algorithm grows drawing
curves easy to adapt the vehicle turning movements. This kind of curves is named
Dubins path. Finally, it is important to consider adjusting many parameters to expand
the tree; it will allow to know how the algorithm is affected when existing variations
of the obstacles, number of iterations and the length of the connection between the
tree’s nodes.
In Sect. 2, the article presents an image processing algorithm for constructing
a global planner. In Sect. 3, it details an analysis to establish a RRT* algorithm
using a geometrically method based on constructing tangent lines that connect two
points with a shortest curve in a Euclidean plane. Finally, in Sect. 4 results will be
presented according to the data obtained in different scenarios and various obstacle
configurations.
2 Image Processing
The main objective of image processing is to remark certain details on an image.

For this reason, processing images from the UAV’s camera is important to obtain a
2D binarized map where obstacles are recognized [9]. All stages are specified in the
next flowchart (Fig. 1):
Implementation of Dubin Curves-Based RRT* … 207
Transformaon
Image Image Image Morphological
from RGB to
Acquision smoothing binarizaon operaons
HSV
Fig. 1 Flowchart image processing
2.1 Image Acquisition
Parrot Bebop 2 was used as UAV. It has a 14-megapixel ‘fisheye’ with a 3-axis image
stabilization system that maintains a fixed angle of the view. The drone’s camera has
to point to the ground (vertical position), for getting vertical camera’s rotation we
have used a software development kit provided by Robot Operating System (ROS).
Experimental results show to an altitude of 4 m above ground level our region of
interest (ROI) covers an approximately metric area of 6 m × 4 m.
2.2 Changing Color Spaces
Object discrimination using red, green, and blue color space (RGB) turns difficult
because the object’s color is correlated with the amount of light exposure at the
moment. For that reason, hue, saturation, and value color space (HSV) is more
relevant to discriminate colors regardless the amount of light at the scene. Change
of color spaces from RGB to HSV in this application allows better adapting of the
obstacles’ algorithm to light changes, in cases which shadows contribute to false
positives [10].
2.3 Image Smoothing
Noise reduction is a very important factor for good image processing. In this case, it
is applied a Gaussian Blur filter [11]. This means that each pixel was affected for a
kernel (neighbor pixels). A kernel of size 5 was used and a Gaussian kernel standard
deviation in an X direction of 0, where the target is not to get a blurred image if else
to eliminate the small percent of noise.
2.4 Binarization
In our application, it was important to separate foreground elements (obstacles) from

the background of the image, binarization, or thresholding methods are commonly
used for that purpose [12]. The obstacles were considered the most outstanding
element on the ground, therefore, establishing a binarized map where the obstacles
have a pixel value of 0 and a pixel value of 1 is a free space. The hue value which
represents red color goes from 0º to 30º and 330º to 0º approximately. Another stage
is using the binarization to detect the target blue [13].
2.5 Edge Dilatation and Erosion
The dilation convolution adds pixels to the boundaries of the obstacles’ structure
reducing noise, while erosion removes pixels on the outside boundaries of the obsta-
cles’ structure. The dilatation and erosion combined result in better filter. It fills holes
(opening) and removes small objects (closing). The opening operation works with a
structuring element a size kernel of 5. The operation affects the pixel and a 5 × 5
matrix size, meanwhile closing operation uses a size kernel of 21 [14].
3 RRT* Algorithm
Dubins curves-based RRT* is used as a global planner. It establishes the path by two
types of configurations divided into curve—straight—curve (CSC) and curve-curve-
curve (CCC), which are the pattern of expansion in the RRT* algorithm. Random
sample paths are generated from combination of straight lines and curves.
3.1 Initial Requirements
Data input to get the path with the RRT* algorithm serve as initial state and orienta-
tion, numbers of iterations, goal state, and step size. To set Dubin curves parameters,
it is necessary to know the minimum turning radius of the robot. Finally, another
important input data is having the binarized map (workspace).
3.2 RRT* Algorithm
According to Noreen et al. [15], RRT* is based in a group of features which allow
tree expansion very similar to RRT algorithm. The difference between the systems
is that RRT* incorporates two special properties called near neighbor search and
rewiring tree operations.
The algorithm is represented by a tree denoted as T = (V, E), where V is a set
of vertices and E is a set of edges. The initial configuration (qinit ) includes a set
of vertices that represents where the tree starts. In each iterations configured (Lines
Table 1 Pseudocode RRT* algorithm

Algorithm RRT*
T = (V, E) RRT*
1:
2:
3:
4:
5:
6:
7:
8:
9:
10:
11:
12:
3–11), the algorithm establishes a random position (qrand ) in free region, (qnear est ) is
searched in the tree according to predefined step size from (qrand ) and immediately
the algorithm establishes a new configuration (qnew ) with Steer function which
guides the system from (qrand ) to (qnear est ). The function Choosepar ent allows to
select the best parent node for the new configuration (qnew ) before its insertion in
tree considering the closest nodes that lie inside of a circular area, finding (qmin ).
Finally, near neighbor operations allow to generate the optimal path repeating the
previous process [16] (Table 1).
3.3 Dubin Curves
The Dubin curves describe six types of trajectories: RSR, LSR, RSL, LSL, RLR, and
LRL. Each configuration comes from an analogy that is denoted by R (right move),
S (straight move), and L (left move) [17].
All these configurations use geometrically computing method based on construct-
ing tangent lines between two circles. The first step has two circles C1 and C2 , with
their respective radius r1 and r2 , where C1 represents coordinates (x1 , y1 ) and C2 as
(x2 , y2 ) (see in Fig. 2).
Fig. 2 Circles initial

configuration
Fig. 3 Inner tangents
Inner tangents. Then a line is drawn between two center points C1 and C2 estab-
−
→
lishing a vector V1 , magnitude D and the midpoint p3 (point between p1 and p2 ) is
calculated, circle C3 is constructed with a radius r3 as we reflected in Fig. 3

D= (x2 − x1 )2 + (y2 − y1 )2 (1)

x1 + x2 y1 + y2
p3 = , (2)
2 2
D
r3 = (3)
2
The next step is to draw another circle C4 located in C1 ’s center, with radius
r4 = r1 + r2 , we obtained pt , which is the intersection between C4 and C3 like
the one shown in Fig. 3. A triangle is built joining the points p1 , p3 , pt and we can
define geometrically that segment pt p1 = r4 and p1 p3 = pt p3 = r3 . The angle
γ = ∠ pt p1 p3 is very important to define the coordinates of pt . The next equation
−
→
determinates the amount of rotation about the x-axis, θ for V2 .
→
−
θ = γ + atan2 V1 (4)
Using θ , it is possible to obtain pt according to the following equations:
xt = x1 + (r1 + r2 ) ∗ cos(θ ) (5)
yt = y1 + (r1 + r2 ) ∗ sin(θ ) (6)
Considering that the inner tangent starts on C1 , it is necessary to normalize a

−
→ −
→
vector V2 = ( pt − p1 ) and multiply it by r1 , the result will allow to find a vector V3
to pit1 from p1 . It resumes next by:
Fig. 4 RRT* algorithm’s

visualization
−
→
−
→ V2
V3 = ∗ r1 (7)
V2
−
→
pit1 = p1 + V3 (8)
−
→
Finally, it is possible to draw a vector V4 from pt to p2 , as shown in the figure.
−
→
Using V4 magnitude and the direction, it is possible to find the inner tangent point
on C2 .
−
→
V4 = ( p2 − pt ) (9)
−
→
pit2 = pit1 + V4 (10)
Outer tangents. The process is very similar to the one of inner tangents, having
two circles C1 and C2 , and considering r1 ≥ r2 , the procedure is the same as before,
C4 is centered at p1 , with a difference the radius r4 = r1 − r2 , after getting pt and
−
→
following all steps performed for the interior tangents, V2 is obtained and the first
outer tangent point pot1 . This condition produces that r4 < r1 . To get the second
outer tangent pot2 an addition is performed by:
−
→
pot2 = pot1 + V4 (11)
The main difference between calculating outer tangents compared to inner tan-
gents is the construction of circle C4 ; all steps keep the same.
In the next figure (see Fig. 4), it can be seen the path establishes using data input.
4 Results
The experimental results produced by RRT* path planning based on Dubins curves
were performed using a CPU with an i7 third generation processor and 16 Gb RAM
memory. For algorithm testing, a scene with a necessary amount of light was chosen
to obtain the correct object segmentation. From this point, the test started with no
obstacles in the scene. Then, obstacles were gradually added to establish four different
configurations, in such way, algorithm functionality and efficiency was tested.
The algorithm was tested in five different cases (see Fig. 5), with no obstacles
(first case), one obstacle (second case), two obstacles (third case), three obstacles
(fourth case), and four obstacles (fifth case).
The algorithm satisfies its purpose because the path established (red line) started
from car’s position to the final configuration (blue circle) avoiding the obstacles.
Fig. 5 Scene configurations. a No obstacles. b One obstacle. c Two obstacles. d Three obstacles.
e Four obstacles
150
Nº Obstacles vs Input data 150
nº itera ons
y = -4.6429x2 + 44.357x + 3
step size 100 R² = 0.9756 100
Step size
Poly. (nº 50 50
itera ons) y = 8.5714x2 - 67.429x + 196
Poly. (step R² = 0.9752
size)
0 0
0 1 2 3 4
nº itera ons 40 80 90 105 110
step size 140 90 70 70 70
nº Obstacles
Fig. 6 Number of iterations and step size processed for each obstacle configuration
Another aspect to consider was that the goal position should have a tolerance radius
of 60 pixels measured from the goal position’s center point.
The first measurement of results (see Fig. 6) showed that the cases in which
obstacles were added, it was necessary to increase the number of iterations to reach
the configuration desired. Forty iterations were the minimum to reach the goal in the
first case, compared with the fifth case where it was necessary at least 110 iterations
to establish a path for UGV to represent.
Based on the data processed, a second-grade polynomial equation was obtained
to represent the number of iterations required for a scene where there are more than
four obstacles in a scene.
y = −4, 6x 2 + 44, 3x + 3 (12)
The ideal value for step size for each case is decreasing if we take into considera-
tion the first test is with no obstacles and the fifth test has four obstacles in the scene.
One hundred forty pixels the maximum step size for the optimal performance of the
algorithm within the limits of workspace and for UGV to follow the path described.
In the last three cases (case 3, case 4, and case 5), the necessary step size is 70 pixels,
although the correlation of data is high by number of tests, the best possible solution
in case the number of obstacles increase would be tested with the same value of step
size.
Processing time required by the algorithm compared with the number of obstacles
is presented in (Fig. 7). The blue line (time 1) describes the necessary time for
expanding RRT* using an optimal data input for each case. The red one (time 2)
describes the time required to reach de goal destination if the data is fixed with
values of 30 pixels for step size and the maximum number of iterations is 1500.
Fig. 7 Processing time

required by algorithm
Nº Obstacles vs Time
100
Time
50
0
0 1 2 3 4
Ɵme 1 2 2 5 5 5
Ɵme 2 6 7 66 70 62
Nº Obstacles
5 Conclusions
From the data aforementioned in this paper, we can conclude that exists a direct
relationship between several iterations, step size, and quantity of obstacles. In case
with few obstacles within the workspace, we can configure the step size in maximum
value of 140 pixels allowing to reduce the number of times that mechanical direction
of UGV has to rotate. The number of iterations is directly proportional to the number
of obstacles because, more obstacles in the scene require to construct more iterations
to reach the goal with an optimal path.
Establishing constant input data was not the most efficient way to execute the
algorithm because if the step size configuration is too low, near 30 pixels, so the
UGV can take an excessive time to follow the path and require too many changes in
the direction of the robot.
This study has determined the importance of configuring initial parameters to
construct the RRT* algorithm. Considering the path planning is based in a global
planner, it is important to get the parameters in lesser time and optimal manner
possible to obtain a better routing of data between UGV and UAV. Also, the study
is the first stage of a global research for developing a collaborative robot system; we
focused on getting a correct path for a mobile robot to navigate without colliding with
objects around it. Finally, this work could be complemented with the development
of pure pursuit control for UGV.
References
1. Yulong, D., Bin, X., Jie, C., Hao, F., Yangguang, Z., Guanqiang, G., Lihua, D.: Path planning
of messenger UAV in air-ground coordination. IFAC-PapersOnLine 50, 8045–8051 (2017)
2. Sivaneri, V.O., Gross, Y.J.N.: UGV-to-UAV cooperative ranging for robust navigation in GNSS-
challenged environments. Aerosp. Sci. Technol. 71, 245–255, 2017
3. Rahimi, R., Abdollahi, F., Naqshi, K.: Time-varying formation control of a collaborative het-
erogeneous multi agent system. Robot. Auton. Syst. 62, 1799–1805 (2014)
4. Melin, J., Lauri, M., Kolu, A., Koijonen, J., Ritala, R.: Cooperative sensing and path planning
in a multi-vehicle environment. IFAC-PapersOnLine 198–203 (2015)
5. Rosa, L., Cognetti, M., Nicastro, A., Alvarez, P., Oriolo, G.: Multi-task cooperative control in
a heterogeneous ground-air robot team. IFAC-PapersOnLine 48, 53–58 (2015)
6. Ropero, F., Muñoz, P., Moreno, M.D.R.: TERRA: a path planning algorithm for cooperative
UGV-UAV exploration. Eng. Appl. Artif. Intell. (2019)
7. LaValle, S.: The RRT Page (1999). [En línea]. http://msl.cs.uiuc.edu/rrt/index.html
8. Abbadi, A., Prenosil, V.: Collided path replanning in dynamic environments using RRT and
cell decomposition algorithms. Modelling and Simulation for Autonomous Systems: Second
International Workshop, pp. 131–143 (2015)
9. Sieberth, T., Wackrow, R., Chandler, J.H.: Automatic detection of blurred images in UAV image
sets. ISPRS J. Photogramm. Remote Sens. 122, 1–16 (2016)
10. Chernov, V., Alander, J., Bochko, V.: Integer-based accurate conversion between RGB and
HSV color spaces. Comput. Electr. Eng. vol. 46, pp. 328–337, 2015
11. Huang, J., Feng, H., Xu, Z., Li, Q., Chen, Y.: A robust deblurring algorithm for noisy images
with just noticeable blur. Optik 168, 577–589 (2018)
12. Chen, S., Li, D.: Image binarization focusing on objects. Neurocomputing 69, 2411–2415
(2006)
13. Dougherty, E.: Mathematical Morphology in image processing. Marcel Dekker, Nueva York
(1993)
14. OpenCV.: OpenCV 2.4.13.7 documentation. [En línea]. https://docs.opencv.org/2.4/modules/
imgproc/doc/filtering.html
15. Noreen, I., Khan, A., Habib, Z.: Comparison of RRT, RTT* and RRT*-smart path planning
algorithms. IJCSNS Int. J. Comput. Sci. Netw. Secur. (2016)
16. Karaman, S., Frazzoli, E.: Sampling-based algorithms for optimal. Int. J. Robot. Res. (2011)
17. Yao, W., Qi, N., Zhao, J., Wan, N.: Bounded curvature path planning with expected length for
Dubins vehicle entering target manifold. Robot. Auton. Syst. 97, 217–229 (2017)
Machine Learning and Multipath
Fingerprints for Emitter Localization
in Urban Scenario
Marcelo N. de Sousa, Rafael L. Cardoso, Henrique S. Melo,

José W. C. Parente Jr. and Reiner S. Thomä
Abstract A hybrid approach is proposed to perform the locate and tracking a non-
collaborative radio frequency emitter using ray tracing (RT) simulation tool, chan-
nel impulse response (CIR), and machine learning estimation. The technique can
enhance communication intelligence (COMINT) systems or even perform the local-
ization using a single sensor in an non-line-of-sight (NLOS) suburban scenario. A
multipath fingerprint can identify the target position using the machine learning
classification engine to perform the matching. Conventional localization techniques
mitigate errors trying to avoid NLOS measurements in processing emitter position,
while the multipath fingerprints proposed uses the reflection information to feed the
pattern matching engine build on a machine learning classification framework. The
method was applied to simulate a tactical scenario, where a navy frigate is in Ipanema
and tries to track an RF emitter target in the Rio de Janeiro streets using only one RF
sensors fixed in an aerostat in a hypothetical counterinsurgency situation.
Keywords Wireless positioning · Multipath exploitation · Hybrid positioning ·

Machine learning · Ray tracing simulation
1 Introduction
The estimation of the source of a signals’ location remains the subject of extensive
research for years and is increasingly receiving considerable interest in signal pro-
cessing scholarly community, including radar, GPS, wireless sensor networks, sonar,
animal tracking, multimedia, and mobile communications. In urban and counterin-
surgency operations, there is an increasing demand to have an information superiority
M. N. de Sousa (B) · R. S. Thomä

Technische Universität Ilmenau, 98684 Ilmenau, TH, Germany
e-mail: marcelo.nogueira@tu-ilmenau.de
URL: http://www.tu-ilmenau.de/it-ems/
R. L. Cardoso · H. S. Melo · J. W. C. Parente Jr.
Military Institute of Engineering, Rio de Janeiro 22290-270, Brazil

https://doi.org/10.1007/978-981-13-9155-2_18
218 M. N. de Sousa et al.
of the tactical scenario, where the communication intelligence systems (COMINT)

need to supply the commanders in different levels with the precise information of
the target position [1].
Multipath is typically perceived as a limitation, [2], but is possible to exploit
interaction effects for localization of the RF emitter using a single sensor. There is
always a trade-off between the performance desired and the available resources in
deploying the number of sensors needed. It is also worthy to use the scenario features
converted by ray tracing simulation to improve the performance of the available
sensors even in NLOS condition. There are various approaches [3] tracking and
emitter position finding where multipath exploitation is performed to locate the target,
again using the image-based approach to identify the sensor position end extract the
location with LoS and NLOS calculation [4].
We propose an approach to perform a RF localization in urban scenario binding
together a ray tracing propagation tool to extract the multipath fingerprint and a
machine learning framework to improve the precision even in NLOS situation. The
idea is to reproduce the framework presented in [5] using only one sensor, contrarily
to similar approaches, where the NLOS measurements are not used in the position
estimation. This approach is relevant to enhance the radio frequency localization
performance in an urban environment because it uses NLOS and geographic database
as an extra layer of information, that can be fused in the position estimation engine.
This paper has the following structure. Section 2 discusses the problem formula-
tion, including the signal data model and the channel impulse response estimation.
An experimental setup, showing the application of the proposed localization method
and including the results of the simulation setup, is discussed in Sect. 3. Finally,
Sect. 4 summarizes the principal aspects of this paper and presents possible research
direction using the proposed methodology.
2 Problem Formulation
Assuming the typical scenario displayed in Fig. 1, Where a frigate if the navy is
placed close to a city cost to support some counterinsurgency operations, the ship
has an RF sensor with direction finding capability, and the sensors are installed in an
aerostat connected to the frigate flying at 100 meters over the sea level. A schematic
view of the sensor, the platform, and the navy frigate is shown in Fig. 2. To map the
reflection points in the environment as virtual sensors, assume a “non-collaborative”
emitter that has been tracked inside the city, the buildings, and the urban scenario
makes it impossible to have a clear line-of-site situation to locate the emitter (Fig. 3).
Using the approach introduced by [6], modified to perform the calculation using
the sensor S1RT and M scatters points at location Ri = [xi , yi , z i ]T . If the scatters’
position are known, it is possible to locate the RF target P using an adapted TDOA
measurements scheme introduced by [7].
The three most significant scatters that affect the signal received S1RT will be
considered as three others TDOA sensor, namely as “Virtual Sensors.” The position of
Machine Learning and Multipath Fingerprints for Emitter Localization … 219
Fig. 1 Schematic view using Google Earth of an emitter in a urban scenario (Ipanema-RJ) tracked
by a navy vessel with a RF sensor at 100 m
Fig. 2 Sensor on aerostat platform in navy vessel at 100 m

Fig. 3 Location error caused by discretization
the sensors, the scatters, and the emitter presents an error, caused by the discretization
of the scenario. As the error also affects the precise reflection points’ information,
the virtual nodes modeled as virtual sensors position also presents a grid error.
The sensor position will be assumed to be TDOA sensor 1, and the three most
significant scatters that affect the signal received S1RT will be considered as three
others TDOA sensor, namely as “Virtual Sensors.” The emitter source is in position
P = [xs , ys , z s ]T ∈ R3 and the TDOA sensors, Si = [Si x , Si y , Si z ]i=1:4 at known
position in 3D coordinate system, as shown in Fig. 4.
Θ̂ = [τ21 , τ31 , τ41 ]T (1)
As shown in Fig. 5, the channel impulse response presents a diffracted ray (α1 , τ1 ),
and three scatters that where reflected in R2 ,R3 , and R4 , in this situation we have the
time of propagation from emitter and the sensor, and the time of propagation from
the sensor to each reflection wall, as follows:
τ21 = a + b − τ1 ;
τ31 = c + d − τ1 ; (2)
τ41 = e + f − τ1 .
Fig. 4 Scatters points as virtual TDOA sensors
The Figs. 4 and 5 can illustrate the differences of the RT virtual nodes data model
with the normal TDOA system. The sensor position and the reflection points are
assumed to be known under certain imprecision. The time differences from the sensor
base (S1RT ) to the other TDOA sensor in Fig. 4 or to the reflection points in Fig. 5, a,
c, e are also known. So, the central question is how to estimate the time differences
and perform the position estimation with this data model of [7]. Assuming that it
is possible to perform the CIR estimation and assign to each ray to each reflection
point in the scenario, these points can work as virtual TDOA sensors. The error in the
reflections points will be assumed as an error in the definition of the station position.
The TDOA computation between the signals received by the sensors is computed
several; first, the signal is filtered (“whitening”) and oversampled using Farrow filters
[8] and generalized cross- correlation function in phase(GCC-PHAT).
3 Proposed Method
This section describes the method proposed to enhance a TDOA localization system
when the emitter is in NLOS position, that is a typical situation of the outdoor channel
characteristics and cannot be avoided. Some methods deal with multipath assuming
Fig. 5 Location using four TDOA sensors
that they are outliers, but this approach is not feasible in a typical suburban scenario.
The proposed method is a combination of channel impulse response estimation and
ray tracing. The first step is to discover the strong multipath specular components
that affect the time measurements, this is performed doing the CIR estimation at
sensor position, after that, ones try to identify the main obstacles in the environment,
that produces the multipath rays.
The multipath parameters, defining the amplitude, time of arrival, and directions
of the incoming ray at the sensor position are calculated. The decision to use the
reflection points as virtual TDOA sensor, instead of using only the direct range
estimation was made because there is no synchronization between the emitter and
the receiver. As shown in Fig. 6, the buildings in the scenario cause reflections and
diffraction in the incoming signal at the sensor.
The method is schematically presented in Fig. 6, first, the ray tracing will perform
several simulations of the outdoor scenario, the output of the simulation is a descrip-
tion of the paths from each component of the signal that connects the emitter position
with the receiver point. The second step is to estimate the CIR to extract the multi-
path information (α1 , τi ) from the receiver signal in each sensor position. In the end,
a neural-network-based on machine learning engine performs the estimation of the
emitter position, enhancing the TDOA performance in the outdoor NLOS situations.
Depending on the precision of the obstacles description and the material properties
Fig. 6 Channel impulse estimation in RF sensor
Fig. 7 Fingerprints using ray tracing
included in the simulation, the ray tracing gives a reasonable representation of the
specular components that characterize the emitter–receiver relationship (Fig. 7).
The next step is to take the signal received by each sensor and make the channel
impulse response (CIR), where the information (α1 , τi ) extracted from the complex
baseband (CBB) received signal is associated with a defined position in the scenario.
The matching procedure between the CIR estimated, and the multipath dataset is
performed using a neural-network-based with a machine learning engine. The reflec-

tion point used as virtual sensors has some uncertainty not precisely described, as
described in [9].
3.1 Ray Tracing Fingerprints
When a ray tracing simulation is performed, there is “preprocessing” information,

which extracts the emitter–receiver visibility. Depending on the desired number of
ray interaction, the image theory allows to identify the reflection points and the
virtual nodes, and all possible rays in the simulation domain can be estimated. Then,
for each Tx-Rx combination, there is a specifically channel impulse response that
defines the time delays for the specular components. Considering that the sensor 1
can extract the angular channel impulse response where the four most robust specular
components are identified and measured in angular and time information. With the
output file of ray tracing, it is possible to search for a given position where are the
main reflectors that the rays bounce before arriving at sensor. It works as if, ones
want to extract the “Field of Vision” in optics ray.
So, in the case of Fig. 8, after this “searching” phase three buildings were found,
and the reflection points are P1, P2, and P3. Now, with the information of the time
difference between the sensor position and each reflection point, the time difference
between the direct rays and the reflected ones are taken from the CIR estimation.
In the end, this configuration looks like a TDOA system where the reflection points
Fig. 8 Wall and edges form the buildings in ray tracing simulation
Fig. 9 Single station estimation location
work as virtual sensors and the position estimation engine uses the time differences
as a modified TDOA system with only one station.
Following the approach of [10], the RT gives the information about the ray path,
describing the edges and wall touched by the rays in the emitter–sensor path. The
Fig. 8 gives a graphic representation of the visibility matrix calculation. Fig. 9 shows
the multipath fingerprints produced by ray trace simulation from sensor in the aero-
stat.
3.2 Machine Learning in Multipath Fingerprint
As described by [11], the machine learning (ML) tool is used to match the multipath
information to predict geolocation of the emitter. It is possible to use the regression
for prediction because it is a class of supervised algorithm that attempts to estab-
lish a continuous relationship between a set of dependent variables (geolocation
coordinates) and set of other independent variables (multipath fingerprints—αi , τi ).
Figs. 10 and 11 show the machine learning framework proposed.
First, we will try to use ML to approximate the position estimation as a target
function for optimization, the semi-supervised or supervised ML makes an approx-
imation to find a target function ( f ). This function should be able to map input
variables (X ), that are the multipath components (αi , τi ), to an output variable (Y ),
that is the emitter position. We used a set of tools as a machine learning framework
to develop and test the application, including:
Fig. 10 Machine learning ray tracing dataset refinement
Fig. 11 Supervised machine learning based on RT simulation
– Jupyter Notebook version 5.1.0 (http://jupyter.org);

– Tensorflow version 1.4.0 (http://tensorflow.org);
– CentOS version 7.4.1708 (https://www.centos.org).
The machine learning and kernel methods can work as a supervised, unsupervised,
or semi-supervised learning framework. In our approach, the ray tracing gives the
initial dataset for the ML and the CIR estimation of the reference or anchor nodes
multipath works as a “reinforcement” or supervised learning. The hybrid approach
uses real measurements and simulations to perform the dataset, which can be classi-
Table 1 Emitter position with single station
Pseudo-code: Emitter Position with Single Station

1: Input:
– Sensor Position (X, Y, Z);
– Geo database with Buildings Description (Walls and Edges);
– Complex Base Band Signal at Sensor 1;
– Direction Finding with with 4 Strongest Multipath Info.
2: Perform Ray-Tracing Simulation:
– from all points from “AoI” and Sensor 1;
– Extract the “Visibility”, Edges and Wall with “mutual vision”;
– Map RT simulation to channel estimation to calculate
reflection points.
3: From the CBB Signal Received at S1 Perform CIR Estimation;
– Assign the First Obstacle in the DoA from 4 strongest Rays;
– Identify the Edges and Wall from the Obstacles.
4: Assign the Reflection Points a TDoA Sensor Position (X, Y, Z).
5: With ToA info and Reflection Position:
Perfom TDoA Localization.
6: Compare the Localization with Virtual TDoA Sensor
7: Output: Emitter Estimation Position.
fied as semi-supervised learning. The network had ten inputs, i.e., five rays. Neural
networks, linear regression, and random forest algorithm were used to achieve the
result, and the random forest proved to be the best one.
One of the problems in the model proposed was the points that have less than
five paths. Then a set of hyper-parameters tuned the linear regression was used,
giving better results. In our formulation, we use the five strongest rays that arrive at
each point of the simulation domain. The proposed algorithm can be summarized in
Table 1.
4 Experiment Setup
The algorithm was applied to perform the position estimation in the region of
Ipanema, as described in Sect. 2, the target is inside an urban area, and the navy
vessel is trying to perform the localization using only one sensor and the ray tracing
fingerprint framework presented in Sect. 3 using just the sensor 1 and the ray tracing
simulation of the area of interest (AOI). The result was compared with a simulation
using the estimation performed by a complete TDOA systems using 4 (four) base
station nodes (S1 , S2 , S3 , and S4 ), deployed in the ocean in front of the AOI.
First, the complete scenario was reproduced in ray tracing simulation tools to
extract the multipath information, then, the most significant reflectors, considering
the TDOA sensor 1 was obtained from the RT output files. The CIR estimation proce-
dure was performed using a signal synthetic generated using a binary pseudo-random
sequence with length 15 (PBRS15). Several averaging and filtering operations were
performed to reproduce the effects of the urban multipath channel on the received
signal S1RT . The CIR estimated, in the time domain, gave the delay and power profile
of the known position building. Then the CIR estimated from the target with up
to three rays (delay and power) was used to search in the fingerprint database the
position of the emitter.
The fingerprints of the rays suing the machine learning framework are presented
in the results of the application the method are presented in Fig. 12. The Table 2
summarizes the result of the method using three and four multipath components, the
error is caused by the scatter position imprecision in the discretization of the scenario.
It is also important to point out that the multipath components are first filtered out
to taking into account only the first bounce reflections, and the ray association using
the real measurements was possible because the emitter stayed in a static position
Table 2 Performance of the single station TDOA location position compared with a complete
TDOA systems
Method Obs Lat Long Error (m)
TDoA 4 [tau 31 , tau 31 , tau 31 ] 49.591715 11.085821 5
sensor
Qualities [84 57 65]%
Duration 20 ms
Power [−100 −99 −101 −95] dBm
Delays [1.064 1.052 0.471] µs
Sol 1 S1 + 3 Paths at R1, R2 and R3 49.592382 11.085037 27
Sol 2 S1 + 2 Paths at R1 and R2 49.592379 11.088463 82
Fig. 12 Fingerprints from the first rays amplitude and delay in a urban scenario (Ipanema-RJ)
Fig. 13 Fingerprints in 3D from the first rays amplitude and delay in a urban scenario (Ipanema-RJ)
during the measurements done. The TDOA location ellipses are shown in Fig. 13, and
the total standard deviation of the position estimator was 89.66 m, and the standard
deviation of each component, using the Chan method was: σ̂ [σx = 28.40 m, σ y =
33.05 m, σz = 78.36 m].
5 Conclusions
The paper proposes a kernel-based machine learning localization scheme based on

TDOA fingerprinting. It describes how to use ML for refining the multipath Finger-
print generated by ray tracing simulation. The method proposed showed the ability
for enhancing the performance of sensors deployed in NLOS scenario. The simu-
lations performed using the single sensor location with the ray tracing technique
could perform the emitter position estimation. Despite the bias and imprecision of
the scenario description, the main obstacles that reflect the signal can be used as
virtual sensors to improve the TDOA system performance.
The effect in localization procedure is not straightforward to analyze, and the ML
engine and ray tracing are two tools that only give reasonable results depending on
the quality of the representations of the obstacles in a geographic database. The ML
can improve the multipath fingerprint generated by ray tracing, but it also depends on
the capability of processing the channel impulse estimation and get an appropriate
multipath characterization of the calibration point.
For the next steps in the research, the authors would like to consider the use of
more features either the signal and the scenario, trying to record more multipath
information and patterns to build a data fusion engine with the cartographic database
and signal processing. It is necessary to make a mapping procedure to assign the

reflected ray from the measurements into the ray tracing. The static scenario and the
specific characteristic of the buildings make the multipath components “well defined”
with the specular components much bigger them the dense multipath components
(DMC).
References
1. Brady, E., Starr, S.: Assessing C3I in support of dismounted operations in complex terrain.
Technical Report, Strategic Perspectives Inc., McLean, VA (2002)
2. Phelan, B.R., Lenzing, E.H., Narayanan, R.M.: Source localization using unique characteriza-
tions of multipath propagation in an urban environment. In: 2012 IEEE 7th Sensor Array and
Multichannel Signal Processing Workshop (SAM), pp. 189–192, June 2012
3. Fertig, L.B., Baden, M.J., Kerce, J.C., Sobota, D.: Localization and tracking with multipath
exploitation radar. In: 2012 IEEE Radar Conference, pp. 1014–1018, May 2012
4. Setlur, P., Devroye, N.: Bayesian and Cramer-Rao bounds for single sensor target localization
via multipath exploitation. In: 2013 IEEE International Conference on Acoustics, Speech and
Signal Processing, pp. 5845–5849, May 2013
5. de Sousa, M.N., Thomä, R.S.: Enhancement of localization systems in NLOS urban scenario
with multipath ray tracing fingerprints and machine learning. Sensors 18(11), 4073 (2018)
6. So, H.C.: Handbook of position location: theory, practice and advances. In: Reza Zekavat,
R.M.B. (ed.) The Oxford Handbook of Innovation, pp. 23–34. Wiley-IEEE Press, ch. 2 (2011)
7. Ho, K.C., Lu, X., Kovavisaruch, L.: Source localization using TDOA and FDOA measurements
in the presence of receiver location errors: analysis and solution. IEEE Trans. Signal Process.
55(2), 684–696 (2007)
8. Kolb, D., Warzügel, S., Schramm, J.: Hardware-reduced system for TDOA-locating of radio
frequency emitters, eP Patent App. EP20,130,160,415, 25 Sept 2013. http://www.google.com/
patents/EP2642312A1?cl=en
9. Yang, L., Ho, K.C.: An approximately efficient TDOA localization algorithm in closed-form for
locating multiple disjoint sources with erroneous sensor positions. IEEE Trans. Signal Process.
57(12), 4598–4615 (2009)
10. Fuschini, F., El-Sallabi, H., Degli-Esposti, V., Vuokko, L., Guiducci, D., Vainikainen, P.: Anal-
ysis of multipath propagation in urban environment through multidimensional measurements
and advanced ray tracing simulation. IEEE Trans. Antennas Propag. 56(3), 848–857 (2008)
11. Alsheikh, M.A., Lin, S., Niyato, D., Tan, H.P.: Machine learning in wireless sensor networks:
algorithms, strategies, and applications. IEEE Commun. Surv. Tutor. 16(4), 1996–2018 (2014)
(Fourthquarter)
Virtual Rehabilitation System Using
Electromyographic Sensors
for Strengthening Upper Extremities
Z. Andrea Sánchez, T. Santiago Alvarez, F. Roberto Segura,

C. Tomás Núñez, P. Urrutia-Urrutia, L. Franklin Salazar , S. Altamirano
and J. Buele
Abstract This work presents a virtual system for the rehabilitation of the upper
extremities, using the MYO Smart Band device for the acquisition of electromyo-
graphic signals produced by the user. Processing and managing of these signals are
done through the SDK provided by the manufacturer of the bracelet which is com-
patible with the MATLAB software. The virtual environment is developed in the
Unity 3D graphics engine, in which three-dimensional objects that were previously
designed in the 3ds Max software are implemented. The application presents the
user with a virtual scenario set in a natural landscape, in which there is a van that
must be driven on a certain path (the complexity is increasing). The videogame is
of low complexity, since it seeks to avoid situations of stress while the rehabilita-
tion process takes place. Each task in the application is associated with a hand and
forearm movement of the user; it means the patient is given an alternative tool that
Z. Andrea Sánchez · P. Urrutia-Urrutia · L. Franklin Salazar · S. Altamirano · J. Buele (B)

Universidad Técnica de Ambato, Ambato 180103, Ecuador
e-mail: jl.buele@uta.edu.ec
Z. Andrea Sánchez
e-mail: ap.sanchez@uta.edu.ec
P. Urrutia-Urrutia
e-mail: elsapurrutia@uta.edu.ec
L. Franklin Salazar
e-mail: fw.salazar@uta.edu.ec
S. Altamirano
e-mail: santiagomaltamirano@uta.edu.ec
T. Santiago Alvarez · F. Roberto Segura
Instituto Tecnológico Superior Guayaquil - Ambato, Ambato 180205, Ecuador
e-mail: salvarez@institutos.gob.ec
F. Roberto Segura
e-mail: rsegura@institutos.gob.ec
C. Tomás Núñez
CELEC EP, Baños 180250, Ecuador
e-mail: leandro.nunez@celec.gob.ec

https://doi.org/10.1007/978-981-13-9155-2_19
232 Z. Andrea Sánchez et al.
allows him/her to perform exercises that improve his/her extremity active mobility,
mitigating the routine effects of a conventional session. To validate this proposal, it
is tested by five retired military personnel in passive state, to whom the using task
ease (SEQ) usability test is applied. The result is (58,8 ± 0,27), which shows that
this interactive interface has a good acceptance when being in the range between 40
and 65.
Keywords Rehabilitation · Virtual reality · Upper extremities ·

Electromyographic sensor
1 Introduction
Military confrontations and wars are considered extreme and challenging situations
that a human being must endure [1]. It is subjected to a multitude of psychomo-
tor, sensory, cognitive, and psychological tests that exert great pressure on military
personnel [2]. Some get to lose their senses or develop some type of trauma or dis-
ability [3]. Among the main intellectual/cognitive disabilities are learning disorders,
generalized developmental disorders, attention deficit disorders and behavior [4]. In
the sensory field, there are visual, auditory, language, speech, and voice disabilities
[2, 3]. Additionally, in the physical aspect, reference is made to motor alterations
at osteoarticular level, nervous system, amputations, and disability in the extremi-
ties [5]. The most common injuries occur in both the lower and upper extremities
[6, 7]. In the upper ones, there is ankylosis, joint stiffness, sequelae of fractures or
traumatisms and complete paralysis of fingers or hands. That is why, with the aim
of preventing or mitigating any of these illnesses, a physical therapy process must
be initiated [8]. Through a cycle of relearning promote and stimulate muscles and
joints to regain their functionality [9].
The rehabilitation process must be evaluated by a qualified physiotherapist who
provides the patient with the tools to accelerate this process. Previously, routine
coordinated physical exercises were developed to produced boredom and fatigue
during the sessions [10]. But this reality has changed when merging these procedures
with advanced technological tools. A clear example of this is the use of compatible
devices with virtual reality (VR) interfaces and auditory accompaniment. Systems
involving VR are an advanced form of human–machine interaction (HCI), which
allows the user to experience extraordinary situations, without the need to move to
another site. These systems achieve a higher level of immersion with the proper
use of sensors, devices, and peripherals that enrich the experience. All this stuff,
focused on the rehabilitation processes, gives the patient the opportunity to perform
exercises in a different way and with greater encouragement [8]. Especially when
it comes to military personnel who have been subjected to stress and have certain
post-war psychological disorders.
Virtual Rehabilitation System Using Electromyographic Sensors … 233
2 State of Art
The implementation of systems that include VR in their architecture for rehabilitation

processes has been gaining relevance in recent years and that is why several related
researches have been developed. A treatment focused on counteracting the incidence
of post-traumatic stress disorder (PTSD) of military personnel using VR is described
in the proposal developed by Achanccaray et al. [11]. A virtual environment inspired
by Iraq/Afghanistan has been developed and includes the dangers found in these
places. This proposal has provided initial reports with positive results, when tested
by active staff and veterans.
Regarding upper extremity treatment, the development of new sensors and devices
allows the development of several options when it comes to rehabilitation. In the prior
research carried out by Levin et al. [12], VR is presented as a tool used in rehabilita-
tion processes of upper extremities. Virtual environments that optimize learning and
motor skills recovery are managed. Based on the respective experimental tests, the
limitations of current technologies are discussed with respect to their effectiveness
and feedback in the process of physical and sensory learning. For its part, Liu et al.
[13] present a rehabilitation system of upper extremities based on electromyographic
sensors (EMG) and portable accelerometers (ACC) for children with cerebral palsy
(CP) which is explained. Using the Android platform, games are developed that seek
to improve motor function under the guidance of a doctor. The proposal has been
validated by three test subjects who have improved their reaction capacity by inter-
acting with this motivating and user-friendly interface. Similarly, López et al. [14]
present an interactive system that uses myoelectric sensors to strengthen upper limbs
in children. Unity 3D software develops interactive interfaces (games) that promote
the realization of therapies by patients. This prototype is tested by 5 users (3 boys
and 2 girls) with ages between 6 and 12 years where respective System Usability
Scale (SUS) usability test is presented.
As can be seen, there are many researches related to this topic, but not focused
on the military field as required. In this context, this work proposes a virtual reha-
bilitation system for upper extremities that interacts intuitively with the patient, in
this specific case of military personnel who suffered a combat injury. Data acqui-
sition for the interface configuration as well as the reading of electromyographic
signals is done using a low-cost electronic device with which the physical gesture
that the person is making can be recognized. Three-dimensional objects have been
made in 3ds Max software and imported into the Unity 3D software, where they are
given hierarchy and animation. In this context, the aim is to present the patient with
an interactive tool that contributes to a comfortable and accelerated rehabilitation
process, mitigating the post-traumatic stress disorders that they develop after their
exposure to stress situations.
This work is composed of five sections, including the introduction in Sect. 1. The
works related to the subject are described in Sect. 2 and the methodology used in
Sect. 3. In Sect. 4 the tests and the results obtained are presented and finally, in Sect. 5
the conclusions and future work.
3 Methodology
This section details the proposed system and the elements that comprise it. The
general diagram is described in Fig. 1.
3.1 I/O Device Configuration
The MYO bracelet is used as the input device and the HTC VIVE virtual reality
glasses and binaural headphones as output devices. Next, the respective configuration
of these hardware elements is presented.
MYO Smart Band. To perform the digital processing of acquired signals, this device
is connected to the MATLAB software through the Myo SDK MATLAB MEX
Wrapper library. Through the SDK, information can be accessed provided from the
eight electromyographic sensors and the inertial measurement unit (IMU). Among
the most relevant data are those obtained from the three-axis gyroscope (angular
velocity) and the three-axis accelerometer (linear acceleration).
HTC VIVE. The connection of this device (virtual reality glasses) and the environ-
ment is generated natively in Unity 3D thanks to the SteamVR plugin. In addition,
reorientation and movement scripts have been made due to the fact that HTC con-
trollers are not used. In this device, the teleportation function is activated by default
in standard mode, which is replaced by signals from the Myo bracelet.
Binaural Headphones. They allow to reproduce the spatial sound that is generated
in the virtual scene and in which the user can spatially identify the origin of the
sound.
Fig. 1 General diagram of the implemented system

3.2 Signal Acquisition
The signal acquisition process is performed based on the flow diagram described in
Fig. 2. The Myo Armband device from Thalmic Labs integrates a set of eight nonin-
vasive sensors, both electromyographic (EMG) and inertial. All sensors collect infor-
mation that is processed in a central module (sensor number 4), which includes batter-
ies and a communication device based on Bluetooth. Through libraries developed in
C++, it has been possible to use it without restrictions letting know the developer
certain parameters such as the measurement of each one of the electromyographic
sensors, processed and unprocessed information of the inertial sensors, transmission
periods, the gesture made by the patient, the values of accelerometers and gyro-
scopes, properties to eliminate the connected device, transmission rate, among oth-
ers. Figure 3 shows a capture of the eight sensor’s signals in a 10 s period executing
the open hand gesture. The amplitude units of each of the sensor signals are denoted
here as maximum factory value (MFV), a dimensionless value used by the developers
to represent the minimum and maximum amplitudes (−1 to 1) that the signal can
have, given the internal preprocesses and validators (within the SDK).
Fig. 2 Flowchart of the algorithm used for signal acquisition

Fig. 3 EMG signal acquisition in 10 s of the gesture: open hand
3.3 Interface Development
The interface design is made based on the flowchart of the interaction of programs,
which also includes the configuration in MATLAB and the database presented in
Fig. 4. Using Unity 3D software, a virtual environment has been developed, in which
objects that simulate an open field with a road across it are placed. The aim of
this application is to drive a van-type vehicle and completing a trajectory that is
previously defined by itself. Tasks have been carried out in a simple way so as not
to produce a stressful situation in patients, but instead they are entertained while
performing their rehabilitation exercises. Script programming allows to identify five
gestures that patients can perform: (i) fist, (ii) open hand, (iii) palm inward, (iv) palm
outward, and (v) join thumb and middle fingers.
The three-dimensional objects’ design has been made in 3ds Max software. Once
all the necessary objects have been modeled in 3D, their compatibility is checked, the
rotation points are located, hierarchies are established, and the axes of each element
are oriented. After that, models are imported from 3ds Max using a format with
extension *.fbx. Generally, the model of the imported objects is in gray color, or it
can have certain properties of previously created materials, but commonly the colors
and textures are eliminated. In this way, within the Unity 3D videogame engine,
colors and textures are assigned to the 3D model, depending on the material and
graphic quality required. In addition, the different tools offered by the palette of this
software are used to incorporate the required grass and road, as shown in Fig. 5.
Additionally, the respective environmental sounds are coupled, which complement
the immersive experience of the patient who can feel how the car drives in the middle
of nature.
Fig. 4 Flowchart of the interaction of programs
Fig. 5 Design of the interface in Unity 3D software

4 Tests and Results
4.1 Test
When executing the application, the patient must perform various gestures with his
hand and forearm in order to perform the tasks entrusted, as shown in Fig. 6a, b.
When the first gesture is made, the vehicle is turned on and you can drive on the
road. Once you are in driving mode, moving the palm of the hand outward produces a
turn to the right; depending on the intensity of the gesture, the rotation is performed.
The movement of the palm of the hand inward indicates that it is going to turn to
the left. When you open your hand, it indicates that you have reached the desired
point and so the vehicle must stop. In case that a bad maneuver has been made, the
reverse option is incorporated by joining the thumb and the middle finger. Patients
performed the entrusted exercises in 30 min’ sessions, 2 times a week. This treatment
lasted 4 weeks and was carried out by 5 retired soldiers in a passive state (it must be
Fig. 6 a Tests performed by a patient and b virtual environment seen by the patient
differentiated that 2 of them required rehabilitation in both arms), with an age range
between 48 and 63 years. The supervision of the therapist is very important, since
it must corroborate that patient performs the exercises correctly so that there is an
improvement in the treatment.
4.2 Results
At the end of the application, each patient data and the values obtained in the
videogame are saved in a text file that is automatically generated with the user’s
name. To determine the level of acceptance of this proposal, the usability test SEQ
has 14 questions, of which 13 have the values of 1–5 points according to the following
scheme: The first seven questions (Q1–Q7) are related to the level of acceptance and
immersion after the user experienced the virtual environment. The following four
questions (Q8–Q11) are related to the effects and discomfort that the system could
cause nausea, disorientation, or ocular discomfort. The next two questions are related
to the difficulty encountered when performing the tests. The last question is open,
so the user can indicate if there is any discomfort when using the virtual system and
its reasons. If the result obtained is in the range of 40–65, the implemented system
is considered acceptable. The questions asked to the users about the virtual system
and the results of the SEQ usability questionnaire are shown in Table 1.
The results of the SEQ test performed by five users after using the virtual system
are: (58,8. ± 0,27). The score obtained is greater than 40. Therefore, it is determined
that patients feel fully comfortable with the developed interface, so it represents
an ideal tool to contribute to the rehabilitation process which requires an extended
period of application.
5 Conclusions
The experimental results after the implementation of this system show that it is a
valid proposal for the rehabilitation of upper extremities. With this, it is pretended to
give the patient a new tool to replace the conventional method; immersing his/her in
a virtual world (fusing the sense of sight and hearing) a wide range of possibilities
for physical rehabilitation is presented. An important fact to note is when dealing
with users who were in active service and were subjected to stress situations, the
execution of simple tasks and the simulation of an environment with nature help
them to mitigate possible side effects of a psychological nature. The application of
the usability test SEQ shows that the interface presented to the user is intuitive, easy
to use, and does not cause discomfort. It should be pointed out that this research has
been evaluated from the point of view of the acceptance generated by this system, but
not with a medical approach, given that the results are not conclusive when requiring
a longer evaluation period.
Table 1 Results of the SEQ test applied to users

Question Result (N =
5)
Mean SD
1. How much did you enjoy your experience with the system? 4.8 0.43
2. How much did you sense to be in the environment of the system? 4.8 0.43
3. How successful were you in the system? 4.4 0.83
4. To what extent were you able to control the system? 3.2 0.71
5. How real is the virtual environment of the system? 5 0
6. Is the information provided by the system clear? 4.8 0.43
7. Did you feel discomfort during your experience with the system? 4 0.43
8. Did you experience dizziness or nausea during your practice with the 4.8 0.43
system?
9. Did you experience eye discomfort during your practice with system? 4.6 0.87
10. Did you feel confused or disoriented during your experience with the 5 0
system?
11. Do you think that this system will be helpful for your rehabilitation? 4.2 0.71
12. Did you find the task difficult? 4.8 0.43
13. Did you find the devices of the system difficult to use? 4.4 0.43
Global score (total) 58.8 0.27
In this context, the authors of this research propose as future work the execution of
this application in civil patients, in order to make a comparison and determine factors
of incidence. In addition, the development of a system that allows the rehabilitation
of the lower limbs is proposed, taking advantage of the already established design.
Finally, these authors propose the use of a control algorithm to expand the number
of gestures presented in this work and thus allow a better recovery process.
Acknowledgements To the authorities of Universidad Técnica de Ambato (UTA), Dirección de

Investigación y Desarrollo (DIDE), Instituto Tecnológico Superior Guayaquil Ambato and Celec
EP, for supporting this work and future research.
References
1. Drakos, N.D., et al.: In good conscience: developing and sustaining military combat trauma
expertise. J. Am. Coll. Surg. 227(2), 293–294 (2018)
2. Russell, C.A., Gibbons, S.W., Abraham, P.A., Howe, E.R., Deuster, P., Russell, D.W.: Narrative
approach in understanding the drivers for resilience of military combat medics. J. R. Army Med.
Corps. 164(3), 155–159 (2018)
3. Wells, T.S., Seelig, A.D., Ryan, M.A., Jones, J.M., Hooper, T.I., Jacobson, I.G., Boyko, E.J.:
Hearing loss associated with US military combat deployment. Noise Health 17(74), 34–42
(2015)
4. Writer, B.W., Meyer, E.G., Schillerstrom, J.E.: Prazosin for military combat-related PTSD
nightmares: a critical review. J. Neuropsychiatry Clin. Neurosci. 26(1), 24–33 (2014)
5. Stinner, D.J.: Improving outcomes following extremity trauma: the need for a multidisciplinary
approach. Mil. Med. 181(4), 26–29 (2016)
6. Wilken, J.M., Roy, C.W., Shaffer, S.W., Patzkowski, J.C., Blanck, R.V., Owens, J.G., Hsu,
J.R.: Physical performance limitations after severe lower extremity trauma in military service
members. J. Orthop. Trauma 32(4), 183–189 (2018)
7. Galarza, E.E., et al.: Virtual reality system for children lower limb strengthening with the use
of electromyographic sensors. In: International Symposium on Visual Computing, ISVC 2018,
pp. 215–225. Springer (2018)
8. Gelman, D., Eisenkraft, A., Chanishvili, N., Nachman, D., Glazer, S.C., Hazan, R.: The history
and promising future of phage therapy in the military service. J. Trauma Acute Care Surg.
85(1), 18–26 (2018)
9. Cools, A., Whiteley, R., Kaczmarek, P.K.: Rehabilitation of upper extremity injuries in the
handball player. In: Handball Sports Medicine, pp. 433–459. Springer, Germany (2018)
10. Quevedo, W.X., et al.: Assistance system for rehabilitation and valuation of motor skills. In:
International Conference on Augmented Reality, Virtual Reality and Computer Graphics, AVR
2017, pp. 166–174. Springer (2017)
11. Achanccaray, D., Acuña, K., Carranza, E., Andreu-Perez, J.: A virtual reality and brain com-
puter interface system for upper limb rehabilitation of post stroke patients. In: 2017 IEEE
International Conference on Fuzzy Systems (FUZZ-IEEE), pp. 1–5. IEEE (2017)
12. Levin, M.F., Weiss, P.L., Keshner, E.A.: Emergence of virtual reality as a tool for upper limb
rehabilitation: incorporation of motor control and motor learning principles. Phys. Ther. 95(3),
415–425 (2015)
13. Liu, L., Chen, X., Lu, Z., Cao, S., Wu, D., Zhang, X.: Development of an EMG-ACC-based
upper limb rehabilitation training system. IEEE Trans. Neural Syst. Rehabil. Eng. 25(3),
244–253 (2017)
14. López, V.M., Zambrano, P.A., Pilatasig, M., Silva, F.M.: Interactive system using myoelectric
muscle sensors for the strengthening upper limbs in children. In: International Conference on
Augmented Reality, Virtual Reality and Computer Graphics, AVR 2018, pp. 18–29. Springer
(2018)
Part VIII
Strategy, Geopolitics and Oceanopolitics
The Portuguese Special Operations
Forces as Instrument of Foreign Policy:
The Case Study of Afghanistan
João Reis and Sofia Menezes
Abstract The aim of this paper is to provide an exploratory overview of the Por-
tuguese special operations forces as instrument of foreign policy. In doing so, we
disclose new dynamics that Portugal employing in fragile states and we discuss its
implications for the North Atlantic Treaty Organization. The research methodolog-
ical approach falls into a qualitative case study research, which includes more than
one method of data collection for triangulation and corroboration purposes. We have
found that Portugal is trying to intensify its international relations within the EU, UN
and the NATO, while the Portuguese special operations forces have been playing an
important role to strengthen its presence within these international institutions. The
new dynamics that Portugal is seeking are based on the democratic values, the rule
of law and the human rights, to dissociate itself from the colonial legacy and thereby
find new avenues of influence. Further research should focus on the new strategic
priorities for the Afghan Government and NATO, as these priorities are changed at
a relatively fast pace.
Keywords Portugal · Afghanistan · Special operations forces · Foreign policy ·

Case study research · European Union · United Nations · North Atlantic Treaty
Organization
1 Introduction
This paper focuses the Portuguese special operations forces as instrument of foreign
policy in Afghanistan. We have selected Afghanistan because this country has played
a major role with great regional empires of Persia, India and China in late medieval
J. Reis (B)
Department of Military Science, Military Academy, GOVCOPP, Lisbon, Portugal
e-mail: joao.reis@academiamilitar.pt
S. Menezes
Department of Management and Leadership, CINAMIL, Lisbon, Portugal
e-mail: sofia.menezes@academiamilitar.pt

https://doi.org/10.1007/978-981-13-9155-2_20
246 J. Reis and S. Menezes
and early contemporaneous period [1]. The country is known as the graveyard of
empires [2] that reputation invokes the British and Soviet defeats in several wars
in the nineteenth and twentieth century, to argue that Afghanistan is an unconquer-
able quagmire that inevitably dooms even the strongest army that venture there [3].
Moreover, Afghanistan was the first major test for state-building in the twenty-first
century and embroiled the USA and its allies for almost two decades [4].
Afghanistan required virtually every North Atlantic Treaty Organization (NATO)
member’s participation, including Portugal, and its participation is far from certain.
Most probably, the economic and political assistance will continue, even after the
Afghan forces assumed full security responsibility at the end of 2014—the new,
smaller, non-combat mission Resolute Support (January 1st, 2015—until recent days)
began to train, advise and assist (TAA) the Afghan security forces and institutions
[5].
We have chosen the well-trained special operations forces instrument due to its
strategic value, as they were the first to be deployed to Afghanistan, and they will
be probably the last to redeploy. The objective of this research is to ascertain the
suitability of the Portuguese special operations forces within the NATO Resolute
Support Mission (RSM) and its relevance as an instrument of foreign policy for
Portugal.
We have structured this paper in four sections after the introduction, as follows:
we begin by a review of the literature; secondly, we describe the methodological
approach; then, we investigate the relevance of the special operations forces to NATO
and Afghanistan; in the last place, we provide the implications and suggestions for
subsequent research.
2 Theoretical Background
2.1 Allied and the Portuguese Army Special Operations

Forces
NATO Allied Joint Publication (AJP) defines that special operations as military
activities conducted by specially designated, organized, trained and equipped forces,
manned with selected personnel, using unconventional tactics, techniques and modes
of employment—special operations deliver strategic or operational-level results or
are executed where significant political risk exists [6]. On the other hand, the NATO
special operations headquarters define special operations as operations that may be
conducted across the spectrum of conflict as part of article 5 (collective defence)
or non-article 5 (crisis response operations) to fulfil NATO’s three essential core
tasks—collective defence, crisis management and cooperative security [7]. The
NATO special operations headquarter also describes the special operations missions
to include suitable combination or all of the principal tasks of military assistance
(MA), special reconnaissance (SR) or direct actions (DA), depending on the cir-
cumstances of each operation. While special operations mission may range from
The Portuguese Special Operations Forces … 247
small unilateral actions to large-scale activities of a combined and joint nature, they
are tailored to contribute to the accomplishment of the defined political and mili-
tary/strategic objectives.
Each type of special operations missions may be defined as follows: DA is the
use of a rapid offensive operations to destroy or capture a target, usually by means
of secrecy and surprise [8, 9]; SR involves covert collection of intelligence typically
in denied enemy areas [9, 10]; MA involves the provision of training and advising
to host nations to improve their skills and capabilities [8, 9].
In Portugal, an order from the Chief of Staff of the army (directive nº 90), October
28th, 2007, determines that special operations are divided in special reconnaissance,
direct actions and military assistance. The forces that develop special operations may
be employed across the spectrum of operations. In view of its unique characteristics,
it constitutes adequate forces to perform tasks within the crisis response operations
(CRO), combat operations for search and rescue (CSAR) and counterterrorism oper-
ations, consisting in forces especially suitable in counterterrorism [11]. The same
directive from the Chief of Staff of the Portuguese army states that: (1) SR objective
is to obtain strategic and operationally sensitive information, it can use sophisticated
methods, techniques (e.g. human intelligence) and equipment [11]; (2) DA is a pre-
cise action, which is usually performed on well-defined targets, high criticism and
high value, with strategic or operational significance. The use against tactical objec-
tives is justified only by the action of criticality or by the inadequacy or unavailability
of the use of other forces [11]; (3) MA consists of a broad framework of measures
in support of friendly or allied forces and populations in peace, crisis or conflict,
in situations of need for security and survival, but also development. The MA also
fits into all activities related to the actions of military-technical cooperation, support
in situations of disaster and conflict mediation [11].
The missions developed by the Portuguese special operations forces are similar
to reference countries and NATO members. This section has the objective to define
the elementary concepts, which are going to be necessary for the following sections.
2.2 Resolute Support Mission and the Afghan Government
The RSM in Afghanistan is focused on train, advise and assist (TAA) the Afghan
security forces. Historically, the USA special operations forces (SOF) have had the
responsibility to TAA foreign military forces. Even if today, the US National Security
Strategies introduced the US military services (Army, Navy, Air Force and Marines)
to organize, train and equip themselves to carry out these activities at a larger scale
with conventional (non-SOF) forces [12]. Livingston [12] remarks this responsibility
in a broader sense of building the capacity of partner states that has been termed
security force assistance (SFA). SFA is part of the USA and NATO strategic goal
of having Afghanistan responsible for their security. Usually, SFA is directly linked
to counterterrorism strategy and is the key to engage underdeveloped and under-
governed nations, often referred to as weak or fragile states.
In Afghanistan, the SFA mission is in part attributed to the NATO special oper-
ations component command—Afghanistan. This command brought together NATO
and other coalition special operations forces into the same operational headquar-
ters, and it demonstrates both that NATO special operations forces have made major
progress in integration and can potentially be a major part of the alliance capability
[9]. The created synergies among special operation forces community achieve signifi-
cant effect due to the combination of different techniques, technologies and expertise.
Long [9] reinforces the idea of NATO special operations forces that actively performs
three main allied special operations forces missions, i.e. military assistance, special
recognizance and direct actions; but particularly, the military assistance mission to
develop the Afghan police special units was its core mission.
Strandquist [13] also performed a research which suggests the USA and their
allies should change their strategy in Afghanistan, since developing military combat
operations against the insurgents is not enough and the future will have to go through
a focus on the development of local communities to counter-insurgency and that the
Afghan central government should win their support.
During the war in Africa (1961–1974), the Portuguese armed forces also used the
same techniques, in order to obtain public support, as it may be one of the key points to
win the war in Afghanistan. This technique was also used in Vietnam and it had been
termed “winning hearts and minds” [14]. The term was firstly used by Louis Lyautey,
a French army marshal and minister of war. The term included not only military, but
also social actions and improvements in the quality of life [15]. Later, Grochowski
[16] published an article concerning the Takur Ghar battle, a short, but intensive,
military operation that engaged the US special forces and Al Qaeda insurgents in
Afghanistan. According to Grochowski [16], the Takur Ghar battle demonstrated
the short lead-time and short-duration high-intensity tempo of the special operations
missions regarding the war against non-state actors and particularly violent extremist
organizations. The course of Afghan conflict and the commitment of the coalition of
special operations forces has focused on the tactical objectives, but with time, this
path has changed, i.e. the current conflict requires special operations forces highly
committed to TAA the Ministries of the Government of the Islamic Republic of
Afghanistan (GIRoA), which creates strategic and political objectives for a rapid
development of the Afghan Security Forces and its population.
2.3 Brief Overview of the Portuguese Foreign Policy
Between April 1974 and January 1986, the Portuguese foreign policy oscillated
between two basic orientations, which marked two distinct phases: the transition
to democracy and the democratic consolidation [17]. Teixeira [17] also portrays
the Portuguese position in the Euro-Atlantic binomial, in the post-colonial relations
regarding the new multilateral triangle and based on foreign policy values, which we
summarize as follows: (1) The Europe-Atlantic binomial remains the geopolitical
equation: Portugal as an EU country seeking to promote and enhance not only the
Atlantic position between Europe and the USA, but also the post-colonial relations
(African Portuguese-speaking countries), the key for the strategic guidelines of the
foreign policy of democracy [17–19]; (2) Portugal seeks participation in terms of
producing international security, such as the military presence in organizations as
NATO, EU and UN (the new multilateral triangle), which has provided not only a
factor of military modernization, but also the external credibility of the Portuguese
State. The armed forces have become a foreign policy tool [17, 19, 20]; (3) The
challenge is in the global plan of international security, in an environment marked by
terrorism, ethnic and religious conflicts; that is why the framework of security has
to be cooperative security [17, 21]; (4) The presence of the Portuguese armed forces
in peacekeeping missions under the umbrella of the UN confirms the diplomacy
of values that had the independence of Timor as one of its major hits [17]; (5)
Lastly, foreign policy always comes from the formulation of interests. However, the
contemporaneous democratic model ignores the interests of the states and focuses
on the human values. The values for Portugal are democracy, the rule of law and
human security [17].
Similarly, Santos [22] argues about the use of the armed forces which is one of the
most important instruments available to support its foreign policy, giving it a specific
weight within the security scope.
3 Methodology
The methodological approach falls into a qualitative case study research. This
research therefore uses a contextual rich data setting from the real world to inves-
tigate a phenomenon which is understudied. We do use multiple sources of data
collection [23], i.e. semi-structured interviews, official documents and direct obser-
vations, for data triangulation and corroboration purposes [24]. The data collection
method focused on ten interviews from two military branches of the NATO Result
Support Mission in Afghanistan; the non-classified documents were collected from
open sources; and the direct observations have collected from informal interviews
and personal field notes of the observed reality. The field study took place in Kabul,
Afghanistan during the following periods: February to July of 2015 and from August
2017 to February 2018. Furthermore, for confidential reasons, all the respondents
kept their anonymity and the collected reports came from unclassified sources. The
goal of the case study is to gain a holistic view of the phenomenon and to understand
the specific reality that is being discussed in this paper [25].
3.1 Data Collection Techniques
The interviews were the primary data collection method for this case study. Quivy
and Campenhoudt [26] reported the interviews as a method that differs from others
250 J. R eis a n d S. M e n e z es
i n t h e e xt e nt t h at it r ef ers t o t h e f u n d a m e nt al pr o c ess es of c o m m u ni c ati o n a n d h u m a n

i nt er a cti o n. T h e s a m e a ut h ors als o m e nti o n e d t h at t his t e c h ni q u e, w h e n pr o p erl y v al-
u e d, e n a bl es t h e r es e ar c h er t o wit h dr a w i nf or m ati o n a n d v er y ri c h r e fl e cti o n el e m e nts
fr o m i nt er vi e ws. H o w e v er, w e k n o w t h at, s o m eti m es, r es p o n d e nts c a n pr o vi d e t h e
i nt er vi e w er wit h t h at h e w a nts t o h e ar, r es ulti n g i n p o or a n d i n a c c ur at e r es p o ns es
[2 4 ], w hi c h m a y b e miti g at e d t hr o u g h t h e us e of ot h er d at a c oll e cti o n t e c h ni q u es.
T h e s a m pl e s el e cti o n c o m pr e h e n d e d t e n i nt er vi e ws, di vi d e d i nt o t w o gr o u ps of
fi v e. B ot h gr o u ps s er v e d u n d er t h e c o m m a n d of t h e N A T O s p e ci al o p er ati o ns c o m-
p o n e nt c o m m a n d — Af g h a nist a n. T h e first fi v e ar e s er vi c e m e n fr o m t h e U S s p e ci al
o p er ati o ns a n d t h e s e c o n d gr o u p ar e s er vi c e m e n fr o m t h e P ort u g u es e ( P R T) s p e ci al
o p er ati o ns f or c es. T h e U S a n d P R T of fi c ers h a v e i nt er a ct e d wit h t h e m ost s e ni or l e v-
els of t h e Af g h a n Mi nistr y of I nt eri or a n d dir e ctl y i n fl u e n c e d t h e l o n g-t er m str at e g y
t o s ust ai n t h e Af g h a n S p e ci al S e c urit y F or c es ( A S S F). T o g et diff er e nt p ers p e cti v es,
w e h a v e als o c oll e ct e d a c o ntri b uti o n fr o m s o m e P ort u g u es e c o m missi o n e d of fi c ers
t h at h a v e w or k e d as st aff of fi c ers fr o m t h e S p e ci al O p er ati o ns A d vis or y A n al ysts.
T h e c as e st u d y pr ot o c ol w as b as e d o n t h e P ort u g u es e N ati o n al D ef e n c e Str at e g y
( P N D S), w hi c h w as a p pr o v e d b y t h e r es ol uti o n of t h e P ort u g u es e c o u n cil of mi nist ers
9/ 2 0 1 3, A pril 5t h, 2 0 1 3. T h e P N D S w as d e v el o p e d a c c or di n g t o t hr e e m ai n v e ct ors:
( 1) n ati o n al s o v er ei g nt y — n e utr ali z ati o n of t hr e ats a n d ris ks t o n ati o n al s e c urit y, ( 2)
a cti o ns t o n ati o n al v ul n er a biliti es a n d ( 3) e n h a n c e m e nt of t h e r es o ur c es a n d n ati o n al
o p p ort u niti es. E a c h v e ct or is di vi d e d of s e v er al str at e gi es ( 2 0 str at e gi es i n t ot al), a n d
e a c h str at e g y is s u b di vi d e d i nt o li n es of a cti o n ( 1 1 6 li n es of a cti o n i n t ot al) —f or t h at
r e as o n, it w as i m p ossi bl e t o dis pl a y all t h e N ati o n al D ef e n c e Str at e g y str at e gi es i n
s h ort p a p er.
T h e n ati o n al v e ct ors, t h e str at e g y a n d t h e s e v er al li n es of a cti o n c o ntri b ut e d
dir e ctl y or i n dir e ctl y t o t h e a c hi e v e m e nt of p er m a n e nt a n d s h ort-t er m n ati o n al g o als
i n Af g h a nist a n. T h e c o ntri b uti o n of b ot h gr o u ps w as b as e d o n t h e c h oi c es of e a c h
( 1) v e ct or, t h e c orr es p o n di n g ( 2) str at e g y a n d t h e ( 3) li n es of a cti o n fr o m t h e P N D S
( vi d e Ta bl e 1 ).
T a bl e 1 Ve ct ors, str at e gi es a n d li n es of a cti o n of t h e P ort u g u es e n ati o n al d ef e n c e str at e g y

V e ct ors Str at e gi es Li n es of a cti o n
E x er cis e s o v er ei g nt y, C o ns oli d ati n g
I nt e nsif yi n g e xt er n al r el ati o ns a n d
n e utr ali z e t hr e ats a n d e xt er n al d ef e n c e
d ef e n c e c o o p er ati o n wit h t h e U S
ris ks t o n ati o n al r el ati o ns
s e c urit y R e affir m t h e str at e gi c i m p ort a n c e of t h e
D ef e n di n g P ort u g al’s
i n cl usi o n of P ort u g al i n N A T O, as w ell
i nt er n ati o n al p ositi o n
as t h e bil at er al alli a n c e wit h t h e U S
C o ns oli d ati n g P arti ci p at e i n i nt er n ati o n al milit ar y p e a c e
e xt er n al d ef e n c e missi o ns, p arti c ul arl y i n t h e c o nt e xt of
r el ati o ns t h e U N, N A T O a n d E U
D ef e n di n g P ort u g al’s Str e n gt h e n t h e pr es e n c e of t h e
i nt er n ati o n al p ositi o n P ort u g u es e i n i nt er n ati o n al i nstit uti o ns
To ensure the data corroboration, we have used other data collection techniques,
i.e. secondary sources, such as direct observations and official documents. We did
not give priority to these data collection techniques, but they were essential for
triangulation and corroboration purposes.
The direct observations came from field notes that were taken during the research
process and from the author’s own experience, since he has also served as a special
operations operator in Afghanistan. Furthermore, we have also collected data from
the frequent discussion with NATO servicemen about their judgement regarding the
Portuguese special operations forces contribution to NATO and, consequently, to
Afghanistan. The informal discussions were registered and once they provided a
better understanding of the phenomenon, they were taken into account.
We also collected official documents from NATO website that were relevant to
provide the Resolute Support Mission perspective and the commander intention about
all NATO countries contribution to the overall mission.
To analyse the collected data, we defined a set of categories and subcategories to
search associations between the various sources of data collection. Then we sought
to establish relations between the various categories in order to draw conclusions.
The NVIVO10 was used as a qualitative data analysis software to organize, analyse
and finds insights into unstructured or qualitative data.
4 Findings
This section provides an empirical summary of the data analysis, which includes
real-life statements from the respondents.
The most valued aspect attributed by the US special operations forces is regard-
ing the knowledge acquisition by the Portuguese forces. The respondents believed
the Portuguese special operations forces are looking forward to acquire new expe-
riences and new competencies through the contact with other international military
forces, especially in counterinsurgent environments. This exchange of experiences
is allowing the Portuguese servicemen to increase their technical skills, which can
be boosted by continued experience sharing with contributing NATO nations. These
relationships mainly promote the exchange of relevant information that allows tech-
nical innovation with European partner. A US special operations serviceman shares
this view:
I believe that Portugal is participating at this NATO mission (Afghanistan) to intensify its
relations with the United States and to better know our modi operandi. This aspect is relevant
for us because it allows for a better interoperability within NATO, but also because we can
work with our allies (special forces operator, 32 years old, US)
The US special operations operators also consider Portugal as a strategic partner

relevant to the USA, as it is a country within the European Union that allows strength-
ening the transatlantic ties, because of its relative proximity to the USA—being the
most western country in Europe. They believe it is increasingly relevant to have a
close cooperation with the Portuguese special operations forces, as well as with all
the European countries, as they take into account the European contribution to the
international security—cooperative security. The arguments are well evidenced in
by a US special forces team leader:
Portugal has a very close relationship with the US. To my best knowledge, we have a US base
in a Portuguese Island (Azores) that works as a strategic platform to Europe. The relationship
we have with Portugal is not only due to its geographical position, since we also have had
several special operations exercises, namely cross training in infiltration techniques (special
forces team leader, 42 years old, US)
From the second group of respondents, it was possible to ascertain that the most
valued aspects were the participation in missions that have contributed to peace and
international security, especially the ones that have strengthening the presence of
the Portuguese in international organizations. They believe that Portugal has a stable
international status as a western democracy, a full member of the EU and NATO,
which gives the country the possibility to actively participate in the international
panorama. The participation in international missions, namely through elite forces,
creates bonds between Portugal and the countries of the alliance and contributes to
improve the country´s visibility and credibility. These arguments are well evidenced
by the statement of a Portuguese special operations operator:
The Portuguese special operations forces were created as a counter-insurgency unity for
the Portuguese operations in Africa, during the colonial war period (1961–74). After the
revolution of April 25th, 1974, the Portuguese special operations had to change their way
of employment. The special operations are now used as an icon of national sovereignty,
which acts in the international context as a co-producer of security while working with their
international partners (special forces operator, 32 years old, Portugal)
These respondents also believe the use of the Portuguese special forces intensifies
the relations and defence cooperation with the US military. A good example they
provided is the participation of the Portuguese Commandos in Kabul Capital Divi-
sion, at the time under the mandate of the International Security Assistance Force
(ISAF) and, very recently, with the deployment of special operations forces to which
were assigned several roles, e.g. intelligence, ministerial engagement, etc., under the
mandate of the Resolute Support Mission (RSM).
The Portuguese special forces have long contributed to the develop of the Afghan National
Defence and Security Forces capabilities. I personally work with the General Command of
Special Police Unit (GCPSU), which is an Afghan police unit responsible for responding
to high-profile attacks and to conduct high-risk arrests. The most important for us is to
participate in the peacekeeping operations and to strengthen the Portuguese presence within
the international institutions i.e. NATO. Yes, the Portuguese special forces are operating in
Afghanistan almost since the beginning of the conflict, I am just a small piece this well-
organized and very professional institution (special forces operator and ministerial advisor,
34 years old, Portugal)
During the discussions, we also noticed that Portugal is seeking new dynamics,
based on the democratic values, the rule of law and the human rights, and thereby find
new avenues of influence. We observed that the Portuguese special operations are
also trying to get more influence within the Afghan Ministry of Interior (rule of law),
but also investing on issues such as the gender integration in the Afghan defence
and security forces (human rights). Therefore, further research should investigate
these topics, i.e. new strategic priorities for GIRoA and the NATO. For instance, it
would be interesting to investigate how the Portuguese special operations forces can
improve the integration of Afghan women into their special operations force units.
The data analysis also demonstrated the Portuguese special operations mis-
sions—military assistance, falls under the overall mission of the NATO special
operations component command. Regarding the political foreign objectives, the Por-
tuguese special operations forces capacities and the missions performed by the NATO
special operations forces are perfectly aligned and have two main goals:
(1) influence the long-term strategy and sustainability of the Afghan defence and
security forces.
I feel that we have an effective influence on the defence and security policy. Our
Afghan counterparts are generals and politicians form the Afghan Ministry of Inte-
rior. At the same time that we have received the milestones from NATO and Portugal,
we can decisively influence the long-term sustainability of the Afghan special forces
(informal discussion) and, to develop activities (2) which can strengthen its interna-
tional present and to stimulate cooperative security relations with their international
partners.
In recent years there has been an increase of special operations servicemen to be deployed
to Afghanistan, which may be a reflection of this policy (field notes)
By using its elite forces in Afghanistan as instrument of foreign policy, Portugal

reaches: 1 national vector, 4 strategies and 4 lines of action, from the PNDS. Table 1
summarizes the selected vectors, strategies and actions by the respondents.
In light with the above, we view the Portuguese strategy in Afghanistan from two
different perspectives. At first, it was interesting to observe that questions related
to the scope of the Portuguese mission in Afghanistan, i.e. military assistance, the
military discourse was focused on strengthening the Portuguese presence within the
international institutions and the military participation in peacekeeping operations,
in particular contexts such as the UN, NATO and EU (Table 1—light grey). However,
when we evaluated the Portuguese foreign policy, we cross-checked the interviews
from both groups of respondents and we verified the existence of a strong relation
between the USA and Portugal, with the intention to keep strengthening the transat-
lantic bonds (Table 1—dark grey).
The Portuguese strategies and line of action are consistent within the PNDS, i.e.
all the strategies and lines of action falls under the same national vector: to exer-
cise sovereignty, neutralize threats and risks to national security (vide Table 1). The
purpose of this vector is to defend national interests and guarantee the security and
defence of Portugal, reduce vulnerabilities and execute the international responsi-
bilities of the state and contribute to the realization of the Portuguese vision of the
constructive values of the international order [27].
5 Concluding Remarks
We have concluded that Portugal is trying to intensify its international relations within
the EU, UN and NATO. To this end, the Portuguese armed forces have been playing
an important role to strengthen its presence within the international institutions.
We have verified that the Portuguese special operations operators are trained well
enough to operate in complex environments, as the Afghan mission falls under one
of the main special operations missions—the military assistance. In other words, the
mission that brings the Afghan Ministerial capabilities to been raised is suitable for
the Portuguese special forces.
Portugal, as a small European state, has also tried to contribute along with the
USA to spread the rule of law. In Afghanistan, once NATO moved from a combat
role to a TAA mission, the Portuguese special operations forces have been devel-
oping a strategy to build the Afghan Government, as well as the Afghan National
Defence and Security Forces. Thus, under the NATO special operations component
command, the Portuguese special operations operators have played one of the most
demanding missions in the contemporary Portuguese military history. We therefore
closely observed some Portuguese servicemen that helped decisively to develop the
capabilities of the Ministry of Interior, in order to operate the Afghan police forces
and its special units, e.g. GCPSU.
What is new is that Portugal is seeking new dynamics—based on the democratic
values, the rule of law and the human rights—to dissociate itself from the colonial
legacy. Portugal is therefore, trying to get more influence, by using its armed forces,
to rebuild fragile states in remote geographical areas, along with their international
partners.
This paper also presents some limitations, as the identity of the respondents may
not be revealed, thus to clarify any questions to address this subject the first author
should be contacted. The number of interviews was also limited, once it was restricted
to the number of respondents available in a specific period of time. The study is also
exploratory in nature and it lacks generalization. The lack of generalization is due
to different political goals and different existing types of special forces; thus, the
results may differ across NATO countries. Nevertheless, we do believe this study
may be useful to other researchers, in the extent that are no similar studies so far;
to practitioners, this paper may be relevant too, as it can provide new and fruitful
insights into the strategic engagement of special operations forces in fragile states.
References
1. Seddon, D.: Imperial designs. Crit. Asian Stud. 35(2), 175–194 (2003)
2. Jones, S.: In the graveyard of empires: America´s war in Afghanistan. WW Norton & Company,
New York (2010)
3. Miller, P.: Graveyard of analogies: the use and abuse of history for the war in Afghanistan. J.
Strateg. Stud. 39(3), 446–476 (2016)
4. Nojumi, N.: American State-Building in Afghanistan and its Regional Consequences: Achiev-
ing Democratic Stability and Balancing China’s Influence. Rowman & Littlefield (2016)
5. NATO.: NATO and Afghanistan. North Atlantic Treaty Organization (2016). http://www.nato.
int. Accessed 1 Dec 2018
6. AJP 3.5.: Allied joint doctrine for special operations. North Atlantic Treaty Organization (2013)
7. NATO SOF.: NATO special operations headquarters: Special operations forces. Military Assis-
tance Handbook. 1st study draft, North Atlantic Treaty Organization (2014)
8. AAP-06.: NATO glossary of terms and definitions. North Atlantic Treaty Organizations. 2-D-7
and 2-S-9 (2013)
9. Long, A.: NATO special operations: promise and problem. Obris 58(4), 540–551 (2014)
10. Dellinger, B.: Special operations command Europe: strengthening partnership for global secu-
rity. Spec. Warf. (2012)
11. Chefe de Estado-Maior do Exército.: Âmbito de atuação, missões e tarefas das unidades de
tropas especiais. Divisão de Planeamento de Forças, diretiva nº9 do Chefe de Estado-Maior do
Exército (2007)
12. Livingston, T.: Building the capacity of partner states through security force assistance. Security
Force Assistance in the Development of Foreign Forces. Nova Science Publishers (2011)
13. Strandquist, J.: Local defence forces and counterinsurgency in Afghanistan: learning from the
CIA´s village defense program in South Vietnam. Small Wars Insur. 26(1), 90–113 (2015)
14. Anderson, D., Ernst, J.: The war that never ends: new perspectives on the Vietnam War. Uni-
versity Press of Kentucky (2014)
15. Tucker, S., Roberts, P.: The encyclopedia of World War II: A political, social, and military
history. ABC-clio (2004)
16. Grochowski, G.: Planning for the next Takur Ghar. Def. Secur. Anal. 31(2), 152–158 (2015)
17. Teixeira, N.: Breve ensaio sobre a política externa portuguesa. Relações internacionais, R:I(28),
51–60 (2010)
18. Ruivo, L.: Conceito estratégico de defesa nacional contributos para uma revisão. Boletim
Ensino, Instituto de Estudos Superiores Militares (2011)
19. Silva, J.: As forças armadas como instrumento de política externa portuguesa no actual contexto
internacional. Trabalho Individual de Longa Duração do Curso de Promoção a Oficial General,
Instituto de Estudos Superiores Militares (2006)
20. Azambuja, M.: As Nações Unidas e o conceito de segurança coletiva. Estudos Avançados 9(25),
139–147 (1995)
21. Serronha, M.: A cimeira de Lisboa: Uma NATO para o século XXI. Instituto de Defesa Nacional
126–5ª série, 37–46 (2010)
22. Santos, J.: O conceito estratégico da NATO: Superar contradições, manter a coesão. Nação e
Defesa, 37–46 (2010)
23. Barratt, M., Choi, T., Li, M.: Qualitative case studies in operations management: Trends,
research outcomes, and future research implications. J. Oper. Manag. 29(4), 329–342 (2011)
24. Yin, R.: Case study research design and methods. Appl. Soc. Res. Methods Ser. 5 (2003)
25. Benbasat, I., Goldstein, D., Mead, M.: The case research strategy in studies of information
systems. MIS Q. 369–386 (1987)
26. Quivy, R., Campenhoudt, L.: Manual de investigação em ciências sociais (1998)
27. Governo de Portugal.: Grandes opções do conceito estratégico de defesa nacional (2013).
https://www.parlamento.pt/sites/COM/Paginas/DetalheNoticia.aspx?BID=5034. Accessed 1
Dec 2018
The Internal–External Security Nexus:
EU Operation Sophia Through the Lens
of Securitization
Ana Paula Brandão
Abstract The construction of EU security actorness has been accompanied by a

narrative on security nexuses associated with the “comprehensive approach”. This
trend has been intensified in recent years as demonstrated by the discourse on the
“refugee crisis”, Daesh activity, hybrid threats and border security. This paper focuses
on the internal–external security nexus analysed through one of its materializations
notably the interface between the CSDP and the AFSJ. Based on the EU Naval
Force Operation Sophia, it is argued that the comprehensive approach underlying
the logic of the nexus is the result of a co-constitutive adequacy: “more security”
(appropriation of policies and instruments of a multifunctional actor for security
purposes) and “more actorness” (securitization of issues in order to promote the
actor and its policies). The main contribution of the paper is to think critically on
why and how security narratives, military instruments and securitization dynamics
serve convergent processes of gaining political and public space for legitimizing
policies and actions.
Keywords CSDP · AFSJ · Security nexus · Securitization
1 Introduction
In a context in which the European Union (EU) discourse has been fertile in iden-
tifying Europe’s challenges in a globalized world, amongst which we find the post-
Westphalian security challenges, it is paramount to reflect upon the rationale and
effects of the evolving EU security actorness. The Union has been innovative in cre-
ating a de facto security community that overcame the European interstate conflict,
and since the end of the Cold War, it endeavoured both to be a security actor (with
global reach and through a comprehensive approach) and to address the multisector
and transnational threats of a complex security environment.
A. P. Brandão (B)
CICP—University of Minho, Braga, Portugal
e-mail: abrandao@eeg.uminho.pt

https://doi.org/10.1007/978-981-13-9155-2_21
258 A. P. Brandão
EU security actorness reinforces and is reinforced by the narrative on the security

nexuses: internal–external security; public–private security; civilian–military secu-
rity; security development. This narrative and its practices raise several questionings
about the what (means the nexus), the why, the how (to implement it) and the “with
what effects”. The paper focuses on a Common Security and Defence Policy (CSDP)
operation which demonstrates in the field the nexus between internal and external
security (IESN), and more specifically the link between the CSDP and the area of
freedom, security and justice (AFSJ).
Resorting to the securitization theoretical framework combined with a concep-
tual matrix on security actorness [1], applied to the analysis of the European Union
Naval Force Operation Sophia (a recent operation dealing with complex areas of law
enforcement, trafficking of human beings and migration in the neighbourhood), it
is argued that it reflects a securitizing move of the European actor explained by the
convergence of opportunity (redefinition of security and prioritization of transna-
tional threats in a globalized world), capacity (legal, institutional and operational
capacity in the field of security), and (ambition to have a global) presence. The
comprehensive approach (CA) underlying the logic of the nexus is the result of a co-
constitutive adequacy: “more security”—appropriation of policies and instruments
of a multifunctional actor for security purposes (security of the EU and of European
citizens); “more actorness”—securitization of issues in order to promote the actor
and its policies.
2 The Internal–External Security Nexus
The interdependence between the internal and external dimensions is a transverse

view to the official EU documents relating to security [2]. What does this interdepen-
dence mean? Both in the political and academic contexts, different expressions, not
necessarily synonymous, have been used to refer to the phenomenon. This wording
cacophony does not facilitate the work of policy-makers and academics. The scien-
tific field has been marked by “empirical ambiguity, theoretical fragmentation and a
lack of scholarly dialogue on this issue” [3].
Historically, the study of security, associated with state polity, was based on the
separation between “the two arms of the Prince” [4]. The complexity of the phe-
nomenon, associated with the diversification of threats and the multitude of actors,
either as providers of security or as a source of threat in the context of intense mobil-
ity and communicability worldwide, bucked the traditional paradigmatic, political
and institutional separation between the internal and external dimensions of secu-
rity defined by the realist legacy. The end of the Cold War and the events of 11
September 2001 potentiated the broad conceptualization of security covering four
areas: security sectors (multisectoral security beyond political and military sectors);
security objects (multiple actors, including individuals and groups beyond the state);
security players, either as security providers or as sources of threat; cross-border
dynamics (transgovernmental cooperation for security; actions of transnational enti-
The Internal–External Security Nexus: EU Operation Sophia … 259
ties for security purposes; perverse transnational actors). In sum, an evolving trend on
internal and external security, from the traditional approach as separated phenomena
(Realism) to the perspectives as interdependent (Neoliberalism) or even as merged
phenomena (Paris School) [5].
As far as the European Union is concerned, the nexus can be applied to different
phenomena which, in short, stem from three dynamics: (a) internalization of external
phenomena; (b) externalization of internal phenomena; (c) cross-border phenomena.
Underlying the in/out narrative is the idea of “globalization of security” associated
with the “predominantly transnational character of postmodern risks” [6]. In this
context, a CFSP that is effective in preventing and combating external threats is
considered to be a condition to ensure the internal security of the European area. In
turn, an effective internal security system is understood as a condition for the former
to be an active policy. This “indissoluble link between internal and external aspects
of security” [7] is explained by several phenomena, namely: Europe’s vulnerability
due to its reliance on an infrastructure interconnected in various areas (transport,
energy, information); the external dimension of organized crime; the global nature
of terrorism, which has increasing resources, including connection through electronic
networks; proximity to troubled areas; regional conflicts that have direct or indirect
impact on European interests; climate change that has a “threat multiplier effect”
[8]. Thus, in the “era of globalization, distant threats may be as much a concern
as those that are near at hand” so “the first line of defence will often be abroad”
[7], and it is therefore necessary to “improve the way we reconcile the internal
and external dimensions” [8]. The most recent events, particularly in the field of
terrorism, have contributed to intensifying the in-out nexus security narrative. In the
words of Federica Mogherini, the fight against radicalization and violent extremism
must continue to be “a priority, not only for internal and security action, but also for
our diplomatic and foreign policy” [9].
The EU agenda on IESN has extended the list of priority issues: transnational
organized crime (since the 1990s), terrorism (since 9/11), migrant smuggling and
human trafficking (post “refugee crisis”), the link between CSDP-FSJ: “The linkages
between internal and external security, including in areas such as irregular migration,
trafficking of all sorts, terrorism and hybrid threats are an increasingly important
context for the further development of CSDP and when considering any possible
new operation/mission” [10]. After the date of entry into force of Lisbon Treaty
changes, the CSDP-FSJ link was the subject of a structured reflection contained in
the joint report of the commission and the HR, which resulted in an operational
roadmap [11] with five areas and 27 lines of action (12 short-time priorities and 15
medium-term priorities). Those initiatives of reflection and operationalization are
concerned specifically to the external component (CSDP missions and operations)
and the externalization of EU internal security (external dimension of AFSJ). The
internalization (of CSDP) component is being approached in a separated vector, in
particular associated with the solidarity clause.
The civilian missions of CSDP are one of the tangible expressions of the IESN,
particularly with regard to security goals, civilian capabilities, planning, situational
awareness and strategic analysis [12]: civilian missions of Rule of Law, Security
260 A. P. Brandão
Sector Reform designated to reduce the criminal activity (organized crime and cor-
ruption) in the host country, through judicial means, thereby contributing to greater
security and stability not only in that country, but also at regional and international lev-
els, and creating the conditions for cooperation with EU and its agencies (EUROPOL,
EUROJUST, FRONTEX); missions sustained by internal police and judicial capa-
bilities; crisis management structures which integrate elements from both domains
(CSDP and FSJ) in the planning of new missions and revision of ongoing missions;
situational reports, risk and threat assessment, early warning and alerting about ter-
rorist attacks and external crisis, which cover both domains (CSDP and FSJ), as
demonstrated by the activity of the EU Situation Centre (SiteCen), EUROPOL and
FRONTEX. The second manifestation of the CSDP-FSJ link regards to intelligence
gathering and sharing whether in Brussels or in the field of the missions/operations,
which has been fostering administrative arrangements between the General Sec-
retariat of the Council and FSJ agencies as well as the interaction between these
agencies and CSDP civilian missions.
The assessment of the first years of implementation of the roadmap (four progress
reports) approved in 2011, and the need to adequate it to situational changes led the
Council to define three priority areas of action [10]: improving situational awareness
and exchange of information within the EU; operationalizing the nexus between
internal/external security, especially by improving the cooperation among the EU
agencies and between these and the EEAS, the joint training, and the coordination and
decision-making process mechanisms; civilian–military convergence and synergies
among EU missions and operations.
Although the documents value the CSDP civilian missions in the link between this
policy and FSJ, we find relevant to focus on a maritime operation which materialize
the IESN through military means used to deal with the (so-called) new type threats
[13].
3 Operation Sophia: The CSDP-AFSJ Link Through

Military Means
In order to tackle the so-called refugee crisis, the EU adopted a comprehensive

approach (CA) based on short-time (emergency support) and medium-long efforts,
multisectoral policies (humanitarian, migration, development, security and external
relations), multilevel (European, national, local) and multi-actor action (EU institu-
tions and governments, partner countries, partner organizations, international agen-
cies, NGOs).
The use of military means (CSDP operation) and the concentration on the sea
(maritime operation) were decided after the tragic sinking and death of hundreds of
migrants attempting to cross the Mediterranean [14]. On 20 April 2015, the European
Commissioner for Migration, Home Affairs and Citizenship presented a ten-point
action plan on migration in response to the crisis in the Mediterranean. The second
action of the plan consisted in the capture and destruction of vessels used by the
smugglers. This was reiterated by the European Agenda on Migration (May 2015)
as an immediate key action.
The Crisis Management Concept of EUNAVFOR MED was approved by the
Council on 18 May and the operation launched on 22 June 2015. The mandate of
the operation “disruption of the business model of human smuggling and trafficking
networks in the Southern Central Mediterranean” by “efforts to identify, capture and
dispose of vessels and assets used or suspected of being used by smugglers and
traffickers” (Article 1, Decision 778/2015). “The objective is less to stop migration
flows than to disrupt smuggling routes and capabilities and, hence, reduce the flows
originating from the Libyan Coast, which has been (together with the eastern route)
the main point of departure of migrants coming to Europe” [13]. The operation is
based on the Articles 42(4) and 43(2) of the TEU. The adoption of the Resolution
2240 by the Security Council of the UN, acting under Chapter VII of the UN Charter,
on October 2015, reinforced the legitimacy of EU military operation.
EUNAVFOR MED Sophia Operation is one of the CA’s elements—a complemen-
tary, not a main instrument. Migration had already been implicitly or explicitly men-
tioned in other EU missions and operations (EUFOR ALTHEA, EUCAP SAHEL,
NIGER, EUTM MALI) [15], but for the first time ever an EU military operation is
explicitly and specifically associated with migration issues. It “is the first EU Naval
Force to operate in the Mediterranean, an area of strategic and economic importance
to Europe, and at the centre of many security challenges that affect Europe as a
whole” [16]. This operation reinforced the maritime dimension of CSDP being the
first EU maritime military operation “aimed to deliver sea-based capacity” [17], with
an explicit coercive mandate and the possibility of deploying means on the territory
of a sovereign state (authorized by a UNSC Resolution) without its consent: “while
the EU had so far adhered to the crisis management principles of consent, limited
coercion and relative impartiality for its own CSDP operations, EUNAVFOR MED’s
mandate contains the possibility of the EU going beyond these principles and coming
close to a peace enforcement situation. In and by itself, this constitutes a qualitative
shift in the EU’s security and defence posture” [13].
The operation involves three phases [18]: phase 1 surveillance—“detection and
monitoring of migration networks through information gathering and patrolling on
the high seas” (June–August 2015); phase 2 operational—search operations “board-
ing, search, seizure and diversion on the high seas of vessels suspected of being used
for human smuggling or trafficking” (2A) in international waters (ongoing since 7
October 2015) or (2B) in the territorial and internal waters of Libya (there are still
no legal and political pre-conditions to transit to this sub-phase); phase 3 opera-
tional—disposal practices—“measures against a vessel and related assets, including
through disposing of them or rendering them inoperable, which are suspected of
being used for human smuggling or trafficking”, in the territory Libya (timeline to
be determined).
In June 2016 and July 2017, the Council approved additional (support) tasks:
training of the Libyan Coastguard and Navy and contributing to the implementa-
tion of the UN arms embargo on the high seas off the coast of Libya; setting up
262 A. P. Brandão
a monitoring mechanism of the long-term efficiency of the training of the Libyan

Coastguard and Navy, conducting new surveillance activities and gather information
on illegal trafficking of oil exports from Libya and enhancing the possibility for shar-
ing information on human trafficking with member states law enforcement agencies,
FRONTEX and EUROPOL [19].
Despite the focus on the fight against migrant smuggling and human trafficking,
there is a clear concern to underline the humanitarian dimension of the operation.
The decision of renaming it “Sophia” in September 2015 was justified by Federica
Mogherini as a tribute to “the lives of the people we are saving, the lives of people
we want to protect” and a way “to pass the message to the world that fighting the
smugglers and the criminal networks is a way of protecting human life” [19]. The EU
public diplomacy/communication strategy also emphasizes this humanitarian goal
as demonstrated by the animated video “EU Global Strategy—The story of Sophia”.
According with the Commander of the Operation, “we will remain committed to
saving lives at sea upon request by the competent Maritime Rescue Coordination
Centre (MRCC) and in accordance with the International Law of the Sea” [20].
According with the UN Secretary-General, the EU should focus on saving lives as a
top priority [21].
The Operation Commander commends the “great results” of Sophia Operation and
its evolving scope: it “has transformed into a multifaceted maritime security operation
carrying out a range of tasks from maritime situational awareness to capacity building
and law enforcement while building an extensive network of contacts; all furthering
the achievement of the mission objective” [16]. The operation demonstrated EU
political will and solidarity, being launched in record time and achieving results both
in the security and humanitarian components without collateral damages: arresting
smugglers, destructing boats and tackling the smuggling networks; saving lives. As of
31 December 2016, Operation Sophia had apprehended 101 suspected smugglers and
traffickers, neutralized or destroyed 372 vessels, conducted 222 rescue missions and
rescued at sea 31,899 migrants and carried out 253 arms embargo events. According
with Lieutenant General Wolfgang Wosolsobe, the first phase of the operation had
two main outcomes in terms of military objectives: “Processes had been ‘developed to
gain information from the migrants’, in order ‘to determine not only their particular
circumstances, but how and by whom their passage was effected’; identification
of a network that exists to traffic women and children with a view to their sexual
exploitation”. Both enhanced “understanding of the traffickers’ networks, personnel
and tactics” [22].
Despite those tangible results, Sophia Operation has been subject of criticism for
its ambitious intentions and objectives [23], ambiguity [24], ineffectiveness [14] and
unintended consequences. One of the main critical voices has been the UK House
of Lords, whose latest report is entitled “Operation Sophia: a failed mission” [25]:
the number of arrests piecemeal is modest in comparison to the scale of smuggling;
given the destruction of boats, the smugglers are now using inflatable rubber craft
from China that are more unsafe; the operation had apprehended only low-level
smugglers (and some of them are in fact co-opted migrants), not the major ones
behind the business model; the deterrent effect of migrant flows is not working
since the smugglers find new routes, generally more dangerous and expensive for
the migrants; the search and rescue missions are feeding hopes of safe passage (“a
magnet to migrants”) and having the unintended consequence of “assisting the job
of smugglers, who now only need their boats to reach the high seas, rather than EU
waters”; “serious abuses of the human rights of migrants by the Libyan coastguard”.
In sum, the operation faces “an impossible challenge” [23] and has no meaningful
deterrent and disruption results, being “the wrong tool with which to tackle migration
in the central Mediterranean” [25, 26].
Operation Sophia is a demonstration of the EU narrative and practice on secu-
rity nexuses (defence-migration, military–civilian security, internal–external secu-
rity, border security–human security) that justify the goal of “more security” and
“more actorness”. These nexuses are present in the trend to construct the link between
the CSDP and the AFSJ [11, 12].
The European Agenda on Migration advanced that migration would “become a
specific component of ongoing Common Security and Defence Policy (CSDP) mis-
sions already deployed in countries like Niger and Mali, which will be strengthened
on border management” [27]. It specifically underlined the role of the operations
in the fight against human smuggling: the identification, capture and destruction of
vessels used by criminal networks as “a powerful demonstration of the EU’s deter-
mination to act” [27]. Those tasks constitute the core business of the operation.
The link between maritime border security and migration was initially associated
with FRONTEX operations. FRONTEX is in the forefront of the maritime border
security related with migration since 2005. The main focus has been the control and
surveillance of external borders through EU Joint operations and the forced return of
migrants to the departure state [24]. After the adoption of the Sea Border Regulation,
on May 2014, the operations also encompass search and rescue: “The objective
of Union policy in the field of the Union external borders is to ensure the efficient
monitoring of the crossing of external borders including through border surveillance,
while contributing to ensuring the protection and saving of lives” [28].
One month later, the Council adopted the EU maritime security strategy (EUMSS).
This strategy defines maritime security as “as a state of affairs of the global mar-
itime domain, in which international law and national law are enforced, freedom of
navigation is guaranteed and citizens, infrastructure, transport, the environment and
marine resources are protected” [29]. It encompasses both the internal and external
aspects of EU maritime security. Among others, it identifies the cross-border and
organized crime, including trafficking of human beings and smuggling of migrants,
organized criminal networks facilitating illegal migration, trafficking of arms as a
threat to the security of the EU, its member states and their citizens [29]. EUMSS left
out the humanitarian (search and rescue) dimension. Both operations (FRONTEX
and Sophia) evolve rescue activities, but while FRONTEX operations are mainly
focused on border management; Sophia Operation is dedicated to the disruption
of migrant smuggling and human trafficking routes and capabilities [24]. Despite
Operation Sophia’s humanitarian component, its rationale is mainly security: Oper-
ation Sophia “is also an example of the renewed strategic and political importance
of the nexus between internal and external security in the present European security
264 A. P. Brandão
environment. Beyond the EU borders, terrorist organizations and irregular migration

flows are profiting from instability and unresolved external conflicts, which eventu-
ally have an impact on EU citizens” [10].
In general terms, Rehrl [15] identifies four potential inputs of CSDP to fight cross-
border crime in general, and migrant smuggling and human trafficking in particular,
and to enhance EU borders security, namely: “border surveillance and prevention
of uncontrolled border crossings”; “processing of irregular migrants, in particular
by providing training and technical assistance as well as capacity building for the
so-called hotspots”; “law enforcement activities against smugglers’ networks by
strengthening intelligence sharing”; “security sector reform in countries either of
origin or transit”. In sum, surveillance and intelligence gathering, training and SSR,
three of which are performed by Operation Sophia.
The third axe if the CSDP-FSJ link is multi-actor and multi-policy coordination.
As requested in Article 8 of Council Decision (CFSP) 2015/778, “EUNAVFOR MED
shall cooperate with the relevant member state authorities and shall establish a coor-
dination mechanism, and as appropriate, conclude arrangements with other Union
agencies and bodies, in particular FRONTEX, EUROPOL, EUROJUST, European
Asylum Support Office and relevant CSDP missions”. General Mikhail Kostarakos
(Chairman of the EUMC) [30] and Enrico Credendino (Sophia Operation Comman-
der) [31] converge in the statement that Operation Sophia has been a success in terms
of complex coordination among several EU actors and Italian authorities, and among
military and non-military actors.
EU security actorness has been evolving since the end of the Cold War by a combina-
tion of opportunity, capacity and (global) presence (ambition). This construction has
been accompanied by the comprehensive approach narrative. This holistic approach
is the result of a co-constitutive adequacy: “more security”—appropriation of poli-
cies and instruments of a multifunctional actor for security purposes (security of the
EU and of European citizens); “more actorness”—securitization of issues in order
to promote the actor and its policies. The internal–external security nexus formu-
lated both in terms of threat perception (transnational, cross-border multifaceted
threats) and response to threats, justifies and is justified by the comprehensiveness
of the approach of a post-Westphalian security actor. Its practices, such as Operation
Sophia, demonstrate how security narratives, civilian and military instruments and
securitization dynamics serve convergent processes of gaining political and public
space for legitimizing policies and actions of the European security actor.
Sophia is a pioneering operation in several aspects: the first Naval EU opera-
tion in the Mediterranean and aimed to deliver sea-based capacity; the first CSDP
operation explicitly and specifically associated with migration issues; the first EU
operation with an explicit coercive mandate and the possibility of deploying means
on the territory of a sovereign state (authorized by a UNSC Resolution) without
its consent. The security rationale underlying the operation results of the merging
of several dimensions: border, maritime, military and human security. In the fight
against migrant smuggling and human trafficking, EUNAFOR MED operationalizes
the CSDP-FSJ by being a law enforcement using military assets, carrying out surveil-
lance, intelligence gathering and training activities involving means and structures
of both domains, coordinating actors of both domains (CSDP and FSJ). For this rea-
son, its achievements and failures enlighten the potential limits and effects of IESN
operationalization.
Acknowledgements This paper has been developed in the framework of the collective research
project: “España ante los nuevos retos de la seguridad marítima: Instrumentos y estrategias en
el marco internacional, europeo y peninsular” (DER2016-78272-R), funded by Ministerio de
Economía y Competitividad (Programa Estatal de I+D+i Orientada a los Retos de la Sociedad
2017–2019). The author also acknowledges the financial support of Fundação para a Ciência e
Tecnologia—FCT (Portuguese Science Foundation).
References
1. For the purpose of this paper, the securitisation process was analysed through the lens of the
Copenhagen School (Buzan, B., Waever, O., Wilde, J.: Security: A New Framework for Anal-
ysis. Lynne Rienner Publishers, Boulder, CO (1998)): the security speech act on Operation
Sophia by EU institutions (Council of the EU, European Council, European Commission);
the (perceived) threat (human smuggling and trafficking networks in the Southern Central
Mediterranean); the securitising instrument (the military operation as an extreme version of
securitisation). The analysis of EU actorness was based on the three elements conceptualised
by Bretherton & Vogler: opportunity (“denotes factors in the external environment of ideas and
events which constrain or enable actorness”), presence (“conceptualizes the ability of the EU,
by virtue of its existence, to exert influence beyond its borders”) and capability (“refers to the
internal context of EU external action—the availability of policy instruments and understand-
ings about the Union’s ability to utilize these instruments, in response to opportunity and/or to
capitalize on presence”)—Bretherton, C., Vogler, J.: The European Union as a Global Actor,
p. 24. Routledge, New York (2006). Two security functions were considered in the analysis of
the IESN: compulsion (military operations) and protection (internal security)—Kirchner, E.,
Sperling, J.: EU Security Governance. Manchester University Press, Manchester (2007)
2. See: the European Security Strategy (2003); the Internal Security Strategy for the EU (2010);
the European Agenda on Security (2015); EU Global Strategy (2016)
3. Eriksson, J., Rhinard, M.: The internal–external security nexus. Notes on an emerging research
agenda. Coop. Confl. 44(3), 243–267 (2009)
4. Pastore, F.: Reconciling the Prince’s Two ‘Arms’. Internal-External Security Policy Coordina-
tion in the European Union. Occasional Paper, n. º 30. ISS, Paris (2001)
5. Bigo, D.: Internal and external security(ies): the Möbius ribbon. In: Albert, M., Lapid, Y.,
Jacobson, D. (eds.) Identities, Borders, Orders, pp. 91–116. University of Minnesota Press,
Minneapolis (2001)
6. Rehrl, J., Weisserth, H.-B. (eds.): Handbook on CSDP. Federal Ministry of Defence and Sports
of the Republic of Austria, Vienna (2007)
7. European Council: A Secure Europe in a Better World—The European Security Strategy.
European Council, Brussels, 12/13 December (2003)
266 A. P. Brandão
8. European Council: 2008. Report on the Implementation of the European Security Strate-
gy—Providing Security in a Changing World. European Council, Brussels (11/12) (2008);
European Council: Internal Security Strategy for the European Union: Towards a European
Security Model (2010)
9. Council of the EU: Foreign Affairs Council (2015)
10. Council of the EU: CMPD Food for Thought Paper “From Strengthening Ties Between
CSDP/FSJ Actors Towards More Security in Europe” (2016)
11. Council of the EU: Strengthening Ties Between CSDP and FSJ—Elements of a Draft Road
Map (15562/11) (2011)
12. European Commission and HR: Strengthening Ties Between CSDP and FSJ Actors—Proposals
for a Way Ahead (SEC/2011/560) (2011)
13. Tardy, T.: Operation Sophia Tackling the Refugee Crisis with Military Means. EUISS Brief
(September). EUISS, Paris (2015)
14. Priconi, M.: EU military Operation Sophia: analysing the shortfalls. Sci. Bull. 2(42), 122–127
(2016); Ventrella, M.: The impact of Operation Sophia on the exercise of criminal jurisdiction
against migrant smugglers and human traffickers. Quest. Int. Law 30, 3–18 (2016)
15. Rehrl, J.: Migration and CSDP. In: Handbook on CSDP, vol. 1, 3rd edn, pp. 104–113. Vienna
(2017)
16. EEAS: EUNAVFOR MED—Operation SOPHIA—Six Monthly Report: 1 Jan–31 Oct 2016
(EEAS/2016/1587) (2016)
17. EEAS (European External Action Service): Common Security and Defence Policy of the
European Union: Missions and Operations Annual Report 2016 (2017)
18. Council of the EU: Council Decision (CFSP) 2015/778 of 18 May 2015 on a European Union
military operation in the Southern Central Mediterranean (EUNAVFOR MED). Off. J. Eur.
Union L 122, 31–35 (2015)
19. EEAS: EUNAVFOR MED Operation Factsheet (2017)
20. EDA (European Defence Agency): Operation Sophia to stop human trafficking across the
Mediterranean. Defence Matters 9, 30–31 (2015)
21. European Parliament: Ban Ki-moon on migration: “Saving lives should be the top priority”.
EP News (2015)
22. Wosolsobe in UK (United Kingdom), House of Lords: Operation Sophia, the EU’s naval mission
in the Mediterranean: an impossible challenge. European Union Committee—14th Report of
Session 2015–16. HL Paper (144) (2016)
23. UK (United Kingdom), House of Lords: Operation Sophia, the EU’s naval mission in the
Mediterranean: an impossible challenge. European Union Committee—14th Report of Session
2015–16. HL Paper (144) (2016)
24. Bevilacqua, G.: Exploring the ambiguity of Operation Sophia between military and search
and rescue activities. In: Andreone, G. (ed.) The Future of the Law of the Sea Bridging Gaps
Between National, Individual and Common Interests, pp. 165–189. Springer, Cham (2017).
https://doi.org/10.1007/978-3-319-51274-7_9
25. UK, House of Lords: Operation Sophia: a failed mission. European Union Committee—2nd
Report of Session 2017–19. HL Paper (5) (2017)
26. See also critics from: Amnesty International on human rights violations (Amnesty International:
European Policies Leaving Migrant Lives Adrift in the Mediterranean (2016)); spokesman
of the Libyan Navy, “the first beneficiaries of this operation are not Europe or Libya, but
rather the traffickers” (Euronews: The first beneficiaries of Operation Sophia are the traffickers.
Insiders (21/12/2016)); Parliament of UK (UK 2016) and Operation Commander on intelligence
gathering; Deputy Director of Global Initiative against Transnational Organized Crime on
lack of full transparency (in Mathari, A.: “Mission impossible” for Operation Sophia. Global
Spotlight (02 June) (2017))
27. European Commission: Communication from the Commission to the European Parliament, the
Council, the European Economic and Social Committee and the Committee of the Regions—A
European Agenda on Migration (COM/2015/240). Brussels (2015)
28. Council of the EU: European Union Maritime Security Strategy (2014)
29. Council of the EU: Regulation (EU) No. 656/2014. Off. J. EU L (89) (2014)
30. Kostarakos in: EDA: Interview military capabilities: “Europe still lacks strategic enablers”.
Defence Matters 11, 36–38 (2016)
31. Credendino in: EEAS: EUNAVFOR MED—Operation SOPHIA—Six Monthly Report: 1
Jan–31 Oct 2016 (EEAS/2016/1587) (2016)
The Evolution of EU’s Maritime Security
Thinking
João Almeida Silveira
Abstract Maritime security is a fundamental vector in the security of the European

Union (EU). Such importance resulted in the adoption of several strategies that con-
tribute to the protection and attainment of maritime interests such as freedom and
safety of navigation, maintenance of supply chains, food and energy security, blue
growth, assumption of the EU as a global actor, among others. The EU Maritime
Security Strategy (EUMSS) is the core strategy guiding the EU in the protection of
the seas, yet EU’s thinking extends beyond the EUMSS. This article explores the
process that leads to the formulation of the EUMSS and the changing functions that
maritime security played in the overall strategic objectives of the Union. It finds
that the EU’s maritime security conceptualization was driven by higher levels of EU
engagement in maritime socioeconomic development and marine environment pro-
tection. Subsequent to perceived successes in its maritime security efforts, the field
was instrumentalized by the Union to support and legitimize its security and inter-
national actorness. Maritime security is perceived as an integral part of EU’s global
strategy, thus, there are attempts to mainstream maritime security in EU policies.
Keywords European Union · Maritime security · EU security strategy
1 Introduction
The maritime domain is currently characterized as space of growing levels of com-

petition and human activities (e.g., economic, cultural, scientific) and afflicted by
a great array of threats and challenges. The number of actors operating at sea has
increased (e.g., NGO’s, criminal organizations), adding to the complexity of a domain
profoundly interconnected with state interests [1–3]. Furthermore, the good state of
oceanic ecosystems is linked with the future and well-being of humanity [4, 5].
J. A. Silveira (B)
Department of Political Studies, Faculty of Social Sciences and Humanities—New University of
Lisbon (FCSH-UNL), Avenida de Berna 26C 1069-061, Lisbon, Portugal
e-mail: joaosilveira@campus.fcsh.unl.pt

https://doi.org/10.1007/978-981-13-9155-2_22
270 J. A. Silveira
Gradually, the EU started to play a greater role in maritime affairs and in its pro-
tection. Since the early 2000s, it adopted over 40 maritime related documents that
covered a myriad of aspects, including, among others, safety and security, environ-
ment, blue economy, governance, and scientific and technological advancements.
Such interest resulted from the perception that the size and complexity of sea mat-
ters are better served by a collective European response, rather than through national
policies. The Union attempted to potentiate the exploitation of seas for sustainable
socioeconomic development, in respect of the environment. As the maritime interests
of the EU grew, it became necessary to protect maritime assets [6–9].
This article explores the dynamic process by which the Union became engaged
with maritime security. More, it tracks the evolution of the EU’s maritime security
strategic thinking, in order to mark strategic shifts and/or enlargements of the concept.
It considers as time period the years between 2000 and 2018. It approaches the topic
through the rhetorical analysis of EU official communications.
The article follows a sequential process. First, it explores the early elements of EU
engagement with maritime matters that explain the added importance of the domain
to the Union after 2000. Second, it outlines the main security concerns that gave
rise to the adoption of a maritime security strategy. Third, it outlines the maritime
security thinking of the Union. Fourth, it explains the instrumentalization of maritime
security to the achievement of the global objectives of the Union. And fifth, it pays
attention to attempts to turn maritime security into a mainstream policy.
2 Theoretical Framework and Methodology
The departing question for this article is: How did the EU’s maritime security strategic
thinking evolves between 2000 and 2018? To answer this question, this article uses the
securitization framework, a constructivist framework, whose objective is to analyze
the ascendant or descendant process of topics within political security agendas [10].
The objective is to mark and analyze the several moments of inclusion of maritime
security within the EU political speech.
To that end, the article makes the rhetorical analysis of official maritime and non-
maritime strategies of the EU. When appropriate, it also makes use of secondary
literature to the better frame or explains the content of the official documents.
There are two main concepts in this article: maritime security and strategy. Regard-
ing maritime security, there is no consensual definition [11–13]; thus, for the purpose
of this article maritime security is understood as a broad concept that includes matters
of national security (seapower), marine safety, blue economy, and human security
[3]. Regarding strategy, this article pays attention to the grand strategy of the EU,
and not just strategy. Thereby, it considers all instruments of possible political use,
and not just the link between policy and military power. In doing so, it considers the
aspects related to economy, finance, technology, military, and others [14, 15].
The Evolution of EU’s Maritime Security Thinking 271
3 Economic–Environmental–Social Triangle
Throughout the 2000s, the strategic conceptualization of the EU toward maritime

security increased exponentially. The awakening for a more robust engagement in
sea matters resulted from two traumatic events: the sink of the oil tanker Erika in
December 1999, in the Bay of Biscay and the sink of the oil tanker Prestige in
November 2002, off the northwest coast of Spain.
These events caused great damages to the marine and socioeconomic environ-
ment and threatened the livelihood of coastal populations. As a result, European
populations demanded measures to prevent such occurrences. To respond to pub-
lic pressures, the EU adopted two maritime safety packages in 2000: the Erika I in
March and the Erika II in December. These packages aimed at improving the safety
of seaborne trade (specifically oil) and prevent accidental pollution by ships [6, 16].
When two years later the tanker Prestige sank, it became clear that added efforts
had to be taken. Henceforth, the EU established the European Maritime Safety
Agency (EMSA) to level-up maritime safety in Europe [17].
Adding to the internal efforts, the EU sought to increase global safety standards.
To that end, it engaged multilaterally in forums such as the International Maritime
Organization (IMO) [18] and bilaterally by including maritime safety and security
concerns in agreements with third countries [19].
During the first years of the 2000s, the Union established a link between mar-
itime economic development, social development, and environmental protection
(in Europe and elsewhere). These elements were framed in an interdependent and
intertwined manner. The Thematic Strategy on the Protection and Conservation of
the Marine Environment of 2005 is a good example of how this conceptualization
worked.
According to the strategy, the erosion of the marine ecological capital jeopardizes
employment opportunities and the generation of wealth. A sound ecological state
of European seas is deemed as indispensable for a diverse, sustainable, and thriving
economic development, which is linked with the well-being of populations (e.g.,
food security, employment, maintenance of cultural ways) [20].
4 Blue Economic Development
As a result of technological developments, it became possible to make a deeper

exploitation of marine resources (e.g., blue energy). Consequently, it was important
for the EU to ready European agents (public and private) to fully exploit the poten-
tialities of the oceans. To that end, the European Commission formulated in 2007
the Integrated Maritime Policy (IMP) designed to assist EU agents to address chal-
lenges such as competitivity, climate change, degradation of the marine environment,
maritime safety and security, and sustainability [7].
272 J. A. Silveira
At the core of the IMP are economic development concerns. The IMP is funda-
mentally an instrument to promote the growth of the maritime industry [21]. As an
overarching policy, the IMP provides an holistic and integrated approach to sea mat-
ters, by including areas such as maritime transport; maritime research, surveillance,
and spatial planning; improvement of national policies; resilience against climate
change; reduction of CO2 emissions and ship pollution; maritime labor law; unsus-
tainable fisheries practices; and development of an European network of maritime
clusters [7].
Each of these areas was further developed in sectoral and regional strategies. For
example, the EU approved strategies to guide an ecosystem-based approach to human
activities at sea [22]; to improve marine research, data availability and accessibility,
and general knowledge over the maritime domain [23–25]; to prevent, deter, and
eliminate illegal, unreported, and unregulated (IUU) fishing [26, 27]; to improve
maritime spatial planning [28]; to promote safe and efficient maritime transport
[29]; to improve safety standards [30]; and to address the idiosyncrasies of European
basins, such as the Atlantic, Arctic, Mediterranean, Baltic, or Black sea [31–35].
Alike the IMP, economic aspects are central to these strategies. In terms of eco-
nomic activities, the 2012 Blue Growth Strategy clarified that the EU was especially
interested in five value chains: (i) blue energy; (ii) aquaculture; (iii) maritime coastal
and cruise tourism; (iv) maritime mineral resources; and (v) blue biotechnology.
Improvements in these areas were deemed possible due to advancements in the
fields of robotics, submersible machinery, and other technologies. Blue economy
industries were perceived as a way to generate growth, innovation, and employment
[36].
5 Security Concerns
The EU anchored its sea power on economic aspects. It was, thus, necessary to engage
in maritime security, for there is no economic development without security [2].
Indeed, even if sparse and, at times, in a neglectable matter, several strategies iden-
tified risks and threats to EU interest. Because most maritime strategies addressed
the European space, which was a safe space, most threats related to the matters of
pollution, environmental risks, and IUU fishing [7, 23, 26, 31–35, 37]. Nonethe-
less, the several regional strategies included a greater array of security concerns,
namely: criminality (including trafficking), illegal migrations, marine accidents, need
for more effective law enforcement activities, piracy, legal disputes, among others
[31–35, 37].
Maritime security concerns found echo also in non-maritime strategies, such as
the Counter-Terrorism Strategy of 2005 that called for higher standards on maritime
and ports security [38] and the 2010 EU Internal Security Strategy (ISS) that called
for improvements on EU maritime borders vis-à-vis illegal trafficking [39].
In strategies that covered imminently international and global sectors such as
the strategy for maritime transport, the most pronounced security concerns included
terrorism, piracy and armed robbery at sea, trafficking, stowaways, and smuggling.
From these, piracy was regarded as “the most urgent priority.” In 2009, the year of
approval of the maritime transport policy, the effects of piracy off the Somalia coast
were already pronounced, and the strategy recognized the grave danger of piracy to
seafarers, fishermen, and passengers [29].
In 2003, the European Security Strategy (ESS) had already included piracy as a
potential threat to EU interest. At the time, it was flagged as a threat that could merit
further attention in the future [40]. Events off the coasts of Somalia in 2007 proved
right the inclusion of piracy in the ESS. An upsurge of piracy from 2007 onward
caused harm to Euro-Asian seaborne trade flows and to EU fishing interests in the
region. Furthermore, the instability caused by piracy had consequences on land (e.g.,
impossibility of deliverance of humanitarian assistance). This event demonstrated
that sea insecurity had negative effects to the EU, which prompted action [1, 41–44].
Under the avail of United Nations Security Council (UNSC), the EU launched
operation Atalanta in the region in order to restore stability [45] and protect EU
economic, diplomatic, and humanitarian interests [44]. With operation Atalanta,
the first EU maritime operation within the Common Security and Defence Policy
(CSDP), it became clear that piracy and maritime instability were real threats to the
EU. Consequently, the 2008 report on the implementation of the ESS increased the
attention of the EU to maritime security developments. Besides piracy, the report
also recognized the possible emergence of disputes over trade routes and maritime
territories [46].
The analysis of the above-mentioned strategies suggests that during the 2000s,
the maritime security agenda progressively penetrated through EU policy documents.
Such introduction was motivated by traumatic events as the Erika and Prestige sunk
and by the upsurge of piracy off the Somalia coast. These events forced the EU
to think and engage in maritime security in its proximity and in the wider world.
Maritime security was framed in a dual capacity of an element necessary for the
protection and enhancement of economic, social, and environmental interests. This
perception, aligned with the sprout of European activities at and from the sea, leads
to the adoption of the EU Maritime Security Strategy (EUMSS) in 2014.
6 European Union Maritime Security Strategy
Designed “as a comprehensive framework contributing to a stable and secure global

maritime domain,” the EUMSS addressed internal and external security elements.
It identified the core maritime interests of the EU, namely: the guarantee of the
territorial integrity, sovereignty, and values of member states; the safe development of
infrastructures (e.g., ports and port-facilities, seabed cables, off-shore installations);
the increase of economic activities at sea; freedom and safety of navigation for
ships, cargo, crews, and passengers; the protection of the marine environment; the
protection of European cultural heritage (e.g., archaeologic objects) and EU citizens;
and the use of the maritime domain for scientific progresses and innovations [8].
274 J. A. Silveira
The strategy consolidated the security risks and threats to EU maritime interest,
namely: interstate conflicts; proliferation; sea denial or sea obstruction; transnational
criminality; terrorism; environmental threats (natural or man-made events) [8].
To tackle maritime security challenges, the EUMSS formulated a comprehensive
and cross-sectoral approach leading to higher levels of internal–external coherence
of the EU. EU actions were to be executed in compliance with the law (national,
EU, and international), in a cost-efficient manner, and bridging military and civilian
authorities and actors,1 local, national and EU agencies, and the maritime industry
[8].
While calling for greater autonomy of EU actions, the strategy asserted the full
respect for the functional integrity of member states competencies, rights, and inter-
ests. Furthermore, the autonomy of action did not equate to isolationism. Instead,
the EUMSS directed toward a multilateral approach, anchored on the strengthening
and support of regional responses [8]. Autonomously, but next to partners,2 the EU
aspired to engage in maritime capacity building, in areas such as maritime gover-
nance and the rule of law; development of the port and maritime transport standards;
borders management; and combat of IUU fishing [8].
In view to the implementation of the EUMSS, the Union approved in December
2014 the EU Maritime Security Strategy Action Plan (EUMSSAP); a holistic plan to
prompt the comprehensive and cross-sectoral approach of the EU. The actions were
divided across five workstrands that addressed specific but intertwined aspects:
Workstrand 1—External Action. Addressed matters of external coordination
between the EU and other international partners. It sought to increase the visibil-
ity of the EU; to foster capacity building in third countries, the peaceful settlement
of disputes, and the international law; to pursued force development for contingency
operations; and to articulate measures to increase the preparedness of EU and mem-
bers states to engage in maritime security contingency operations [47].
Workstrand 2—Maritime Awareness, Surveillance and Information Sharing. Pur-
sued the implementation of the Common Information Sharing Environment (CISE)
to further cross-sectoral cooperation and interoperability at national and EU level. It
addressed cross-border cooperation and information exchange to optimize surveil-
lance; and supported the coordination of the several EU surveillance initiatives [47].
Workstrand 3—Capability Development. Fostered technological development,
harmonization of instruments to favor interoperability, and standardization and certi-
fication in the maritime domain. It was interested in exploring dual-use and multipur-
pose capabilities (e.g., civil–military capacities, integration of cyber dimension into
the maritime domain) and in promoting the sharing of best practices, risks analysis,
and threat information [47].
1 E.g., law enforcement, border control, customs and fisheries inspection, environmental authorities,
maritime administration, research and innovation, navies or other maritime forces, coast guards,
intelligence agencies [8].
2 E.g., North Atlantic Treaty Organization (NATO), the United Nations (UN), the African Union
(AU) or the Association of Southeast Asian Nations (ASEAN).

Workstrand 4—Risk Management, Protection of Critical Maritime Infrastructures

and Crisis Response. Promoted EU joint risk analysis; the enhancement of cross-
sectoral and cross-border cooperation for crisis response and contingency planning;
the assessment of the resilience of maritime transport infrastructures against natural
or man-made disasters; and promoted mutual understanding and interoperability
among maritime security actors [47].
Workstrand 5—Maritime Security Research and Innovation, Education and Train-
ing. Promoted the establishment of knowledge and training networks; fostered
civil–military maritime research (e.g., though public–private partnerships); and pro-
moted inter-agency joined-up exercises [47].
The EUMSS and its action plan provided a coherent approach to maritime secu-
rity. The documents balanced internal and external aspects of security and bridged
maritime and non-maritime aspects of the Union’s action. Besides directing the EU’s
maritime strategy toward the protection and enhancement of strategic interest, the
EUMSS seems to work as a supporter of the EU’s international actorness.
7 Instrumentalization of Maritime Security
In 2016, the International Ocean Governance Agenda (IOG) continued the pursuit of
ensuring EU’s coherence, by bridging several features of maritime security. On the
document, the EU presented itself as the global champion of oceans sustainability
and protection. To build its case, the EU anchored its thought on internal and external
progress and action it had developed thitherto [48].
On the internal front, it deferred to the developments within the IMP that con-
tributed to an overall increase of oceans sustainability. Furthermore, it placed great
focus on the environmental aspects of its policies. On the external front, the EU’s
contribution to international maritime security was the main argument. It notably
called attention to EU’s action in thwarting piracy in the Horn of Africa and its
anti-smuggling activities in the Mediterranean Sea [48].
The agenda was built around the idea of the oceans as a common good. It mim-
icked the EUMSS comprehensive and multilevel approach to ocean governance and
sustainability by including several sectors of the maritime economy; by establish-
ing links with nonexclusive maritime aspects of security such as cyber security; by
making an effort to consolidate and advance the internal/external nexus of EU; and
by pushing cultural components into the agenda [48].
Overall, environmental features were very much embedded into the agenda, and it
established a close link between economic and environmental dimensions. It asserted
that only a sustainable use of maritime resources could lead to long-term growth.
According to the EU, the environmental and economic link is particularly relevant and
requires global and multilateral action [48]. This indicates that the ideas formulated
in the early 2000s made a school in EU’s maritime conceptualization.
Also in 2016, the Global Strategy for the European Union’s Foreign and Security
Policy integrated maritime security within the broader EU approach to its foreign
276 J. A. Silveira
and security policy. The document presented the EU as a global maritime security
provider. According to it, maritime global growth is linked with EU interests of
supporting an open and fair economic system; thus, it is necessary for the EU to take
an active role in its protection. The strategy reinforced a multilateral approach that
perceived maritime security as a matter of external cooperation (e.g., with NATO)
and that required building capacities in Europe and elsewhere [9].
Again, the support for an international order based on the rule of law was present,
with the defense of the law of the seas as an important instrument for the peaceful
settlement of disputes. Also in conceptualizing its relations with others regions, the
Union included maritime security as one of the topics for joint development [9].
In its path toward coherence, the strategy argued for greater coordination between
internal and external agencies and instruments, such as CSDP missions and opera-
tions, European Border and Coast Guard, and EU specialized agencies to enhance
border protection and maritime security, to save lives, fight cross-border crime, and
disrupt smuggling networks. To increase the EU’s preparedness and effectiveness,
the strategy called for the improvement of maritime security-related capabilities [9].
These two documents seem to portrait maritime security not just as a supporter
element of EU’s internal and external actorness, but also as a legitimizer of EU’s
action at home and abroad. Consequently, maritime security serves to protect and
enhance the economic, environmental, and social interests of the EU and to support
and legitimize the security and international actorness of the EU. Maritime security
is, thus, not just a strategic component of EU’s maritime actorness, but also an active
component of EU’s global strategy.
8 Mainstream Maritime Security
The global strategy of the EU of 2016 is the core document guiding the EU toward a
more coherent and coordinated action. It argued for a holistic approach of the Union to
which all policies and instruments should contribute. The document re-focused EU’s
attention to the European continent and to the internal adaptation to cope challenges
faced by the EU as, for instance, the Russian behavior that led to the annexation
Crimea in 2014 and the effects of the migration crisis in the Mediterranean [9].
In 2017, the Council Conclusion on Global Maritime Security incorporated
such re-focusing. Overall, the conclusions reaffirmed the objectives, interests, and
approach of the EUMSS, yet it established as areas of particular attention the Mediter-
ranean Sea, the Baltic Sea, the Black Sea, and the Arctic region [49].
Regarding the Mediterranean, the document pushed for increasing levels of coor-
dination between EU instruments such (e.g., EUNAVFORMED Operation Sophia,
Frontex, European Fisheries Control Agency, EMSA) and between EU instruments
and other actors operating in the Mediterranean in order to increase security levels,
namely, to respond to the migration influxes from North Africa [49].
Oddly, however, the effects of the Russian annexation of Crimea to the maritime
security of the EU were not addressed, and the document only refers to regional
cooperation in its approach to the Baltic and the Black Sea basins [49]. Considering
that the annexation of Crimea considerably changed the security spectrum of Europe
and that the existent strategies for those basins predate the annexation, it is a consid-
erable gap the absence of further strategic guidance to address the changing regional
order.
Despite the European re-focus, the strategy also considered relevant to engage
in confidence-building measures in the South China Sea, which the EU regards as
an area of potential regional conflicts. Furthermore, the document also demonstrates
attention to the situation in the Gulf of Guinea. In both cases, the EU’s approach is
based on supporting local ownership and regional institutions [49].
In 2018, the action plan for the EUMSS was updated. The principles of cross-
sectoral approach (civil–civil, civil–military, and military–military cooperation),
functional integrity, respect for rules and principles, and maritime multilateralism
of the EUMSS were maintained, as well as the defined approach to maritime secu-
rity [50].
The document presented a total of 90 actions. Those were divided between the
horizontal issues (56 actions) and regional and global affairs (34 actions).
The section horizontal issues devises the actions to the fulfillment of the EUMSS
and is subdivided in six main areas:
– Awareness raising. Promotes awareness regarding the EUMSS.
– International cooperation. Promotes the mainstream of maritime security in the
EU’s foreign action. The objective is to enhance cooperation with other parties
(e.g., organization, third parties, private sector), to promote international law, to
foster capability building in regions that require, and to bridge internal and external
components of the EU in order to increase its the internal–external coherence.
– Maritime awareness. Promotes the increase of maritime awareness through the bet-
ter use and sharing of information produced by several EU agencies, programmes,
activities, and policies (e.g., MARSUR, SatCen, CISE, EMSA, ESA).
– Capabilities development, research, and innovation. Promotes research and devel-
opment of capabilities and technologies of both the civilian and military sectors,
through synergies and multiple EU-funded grants. The objective is to reduce the
fragmentation of EU security demand. There is a great tonic in the development
of dual-use technologies for multipurpose use. In this regard, the enhancement of
maritime data is deemed crucial.
– Risk awareness and management. Promotes a comprehensive approach to risk
management and common analysis. The objective is to increase the resilience of
maritime infrastructures.
– Education and training. Promotes training and education initiatives to increase the
readiness of operations in tackling maritime issues. The objective is to standardize
procedures within the EU, and if possible, integrate external partners.
Regional and global affairs address specific needs of maritime areas in Europe
and in the world. From the 34 actions defined only nine address areas outside the
European geography: Gulf of Guinea, the Indian Ocean, and Indian and Pacific Ocean
[50]. This division appears to confirm the EU’s geographic re-centering to Europe.
278 J. A. Silveira
The actions defined for regional and global maritime affairs pursuit six funda-
mental objectives: mainstream maritime security in EU sea basins policies (existent
and future); increase cooperation and coordination among EU agencies, projects,
and instruments, and between these and actors outside the EU; develop capabilities
in third countries; develop EU capabilities and encourage research and development
projects; increase awareness of maritime security; and support regional development,
both in the EU and in the wider world [50].
Overall, the Council Conclusion on Global Maritime Security and the new action
plan for the EUMSS do not significantly alter the EU’s maritime security strategy.
These documents do re-center the attention to Europe, but maintain an international
take, and the objective of asserting the EU at the global stage. Maritime security
remains an important piece of EU’s conceptualization, for it “is an essential precondi-
tion for development, job creation, research in the maritime and marine environment
and global oceans governance” [49].
The more inward approach that these documents seem to convey appears to be a
maneuver to strengthen the EU at home, so that the EU can project a stronger and
more coherent image abroad.
9 Conclusion
Maritime security observed a gradual and incremental evolution within EU strategic

conceptualization between 2000 and 2018. At the core of the EU, strategic thinking
is the economic–environmental–social triangle that drives and supports deeper levels
of EU engagement with maritime security. It was the growing weight of economic
and social interest that prompted the EU into a more active role in maritime security
in order to protect and enhance its interests.
In its engagement, the Union follows a comprehensive approach that links internal
and external elements. More are the acts within the international community, and in
partnership with third countries. Maritime security is formulated as a task to be
executed in a multilateral setting and requires the level-up of maritime capabilities in
Europe and elsewhere, as well as added instruments for cooperation and coordination.
With perceived successes in its maritime and maritime security policy, the Union
instrumentalized maritime security in order to support and legitimize its security and
international actorness. Progressively, the EU brought maritime security objectives
and instruments to the service of overall EU strategic objectives. There is a notorious
attempt to mainstream maritime security into the EU’s strategic speech, in order to
increase the coherence of the Union, and consolidate its position in the global stage.
References
1. Behr, T., Brattberg, E., Kallio, J. et al.: The Maritime Dimension Of CSDP: Geostrategic
Maritime Challenges and Their Implications for the European Union. European Parliament,
Directorate-General for External Policies of the Union, Directorate B, Policy Department
(2013)
2. Germond, B.: The Maritime Dimension of European Security: seapower and the European
Union. Palgrave Macmillan, Houndmills, Basingstoke, Hampshire (2015)
3. Bueger, C., Edmund, T. (2017). Beyond Seablindness: a New Agenda For Maritime Security
Studies. Int. Aff. 0, 1–19
4. Noone, K., Sumaıla, R., Díaz, R. J.: Valuing the Ocean. Extended Executive Summary. SEI -
Stockholm Environment Institute, Stockholm (2012)
5. Figueres, J.M., Lamy, P., Podesta, J.D.: The Fight for Ocean Health. In: Project Syn-
dicate https://www.project-syndicate.org/commentary/un-sdg-14-ocean-regeneration-by-jose-
maria-figueres-et-al-2017-02(2017). Accessed 28 Feb 2017
6. European Commission: Maritime Safety, Erika I Package (2000)
7. European Commission: An Integrated Maritime Policy For The European Union. Office for
Official Publications of the European Communities, Luxembourg (2008)
8. European Commission: European Union Maritime Security Strategy (2014)
9. High Representative: Shared Vision, Common Action: a Stronger Europe – A Global Strategy
for the European Union’s Foreign and Security Policy (2016)
10. Buzan, B., Waever, O.: Regions and Powers. The Structure of International Security. Cambridge
University Press, Cambridge (2003)
11. Chapsos, I.: Introduction: Strengthening Maritime Security Through Cooperation. NATO Sci.
Peace Secur. Ser. E. Hum. Soc. Dyn. pp. 1–7. https://doi.org/10.3233/978-1-61499-536-4-
1(2015)
12. Bueger, C.: What is maritime security? Mar. Policy 53, 159–164 (2015). https://doi.org/10.
1016/j.marpol.2014.12.005
13. Chapsos, I., Malcolm, J.A.: Maritime security In Indonesia: towards a comprehensive agenda?
Mar. Policy 76, 178–184 (2017). https://doi.org/10.1016/j.marpol.2016.11.033
14. Gray, C.: Strategy and History. Essays on Theory and Practice. Routledge, New York (2006)
15. Berindan, I.: Not another ‘Grand Strategy’: what prospects for the future European security
strategy? Eur. Secur. 22, 395–412 (2013). https://doi.org/10.1080/09662839.2013.808189
16. European Commission: Maritime Safety, Erika II Package (2000)
17. European Parliament, Council of the EU: Directive 2002/84/EC of the European Parliament
and of the Council of 5 November 2002 Amending the Directives on maritime Safety and the
Prevention of Pollution from Ships (2002)
18. European Council: Copenhagen European Council 12 and 13 December 2002. Presidency
Conclusions (2002)
19. European Commission: European Neighborhood Policy - Strategy paper (2004)
20. European Commission: Thematic Strategy on the Protection and Conservation of the Marine
Environment (2005)
21. Germond, B.: The EU’s security and the sea: defining a maritime security strategy. Eur. Secur.
20, 563–584 (2011). https://doi.org/10.1080/09662839.2011.635648
22. European Parliament Council of the EU: Directive 2008/56/EC Of The European Parliament
and of the Council of 17 June 2008 Establishing A Framework For Community Action in the
Field of Marine Environmental Policy (Marine Strategy Framework Directive) (2008)
23. European Commission: A European Strategy for Marine and Maritime Research (2008)
24. European Commission: Marine Knowledge 2020 Marine Data and Observation For Smart and
Sustainable Growth (2010)
25. Council of the EU: Council Conclusions on Integration of Maritime Surveillance. Towards The
Integration of Maritime Surveillance: a Common Information Sharing Environment for the EU
Maritime Domain (2011)
280 J. A. Silveira
26. Council of the EU: Council Regulation (EC) No 1005/2008 of 29 September 2008 Establishing
a Community System to Prevent, Deter and Eliminate Illegal, Unreported and Unregulated
Fishing, Amending Regulations (EEC) No 2847/93, (EC) No 1936/2001 and (EC) No 601/2004
and repealing Regulations (EC) No 1093/94 and (EC) No 1447/1999 (2008)
27. European Commission: Commission Regulation (EC) No 1010/2009 of 22 October 2009 lay-
ing down detailed rules for the implementation of Council Regulation (EC) No 1005/2008
Establishing a Community System to Prevent, Deter and Eliminate Illegal, Unreported and
Unregulated Fishing (2009)
28. European Commission: Roadmap for Maritime Spatial Planning: achieving Common Princi-
ples in the EU (2008)
29. European Commission: Strategic Goals and Recommendations For The EU’s Maritime Trans-
port Policy Until 2018 (2009)
30. European Parliament, Council of the EU: Directive 2009/18/Ec Of The European Parliament
and Of The Council of 23 April 2009 Establishing The Fundamental Principles Governing The
Investigation of Accidents in the Maritime Transport Sector and Amending Council Directive
1999/35/EC and Directive 2002/59/EC of the European Parliament and of the Council (2009)
31. European Commission: Black Sea Synergy - A New Regional Cooperation Initiative (2007)
32. European Commission: European Union Strategy for the Baltic Sea Region (2009)
33. European Commission: Towards an Integrated Maritime Policy For Better Governance in the
Mediterranean (2009)
34. European Commission: Developing a Maritime Strategy for the Atlantic Ocean Area (2011)
35. European Commission: An Integrated European Union Policy For The Arctic (2016)
36. European Commission: Blue Growth: opportunities For Marine and Maritime Sustainable
Growth (2012)
37. European Commission: Action Plan For A Maritime Strategy In The Atlantic Area. Delivering
Smart, Sustainable and Inclusive Growth (2013)
38. Council of the EU: The European Union Counter-Terrorism Strategy (2005)
39. European Commission: The EU Internal Security Strategy in Action: Five Steps Towards a
more Secure Europe (2010)
40. European Council: A Secure Europe in a Better World. European Security Strategy (2003)
41. Weber, A.: EU Naval Operation in the Gulf of Aden (EU NAVFOR Atalanta): Problem
Unsolved, Piracy Increasing, Causes Remain. In: The EU as a Strategic Actor in the Realm of
Security and Defence? pp 70–83. SWP, German Institute for International and Security Affairs,
Berlim (2009)
42. Larik, J.: Europe’s Fight Against Piracy: from the Barbary Coast to Operation Atalanta. Fon-
dation Pierre du Bois (2013)
43. Silveira, J.: EUNAVFOR Atalanta: O Compromisso da União Europeia com a Segurança
Global. Universidade do Minho (2015)
44. Silveira, J.: European Union Security: the Threat of Piracy. In: VIII Congress APCP (Portuguese
Association of Political Science). Faculdade de Ciências Sociais – New University of Lisbon,
Lisbon, Portugal (2016)
45. Council of the EU: Council Joint action 2008/851/CFSP of 10 November 2008 on a European
Union military operation to contribute to the deterrence, prevention and repression of acts of
piracy and armed robbery off the Somali coast (2008)
46. European Council: Report on the Implementation of the European Security Strategy. Providing
Security in a Changing World (2008)
47. Council of the EU: EU Maritime Security Strategy Action Plan (2014)
48. European Commission: International Ocean Governance: an Agenda For The Future of Our
Oceans (2016)
49. Council of the EU: Council Conclusions on Global Maritime Security (2017)
50. Council of the EU: Council Conclusions on the Revision of the European Union Maritime
Security Strategy (EUMSS) Action Plan, 26 June 2018
The Obstacles Women Face in Gaining
Access to Special Operations Forces
João Reis, Rafael Gonçalves, Sofia Menezes and Manuela Kaczynska
Abstract The aim of this article is to provide exploratory insights regarding women
integration in the special operations forces. The methodological approach is qualita-
tive and employs a case study research in the Portuguese Army. It includes multiple
sources of data collection for corroboration purposes, namely semi-structured inter-
views, direct observation and official documents. Although it has been considered a
hot-topic in the Portuguese armed forces, and as the Portuguese Republic establishes
the legal means to integrate women in all professional areas, reality shows that up to
the present days no women have been part of the special operations forces. In light
with the above, we evidence some obstacles that women are facing in gaining access
to the Portuguese special operations forces, and we present Portugal in comparative
perspective with two successful case studies.
Keywords Women · Special operations forces · Portuguese Army · Military ·

Gender · Integration · Case study
J. Reis (B)
Department of Military Science, Military Academy, GOVCOPP, Lisbon, Portugal
e-mail: joao.reis@academiamilitar.pt
R. Gonçalves
Department of Military Science, Military Academy, Lisbon, Portugal
e-mail: goncalves.inf@gmail.com
S. Menezes
Department of Management and Leadership, CINAMIL, Lisbon, Portugal
e-mail: sofia.menezes@academiamilitar.pt
M. Kaczynska
Tadeusz Kościuszko Land Forces Military University, Polish Land Forces Military University,
Wrocław, Poland
e-mail: manuelakaczynska@gmail.com

https://doi.org/10.1007/978-981-13-9155-2_23
282 J. Reis et al.
1 Introduction
Special operations forces have become increasingly central to the wars of the twenty-
first century [1]. Every country has military units composed of personnel specially
trained and equipped, with higher level of readiness than general military units and
typically regarded as elite—these forces are collectively known as special operations
forces [2].
Despite consistent downsizing, over the past two decades the armed forces of
the industrial democracies have seen extensive growth in special operations forces
(SOF). By means of increasing numbers of personnel and of more frequent deploy-
ments, SOF units have wielded considerable influence in conflicts around the world
[3]. The remarkable growth of SOF is part of the military organizations and has
had to dramatically adapt to their diversity and changing external and internal envi-
ronments [4]. As Shamir and Ben-Ari [4] argue, SOF are unique in that they are
specialized generalists, local-level integrators that link between the tactical opera-
tional and strategical levels of actions, therefore being hybrid forms of organizational
response to environmental pressures.
In Portugal, women are not precluded from serving in any military unit by law;
however, the reality is that up to the present days no women have been part of
the Portuguese SOF. The reasons are yet to be determined, since all the statistics
prove the commitment of women to combat—“in approximately 10 years of combat
operations in Iraq and Afghanistan, over 283 thousand female members have been
deployed, over 800 have been wounded and over 130 have died” [5, p. 1]. Burrelli
[5] reinforces that, on numerous occasions, women have been recognized for their
heroism, two of which earning US Silver Star medals.
We have conducted a research of the literature and we have detected the existence
of similar articles. What is the novelty brought by our article? Most gender studies
on special operations forces are from the United States (US) Army, mainly after the
congress mandated that US SOF adopted gender-neutral occupational performance
standards by 2016. After that year, females were no longer barred from applying to
SOF rigorous pipeline, and as shown in operations in Afghanistan and Iraq, female
enablers made significant, unique contributions to ground combat and SOF missions
[6]. Gasca et al. [6] even went further, by arguing that many of these unique con-
tributions, such as the ability to interact with indigenous children and female adults
(forbidden by local customs or religious beliefs from interacting with male service
members), were critical to mission success. Moreover, women have been employed
successively in SOF units worldwide, though some Western countries still subsist to
that trend. It is to that extent that our article provides differentiating contribution, as
it analyses the reasons that led a Western European Army to not embrace women in
their ranks.
This article is organized as follows: in the next sections, we review the litera-
ture and discuss the gender concept on the military forces. Then, we describe the
methodological approach, the study sample and the data analysis.
The Obstacles Women Face in Gaining Access to Special … 283
The following sections discuss the empirical results, by providing statements from
the respondents. Finally, we present the academic and practical insights, limitations
and guidelines for future research.
2 Literature Review
The issue of women in the military has always been, and is likely to remain, one
of the most controversial issues in the study of military institutions [7]. Therefore,
the issue of women’s empowerment and gender equality is at the top of agendas
across the world, as gender inequality is widespread in all cultures [8]. By the begin-
ning of the twenty-first century, most Western democratic countries had admitted
and increased the number of women in their armed forces; although restrictions
persisted, many were lifted: women were progressively allowed to enter military
academies and given access to a wide variety of positions and functions [9]. As the
global momentum towards the acceptance of women in ground close combat (GCC)
continues to increase, the USA decided, in December 2015, to open all GCC role
to women without exception—a controversial and, some say, politically motivated
decision [10]. The number of women service in the military and deployed to active
duty is unprecedented in the history of developed economies—in the US Armed Ser-
vices, women constitutes approximately 14.5% of the 1.4 million active component
and 18% of the 850,000 reserve component [11]. Yet, little attention has been paid
to women integration on military life. Thus, the history of women’s military partic-
ipation shows that they have systematically been excluded from warfare as warriors
or have participated only in exceptional circumstances [7].
In the USA, between 2003 and 2013, women were technically banned from direct
assignment to ground combat units; however, the US military deployed all-female
counterinsurgent teams in Iraq and Afghanistan [10, 12]. Greenburg [12] also argues
that in various forms, these teams searched Iraqi women at checkpoints and in home
raids, provided medical assistance to Afghan women and children. Adding to this,
women participated in highly combative special operations missions alongside Army
Rangers and Green Berets in Afghanistan. Although this empirical evidence has been
demonstrated, women are still being marginalized in several military institutions
around the world.
The gender theme has raised attention all over the world, and it is not just about the
USA. In 2016, the Canadian Armed Forces (CAF) launched a recruitment campaign
aimed specifically at women, in which the military’s—and society’s—ambivalence
towards female participation was palpable—one of the goals of the campaign was
to recruit sufficient women so as to change the CAF’s violent, sexist and sexualized
culture [13]. The stereotype has made women’s integration in the armed forces chal-
lenging and supports a gender binary “that characterizes men as active, women as
passive” and is damaging their credibility as combat soldiers [13]. The reality is that
many nations still do not allow women to serve in combat roles or in submarines
[14]. Still, a greater number of combat-support and possibly service-support roles
284 J. Reis et al.
performed by women will increasingly mean they will find themselves in combat
zones [10].
Szayna et al. [15] have recently conducted a survey about women integration in
SOF and concluded an opposition to opening SOF specialties to women in both deep
and wide, with high levels of opposition across all SOF elements. Those authors’
analysis of responses to open-ended questions suggests that this opposition also is
deep-seated and intensely felt. Szayna et al. [15] concluded that principal sources
of this opposition are the belief among SOF personnel that women do not have the
physical capabilities and other to meet the demands of the specialties; however, the
lower level of opposition to women in SOF units than specialties, as well as the
fact that about four in ten of our respondents agreed that women might be helpful
in conducting sensitive operations and communicating with local populations might
present additional opportunities for the participation of women in SOF. Sunde and
Kristiansen [16] argue in the same direction, stating that women in certain urban spe-
cial reconnaissance (SR) roles will increase the SR capacity in small states NATO
SOF. As the authors studied the Norwegian Special Operations Forces (NORSOF)
case, the found that conducting SR in urban areas often involves indirect or direct
human contact. Female SR operators will tear up the stereotypical signature of male
operators, and thereby provide operational flexibility and will enable a SOF unit
to handle a broader range of SR missions, consequently increasing the strategic and
political value of that unit [16]. Meanwhile, Knarr et al. [17] conducted a survey with
Retired SOF Senior Leaders who agreed that any chance of success requires no com-
promise in the assessment and selection standards across the SOF enterprise (e.g. “Do
not lower the standards for physical and mental endurance”, “Keep the standard!”),
this followed a focus group concerned with the ability of mixed-gender teams to
effectively perform direct action missions, perceived to require the highest level of
physical capabilities. From a NATO SOF integration, Kristiansen [18] suggests to
analyse the different roles and to produce a selection and education programme that
suits the standards for these specific roles, instead of forcing women through the
SOF selection and education programmes that already exists. In Kristiansen’s [18]
view, the physical standards for women NATO SOF need to be adapted, based on
potential relevant roles for female operators, such as the level of risk in NATO SOF
is expected to be handled in order to accomplish high-risk missions.
3 Methodology
This research is qualitative and exploratory, trying to explain a phenomenon for which
little or no empirical data exist [19]. To that end, we have conducted a case study to
enable the research to acquire in-depth and holistic understanding of multiple aspects
of the phenomenon, as well as the interrelationships between different aspects [20].
We have built this study on multiple sources of data collection, for triangula-
tion and corroboration purposes. Those sources of data collection consist on semi-
structured interviews, direct observations and institutional documents. A case study
that relies on multiple sources prevents the exclusive reliance on a single data col-
lection method, thus aiding to neutralize any bias inherent to a particular data source
[21].
The participants for the interviews were selected according to the following char-
acteristics: six servicemen with the army special operations course and six service-
women from the Portuguese Army. The researchers made use of their personal net-
work in the army to identify the best respondents. The selected respondents were
chosen according to their different functional areas and different levels of responsi-
bility in the military organization. Thus, the study was carried out from operational
to administrative areas, and the respondents’ ranks ranged from master sergeant to
colonel. The direct observations involved systematically seeing and listening [22] and
mainly consisted in informal interviews and notes that were collected into a research
diary. Those notes contained the respondents’ reactions and expressions to certain
type of questions, while the informal interviews had mainly served to deepen the
knowledge that otherwise would not possible to ascertain during the formal inter-
views. The analysis of institution documents served essentially to understand the
Portuguese concept of special operations forces, their main missions and to under-
stand into what extent the training/course can be an obstacle to women integration.
Similar to other studies, before proceeding, we took the following details in con-
sideration: first, the women in SOF survey dealt with a highly politicized issue in
Portugal as respondents might have viewed the survey as an influence to policy,
rather than an opportunity to identify the potential challenges and opportunities for
smoothing implementation and to change mentalities [15]. Therefore, we believe
some respondents might not have answered in an overt way—to mitigate a possible
disruption, we have cross-checked these answers with secondary sources of data col-
lection. In order to enhance the reliability and validity of the case study research, we
have followed a research protocol (Appendix) and the transcriptions were double-
checked by the participants to avoid misinterpretations.
After we collected the data, all the text was analysed according to the technique of
content analysis [23]. We had preliminary categorized textual data into categories, to
identify consistent patterns and relationships between variables in a way of reducing
data and making sense of them [21]. The study was resorted to the use of qualitative
data analysis software known as NVivo, which allowed to code in order to build
and hierarchize categories and subcategories to identify the main stream and/or new
ideas.
4 Findings
This section presents the results of the case study research. In addition, we also
provide an insightful contribution, by comparing two European realities, Portuguese
and Polish, from military point of view of the authors. Finally, we present an inspiring
case, witnessed from a military mission in Afghanistan.
286 J. Reis et al.
4.1 The Case Study of the Portuguese Special Operations

Forces
The special operations forces have a complex organization, a diverse set of capa-
bilities and a broad range of officially assigned missions, all of which can make it
difficult to understand exactly who special operation forces are and how they should
be used [24]. In Portugal, the army special operation forces are especially assigned to
conduct operations of special reconnaissance, direct actions and military assistance.
Moreover, the directive nº 90 from the Chief of Staff of the Portuguese Army (CSPA)
states that, due to SOF unique characteristics, they act within the spectrum of crisis
response operations, combat operations for search and rescue and counterterrorism
operations [25, pp. 2–4].
The respondents have supported the ideal of combat-support and service-support
roles to be performed by women in the special operations forces. At the same time ser-
vicewomen argued that women might be helpful in conducting sensitive operations
with local population, they have also claimed that if women gather enough physi-
cal and psychological robustness, they should be included in SOF combat missions.
In addition, informal interviews have revealed that until very recently the physical
component of SOF in Portugal was highly valued; this information is also corrobo-
rated with internal documents. However, this paradigm has been changing, while the
technical and intellectual component is highly appreciated, e.g. in order to perform
complex missions that involves modern military technologies [26]. Although some
scholars and SOF servicemen make a distinction about different physical capabilities,
there are no scientific evidence concerning technical and intellectual discrepancies
between men and women. Furthermore, a survey in Portugal shows that currently
the number of women attending higher education in Portugal surpasses that of men
[27].
One of the possible solutions to integrate women into the Portuguese special oper-
ations forces is what servicewomen call “pioneers”, that is, a group of women being
able to successfully undertake a special operations course and combat operations
alongside with men. Demystifying the subject, thrust and inspiration may be the
formula that women need to undertake a life in SOF units. The interviewed service-
women are well informed about the progress made until the present days in both US
and Afghan SOF. If women are willing to attend and complete a special operations
course, they will encourage and inspire others to move forward as they have proven
that this opportunity is within the reach of every person. This opportunity is getting
closer, because although official documents show that the special operations course
is still very demanding from the physical point of view, it is increasingly becoming
more technical, whereas the adaptability, the complex technical and tactical capa-
bilities of SOF students are pushed to limit. Nevertheless, direct observations have
pointed out that some servicewomen are reluctant about the robustness level that
would be imparted to them during the SOF course, when compared with their male
counterparts, in order to make them give up. Apparently, the idea of SOF operators
is that integration of women into the special operations forces might be identified
as a reduction of the parameters of physical robustness, which in their mind could

place the Portuguese SOF in a situation of fragility when compared with other special
forces (e.g. Commandos).
The Portuguese paratroopers have a different understanding—they are considered
a special force, but they also integrate women in their ranks. The Portuguese armed
forces included women for the first time in 1961, when Isabela Bandeira de Mello,
the first civilian parachutist women in Portugal, influenced the armed forces to create
an all-female corps of nurses to be deployed to the colonial war [28]. The first women
were then trained as parachutists, with the mission of providing nursing assistance
during the war in Portugal’s overseas territories in Africa and served in combat areas
in Angola, Mozambique and Guinea-Bissau during the 13-year colonial conflict
[29]. On numerous occasions, these female paratroopers had been recognized for
their heroism: an example is Maria Justino da Silva, who was wounded in combat in
Mozambique, during an evacuation mission to rescue wounded soldiers [30]. Since
those days, women from the paratrooper forces had been well accepted and integrated
in all types of military operations.
We also considered valuable to present two successful cases and thus to present
Portugal in comparative perspective. The next cases were selected due to the authors’
affinity with those realities, mainly due to their professional experiences.
4.2 Comparison with European Countries—The Polish

Model
Currently, Poland employs women in close combat roles [31]. Nowadays, women that
complete 5 years of university assume military posts [32]. According to Cawkill et al.
[32], from 2003 to 2008 the Polish armed forces made outstanding progresses, as they
achieved 1153 servicewomen in the army; moreover, females platoon commanders
have been deployed to Polish military contingents in Iraq and Afghanistan. Although
the Polish Army is given as an example, several European countries integrate women
into combat operations—Denmark, Finland, France, Germany, Netherlands, Norway,
Romania, Spain and Sweden [31]. Moreover, recent studies argue that female sol-
diers’ experiences refute their male colleagues’ assumptions, regarding their ability
to serve in combat units [33].
Polish Special Forces comprise five service operation units. The Polish Grupa
Reagowania Operacyjino Mobilnego (GROM) is a special operations force unit that
has already gained an excellent reputation within the world’s special forces commu-
nities, while the GROM training is made as realistic as possible as all the training is
mainly carried out using live ammunition [34]. Their training is based on three main
areas: black tactics that are focused on counterterrorism operations on the ground,
green tactics-special operations in enemy’s rear and activities associated to the releas-
ing of prisoners of war and blue tactics which involves counterterrorism operations
on the sea. Polish Jednostka Wojskowa Komandosów (JWK) is specialized in special
288 J. Reis et al.
reconnaissance. Special reconnaissance reinforces national and allied measures and

systems to gain information that is relevant to carry out operation in extreme circum-
stances. The youngest polish special operation force unit is Jednostka Wojskowa
AGAT (JW AGAT) that was set up in 2012. This unit involves counterterrorism
operation in urban and rural areas. The fourth and last unit is Jednostka Wojskowa
NIL (JW NIL); they operate as a support for other special operations force units
by providing access to information, commanding and logistic support. According
to the constitutional principal of gender equality, both women and men can become
professional soldiers; these rules for entering, fulfilling and terminating military ser-
vice provide equal rights and place the same requirements on both men and women
[35]. Currently, this principle is also applied to the special operations forces, as they
integrate women at SOF units, especially in logistics areas, but they are not directly
involved in tactical operations—this is identified as their main observed obstacle.
4.3 The Inspiring Struggle—The Afghan Case
One of the co-authors of this article has served in the NATO Special Operations
Component Command-Afghanistan (NSOCC-A). During his duties, as advisor of
the Ministry of Interior, he followed the commitment of the North Atlantic Treaty
Organization (NATO) gender advisors while accomplishing their duties (train, advise
and assist) in order to increase women’s representation at all levels of the Afghan
National Defence Security Forces (ANDSF). Women operating in the Afghan special
security forces (ASSF) may integrate the General Command of Police Special Units
under the Ministry of Interior (MoI) or service with the Afghan National Army
Special Operations Command under the Ministry of Defence (MoD). Recruiting
women to ASSF is clearly a higher challenge when compared with most of the
Western countries. Despite this fact, gender discrimination remains an issue within
the ANDSF; our belief is that, being pushed by NATO advisors, the Afghan leaders
are progressively changing their minds and accepting females’ integration into police
and army special forces. The sacrifices made by the Afghan women in the line of
duty might be seen as an inspiration to developed countries that have not yet been
able to successfully integrate women in their SOF.
In this article, it is concluded that although the physical component of the SOF in
Portugal is highly valuable, this paradigm has been changing. Due to some changes,
motivated by the use of military technologies, the technical and intellectual compo-
nent is now appreciated in the Portuguese SOF. Although SOF operators defend a
distinction concerning different physical capabilities between men and women, there
is no scientific evidence concerning the technical and intellectual discrepancies. It is
with the aforementioned arguments that we support the integration of women into
the Portuguese SOF; in addition, existing examples have shown not only the feasi-
bility of their integration, but also new acting opportunities in all types of military
operations.
This article is not free of limitations: due to confidentiality reasons, we have
not provided any information about key informants and the respective military unit
they serve. It is also limited because of its exploratory nature, but we hope that we
can encourage future investigations as it may also be relevant to fill a gap in the
literature. However, case studies are difficult to generalize. Thus, the same study in
other countries may carry the risk of obtaining contradictory findings. This issue
should not be considered a problem itself. To mitigate the generalization aspect, we
suggest to adapt our results to each SOF’s and country realities.
We instigate academics and practitioners to provide new and insightful contribu-
tions to this emergent topic as it represents a fertile opportunity for future research.
Future research may focus on the role and impact of women in the Portuguese para-
troopers. In addition, we also suggest the role and impact of women in the modern
Afghan Armed Forces.
Acknowledgements This project was started under the sponsorship of Erasmus+ programme,
which allowed students from different European countries to cooperate. We would like to thank
the Portuguese and Polish Army, for their outstanding contribution. Lastly, but not least, we would
also like to thank the participation of the anonymous respondents, which greatly contributed to the
production of the final results.
Appendix
Interview protocol—SOF Personnel/Military Women
1. Introduction
• Explain the purpose of the interview

• Explain the reasons for tape recording
• Mention the length of the interview
• Discuss the confidentiality purpose.
2. Generic information
• What is your current professional activity?

• For how long did you work (have you been working) at the Special Operations
Forces? (this question is not applied to women)
• For how long have you been working to the Portuguese Army?
• Age?
290 J. Reis et al.
3. Generic questions
• What do you think about the integration of women in the Special Operations
Forces?
• If you agree to it, in what branches? Why that branch?
• Should women perform tactical/combat operations?
• If not, why?
3.1. Specific questions
– Do you know any international model of integration of women in the
special operations forces?
– If so, can you give some examples?
– Which model do you consider the most appropriate and why?
– As you know, there are no women in the Portuguese special operations.
In your opinion what is the reason?
– What can be done, so that the integration of women in this type of
forces would be successful?
– In your opinion, what is the greatest contribution that women can bring
to the special operations forces?
4. Closing comments
• Is there anything to add which remained unsaid?

• Are you available to revise the transcription of the interview?
• Thank you for your participation in the interview.
References
1. Long, A.: NATO special operations: promise and problem. Orbis 58(4), 540–551 (2014)
2. Kirby, S., Harrell, M., Sloan, J.: Why don’t minorities join special operations forces? Armed
Forces Soc. 26(4), 523–545 (2000)
3. Turnley, J., Michael, K., Ben-Ari, E.: Special Operations Forces in the 21st Century: Perspec-
tives from the Social Sciences. Routledge (2017)
4. Shamir, E., Ben-Ari, E.: The rise of special operations forces: generalized specialization, bound-
ary spanning and military autonomy. J. Strateg. Stud. 41(3), 335–371 (2018)
5. Burrelli, D.: Women in combat: issues for congress. In: Library of Congress Washington DC
Congressional Research Service, pp. 1–17 (2013)
6. Gasca, F., Voneida, R., Goedecke, K.: Unique capabilities of women in special operations
forces. Spec. Oper. J. 1(2), 105–111 (2015)
7. Carreiras, H.: Gender and The Military: Women in the Armed Forces of Western Democracies.
Routledge (2006)
8. Bayeh, E.: The role of empowering women and achieving gender equality to the sustainable
development of Ethiopia. Pac. Sci. Rev. B Humanit. Soc. Sci. 2(1), 37–42 (2016)
9. Carreiras, H.: Gender relations in the armed forces: does the draft make a difference? Sicherh.
Frieden (S+F)/Secur. Peace, 19–24 (2012)
10. Fitriani, Cooper, R., Matthews, R.: Women in ground close combat. RUSI J. 161(1), 14–24
(2016)
11. Boyd, M., Bradshaw, W., Robinson, M.: Mental health issues of women deployed to Iraq and
Afghanistan. Arch. Psychiatr. Nurs. 27(1), 10–22 (2013)
12. Greenburg, J.: New military femininities: humanitarian violence and the gendered work of war
among US servicewomen. Gender Place Cult. 24(8), 1107–1126 (2017)
13. Lane, A.: Special men: the gendered militarization of the Canadian armed forces. Int. J. 72(4),
463–483 (2017)
14. Pinch, F., MacIntyre, A., Browne, P., Okros, A.: Challenge and Change in the Military: Gender
and Diversity Issues. Wing Publishing Office (2004)
15. Szayna, T., Larson, E., O’Mahony, A., Robson, S., Schaefer, A., Matthews, M., Polich, M.,
Ayer, L., Eaton, D., Marcellino, W., Miyashiro, L., Posard, M., Syme, J., Winkelman, Z.,
Wright, C., Zander-Cotugno, M., Welser, W.: Considerations for Integrating Women into Closed
Occupations in US Special Operations Forces. Rand Corporation (2016)
16. Sunde, F., Kristiansen, M.: Defying stereotypes: the untapped potential of integrating female
reconnaissance operators into small state NATO SOF. Small Wars J. (2018)
17. Knarr, W., Turnley, J., Stewart, D., Rubright, R., Quirin, J.: Special operations forces mixed-
gender elite teams (SOFMET): examining socio-cultural dynamics of SOFMET. Joint Special
Operations University, Centre for Special Operations Studies and Research, 1–117 (2014)
18. Kristiansen, M.: Women: a NATO special operations forces force multiplier. Small Wars J.
(2017)
19. Yin, R.: Case Study Research: Design and Methods. Sage, Thousand Oaks (2003)
20. Gummesson, E.: Qualitative Research in Management. Qualitative Methods in Management
Research. Sage, London (1991)
21. Given, L.: The SAGE Encyclopaedia of Qualitative Research Methods, vol. 1–2. Sage, Thou-
sand Oaks (2008)
22. Taylor-Powell, E., Steele, E.: Collecting Evaluation Data: Direct Observations. University of
Wisconsin, Madison (1996)
23. Mills, A., Durepos, G., Wiebe, E.: Encyclopedia of Case Study Research. Sage, California
(2010)
24. Robinson, L.: The future of US special operations forces. Counc. Foreign Relat. 66, 1–39
(2013)
25. Chief of Staff of the Portuguese Army (CSPA): Âmbito de atuação, missões e tarefas das
unidades de tropas especiais. Divisão de Planeamento de Forças, diretiva nº 90 do Chefe de
Estado-Maior do Exército Português (2007)
26. Knie, J.: Special Forces Skills: Are they Special Enough? U.S. Army War College, Carlisle
Barracks PA. (2002)
27. Pordata: Students enrolled in higher education: total and by sex. Statistical database. https://
www.pordata.pt/en/Search/students%20in%20Portugal,%20sex (2018)
28. Torrão, S.: Anjos na Guerra. Oficinas do Livro, 1–168 (2013)
29. Carreiras, H.: Women in the Portuguese armed forces: from visibility to eclipse. Curr. Sociol.
50(5), 687–714 (2002)
30. Grão, L.: Enfermeiras pára-quedistas 1961–2002. Prefácio (2006)
31. Haring, E.: What Women Bring to the Fight. Army War College Carlisle Barracks PA. (2013)
32. Cawkill, P., Rogers, A., Knight, S., Spear, L.: Women in Ground Close Combat Roles: The
Experiences of Other Nations and a Review of the Academic Literature. Defense Science and
Technology Labroratory, Porton Down (2009)
33. Doan, A., Portillo, S.: Not a woman, but a soldier: exploring identity through translocational
positionality. Sex Roles 76(3–4), 236–249 (2017)
34. Ryan, M.: The Operators: Inside the World’s Special Forces. Skyhorse Publishing Inc. (2008)
35. MNDRP, Ministry of National Defence Republic of Poland—Military service. http://
archiwalny.mon.gov.pl/en/strona/333/LG_260_261 (2018). Accessed 31 Oct 2018
The Transformation of the Defense
and Security Sector to the New Logistics
4.0: Public–Private Cooperation
as a Necessary Catalyst Strategy
Manuel A. Fernández-Villacañas Marín
Abstract The logistical needs to develop and maintain both military and police
weapon systems and support material are being met to respond to a set of certain
risks, which now imply the invasion of the territory as well as other threats, such as
international terrorism, cyberattacks, corruption, drug and arms trafficking, global
organized crime, or the trafficking of human beings. The strategic support that will
guide the new operative concepts and together with the acquisition of new military
and police capabilities is based nowadays, and will be increasingly in the future, on
technology and logistics capacity. In this sense, the concepts of the new logistics
have experienced a great catalyze with the digital transformation, which requires a
complete organizational, cultural, and strategic reinvention of both the armed and
police forces and the defense and security industry. The qualitative research con-
ducted has detected a gap in the reactive adaptation of the armed and police forces
to the new logistics 4.0, in relation to its proactive implementation by the main
companies of this sector. The implementation of public–private cooperation models,
which in addition to involving advantages in the improvement of military and police
efficiency as well as business development for the industrial sector, will involve the
permanent and bidirectional transfer of knowledge among the structures implicated,
including the adaptation and permanent dynamic adjustment to the great changes
that the exponential digital evolution will imply.
Keywords Digital transformation · Logistics 4.0 · Defense and security industry

4.0 · Military logistics · Armed and police forces · Public–private cooperation
M. A. Fernández-Villacañas Marín (B)

Spanish Air Force, Madrid, Spain
e-mail: mfermar1@ea.mde.es
Higher Technical School of Aeronautical and Space Engineering, Technical University of Madrid,
Madrid, Spain

https://doi.org/10.1007/978-981-13-9155-2_24
294 M. A. Fernández-Villacañas Marín
1 Introduction
The vast changes that have taken place in this decade have made it necessary and
evident for the armed and police forces of the international community, to understand
the nature and current factors of the global environment that condition their internal
activity. The risks and threats in our days and, therefore, the problems of defense and
security are much broader than those of other times as a result of their global impact,
the characteristics and possibilities of the new digital technologies applied, as well
as of the “intelligent” armament used [6].
Accordingly, the current logistic needs regarding the development and mainte-
nance of weapons systems and their military and police support material are being
covered by realistic and adjusted public budgets to give sufficient response to a set
of certain risks, that within the framework of the New World Order, moderately
implies the usual threat of invasion of the territory, and a significant presence of a
new and diverse set of threats, such as international terrorism, cyberattacks, corrup-
tion, drug trafficking and trafficking of international arms, global organized crime,
or the trafficking of human beings, which are supported by anonymous groups will-
ing to achieve their objectives at the expense of insecurity and international social
destabilization [8].
The strategic support that will guide the development of new operational con-
cepts and the acquisition of new military and police capabilities is based on three
key premises. Firstly, that the present and future strategic environment is uncertain,
complex and conflictive; secondly, that no crisis can be resolved satisfactorily with
isolated military and police employment, which must be combined and integrated
with other initiatives of a civil, political, economic, humanitarian, or informational
nature; and finally, that the military and police forces must have a balanced and an
adequate catalog of capabilities both for the conventional conflicts and for irreg-
ular and hybrid combat, whose source of “military and police advantage” will be
technology and logistics capabilities [10].
All of this has involved, since the end of the previous decade and the beginnings
of the present one, a transformation of the defense and security sector, which has
been largely driven by the development and implementation of the so-called new
logistics in its adaptation to the global world. However, the new technological sce-
nario represented by digital transformation, and in its scope, the implementation of
the concepts of Industry 4.0 and logistics 4.0 in this sector [15], both in its public and
private sphere, requires a complete organizational, cultural, and strategic reinvention
both the armed and police forces and the international industrial sector of defense
and security, thus providing the international community and each of the countries
with a dynamic redesign capability that can respond to the needs arising from the
processes of exponential technological evolution, that this new paradigm carries with
it.
The Transformation of the Defense and Security Sector … 295
2 Background and Literature Review
2.1 The New Logistics: Creating Competitiveness and Global

Sustainable Development
In recent years, logistics has been acquiring a growing importance in the strategy
of companies and has become an essential factor for improving their competitive-
ness in a constantly changing global market. Thus, the application of improvements
in the performance of methodologies and technologies in the logistics field, which
have developed exponentially with the Internet, implicitly leads to obtaining com-
petitive advantages. These advantages are based not only on increasing efficiency
and effectiveness in logistics management, but also on increasing the value provided
to customers and global sustainability.
The new logistics is based on a conceptual transition from a “push” approach to a
“pull” approach, developing new logistic models that are more efficient, sustainable,
intelligent, agile, adaptable, scalable, and resilient. The following figure summarizes
the main aspects of traditional logistics and its evolution to the new logistics [9]
(Fig. 1).
On the other hand, the sustainability of global logistics [3–5] and of supply-chain
management represents a strategy of transparent integration of the social, environ-
Fig. 1 Traditional logistics versus the new logistics

Fig. 2 La prospectiva de evolución hacia el Physical Internet en 2050. Source http://www.etp-

logistics.eu/wpcontent/uploads/etpalice/Road_map_Alice_PIE.jpg, 17.12.2018
mental, and economic objectives of the organization, within the system of coordi-
nation of the main inter-organizational business processes, in order to improve the
long-term economic results of each company and its supply chains [1].
The development prospective of the new logistics with a long-term projection has
shaped the theory of Physical Internet [14]; a vision that has been adopted both in
the USA and in the European Union as the conceptual goal of logistics for 2050. It
involves achieving an open global logistic system, based on a physical, digital and
operational interconnectivity, through encapsulation, interfaces and protocol design,
in order to move, store, perform, provide, and use physical objects throughout the
world in an economically, environmentally, and in a socially efficient and sustainable
way. The aim is to eliminate inefficiencies in global transport and waste management
of logistics networks, in a similar way to what the Internet raised worldwide for
information flows [13] (Fig. 2).
It will be necessary to create an open market for the transport of goods, with shared,
open and adaptable distribution chains, in which the products will be transported in
modular, standardized, and intelligent containers, allowing each unit to be followed
and controlled.
For this, it will be necessary to achieve a full level of global collaboration, redefin-
ing the competitive space, taking it out of the logistics processes, and taking it only
to the points of sale in which customers define the market share of each product.
The viability of the Physical Internet will obviously require the implementation of
existing digital-enabling technologies, such as the Internet of things, cyberphysical
systems, collaborative robotics, e-commerce, blockchain, virtual reality, augmented
reality, additive manufacturing and 3D printing, artificial intelligence, big data, data
analytics, predictive maintenance, simulation, collaborative economics, cloud com-
puting, cybersecurity, and machine learning, but above all, the incorporation of other
more advanced and complex innovations resulting from the exponential technologi-
cal evolution that are being developed now and in the future.
2.2 Logistics 4.0: The Transformation of the New Logistics

to the Digital Domain
Digital transformation can conceptually be defined as the process of organizational,

cultural, and strategic reinvention, both of companies and of public entities, necessary
for the integral application of the technology that we call digital, which generates,
processes, stores, and uses data, information and intelligence, to improve its perfor-
mance as well as its ability to quickly adapt to disruptive or radical changes generated
in the environment. This technological scenario, an authentic new paradigm, is induc-
ing the emergence of shorter and shorter management cycles, in which the environ-
ment changes continuously and increasingly faster, with a pace that we assume will
continue to accelerate in the future, with an approach in exponential technological
evolution as has already been highlighted.
The main catalyst for change and the cause of this continuous acceleration is
the digital revolution, motivated by the expansion of the Internet, information and
communication technologies, as well as the universalization of its use, which is in turn
involving the transformation of our lifestyle. In this way, we are today fully imbricated
in a process of deployment of disruptive technologies that are profoundly modifying
the business reality and the society in general. Although computer science signified
a cyclopean advance in the mechanization and automation of the processes, and the
later connection of the equipment between them, it also generated the beginning of
a formidable process and distribution capacity of the information, the digital factor
has exponentially multiplied the connectivity of all public and private actors, among
which, logically, citizens are included as protagonists.
The inclusion of digital technologies and the transformation induced by them
defines what has come to be called the Fourth Industrial Revolution. Effectively
globalization, the universalization of the use of the Internet, the full automation of
processes and the digitization of information, have led to an authentic Industrial
Revolution, in which the hybridization of the physical and virtual world is taking
place; products, machines, tools, factories, warehouses, and vehicles are intercon-
necting each other and work automatically, an interconnection of all the elements
of the value chain that is becoming intelligent and that is leading to the creation of
authentic centered networks in the clients. In this way, the environment appears to
be changing, hybrid, competitive, global, connected, and interdependent.
The result of all this will allow us to achieve immediate answers in decision-
making based on the information captured in real time, through intelligent systems
and processes, without variability or errors, with full traceability in the process chains
and total sustainability. All of this leads us to a new situation that will affect the way
of producing and controlling processes, applying logistic models, and developing
marketing and commercialization strategies.
In this way and totally linked to the concept of Industry 4.0, the concept of logis-
tics 4.0, which develops and assumes the concepts of the new logistics, transcends
and emerges due to its importance. That is that logistics, as a key element of the
industrial sector, cannot be kept aside. The Smart Factory implicitly assumes that
manufacturing plants, facing a traditional centralization derived from the search for
economies of scale by volume, now consider an intelligent relocation by factors of
specialization and synergies, giving rise to the creation of different and immense
networks distant from interconnected production units, which will imply that the
raw materials, components, and semi-finished products require greater mobility, be
perfectly synchronized and guaranteed, participating in global supply networks and
intra-industry trade. On the other hand, the life cycles of products in international
markets tend to be continuously reduced and variations in their demands are increas-
ingly difficult to estimate, so it is crucial to have greater connectivity and integration
between the final and initial links of the global supply chains, as well as a greater
capacity for the aggregation of data and automatic estimation, rigorous and real-time
estimation of the different demands, with expert systems and artificial intelligence
that assist the decision-making processes [9]. For all this, the new logistics 4.0 will
involve the optimization and full connection of all elements and processes of the sup-
ply chain, which should generate improvements in the efficiency and effectiveness
of the management of orders and shipments, a production oriented to the individ-
ual customer which increasingly demands more personalized offers, the geolocation
of customers, the simultaneous multimodal omnichannel, and the optimization of
global routes, a full capacity for adaptation, the total international traceability of
merchandise, the reduction of stock and space necessary for storage, automation
of payments, etc. Smart logistics will not only better meet new demands, but will
also more accurately locate and thoroughly know current and potential customers
through the processing and interpretation of the data collected and the optimization
of processes. Integration, holistic vision, coherence, innovation, and flexibility are
the key concepts to support the development of the new logistics 4.0.
Finally, it is considered necessary to emphasize that the digital transformation
and the emergence of Industry 4.0 and logistics 4.0, beyond the technological devel-
opments and their deployment, and the reinvention of strategy, organization and
corporate culture, requires, on the one hand, the development of a new holistic style
of a humanitarian leadership, but also, on the other, the availability of a permanent
digital leadership capacity that achieves the most critical element for the success of
the analyzed evolution to be collaborate and be fully involved: This is the human
factor.
3 The Problem: The Imbalances and Delays Between

the Adaptation to the New Logistics 4.0 of the Armed
and Police Forces, and the Companies in the Defense
and Security Industry
In order to know the state of involvement and development regarding the digital
transformation in the companies of the defense and security industry in Spain, a
qualitative research has been carried out to the main ones that accounted for approx-
imately 85% of sales in 2016 of this sector (according to the Industrial of Defense
Report of the Spanish MoD), all of them with European implantation and global pro-
jection (Airbus Defense and Space, Airbus Military, Indra Systems, Navantia, ITP
Turboprop Industry, Expal Systems, Airbus Helicopters Spain, Cepsa, Iveco Spain,
Santa Bárbara Sistemas, Naval Constructions P. Freire, Spanish Company of Aero-
nautical Systems, ARPA, etc.). The study has been complemented with the vision
of the two associations that bring together most of the companies of the sector in
Spain: the Spanish Association of Technology Companies of Defense, Aeronautics
and Space (TEDAE), and the Association of Companies that contracting with the
Public Sector and especially with the Armed and Security Forces (AESMIDE).
The results show that almost all companies consider that digital disruption implies
opportunities for improvement in the efficiency, quality, and flexibility of processes,
which will produce the potential increase in operating profits. The defense and secu-
rity industry is facing a new global context that poses major strategic and structural
challenges, against which practically all have proactively raised lines of action.
However, the analysis as to the levels of initiative and implementation of these
lines of action, in regard to the armed forces and police forces which are customers
of these companies, both nationally and internationally, denotes a reactive attitude.
They are aware of the implications that the digital transformation will have on their
operation levels and sustainability of their systems, but this has not translated into
decided and effective action plans, possibly due to the lack of funding to address
them.
In short, the probable existence of a gap in the reactive adaptation in the armed
and police forces to the new logistics 4.0 is detected in relation to its proactive
implementation by the main companies of the defense and security industry, with a
delay and desynchronization that predictably will tend to grow due to the exponential
technological evolution.
4 The Solution: Public–Private Cooperation Models

as a Necessary Catalyst Strategy. From Operational
Outsourcing to Cooperation
Outsourcing, as a strategic process of the transfer of public activities of the armed

and police forces to the private sector, has been an aspect in the defense and security
sector in continuous development since the beginning of the 1990s in almost all
the countries. Since then, their governments have opted for the implementation of
outsourcing solutions in response to the growing demand for the in-service support
of their weapons systems and progressive scarcity of resources, exploring better
solutions in both technical and economic terms. Factors such as the acceleration
of the technological evolution of armament and support material, which has made
it increasingly difficult for the armed forces and police to maintain the necessary
autonomy to ensure proper maintenance of their systems, or the increase in the
cost of these equipment and the relative shortage of resources, especially qualified
human resources, led to raise in the international arena to operational outsourcing as
an effective solution to the new situation generated [12].
The management technique of outsourcing arises in the post-industrial era when
competition in global markets begins [11]. Through the externalization of those
activities that are not critical, the entity specializes in its core activity, improving
its competitiveness and the ability to adapt to rapid changes in the environment.
But more than generic outsourcing (outsourcing), as a transfer of certain activities,
managed in a localized manner in the short term and with specific objectives, the
concept of cooperation appears as a public–private co-management (co-sourcing).
It is a strategy of global restructuring to obtain results that allows the best efforts of
the organization to be oriented toward its nuclear activities and, with it, toward the
fulfillment of its mission and objectives. The cooperative vision is to consider the
provider of services as a partner with which a “win to win” relationship is established.
To the extent that the organization internalizes that its fate is associated with that of
its service provider and vice versa, the need to seek relationships in which there is
mutual benefit becomes clearer [2]. Outsourcing, which in principle was proposed
as a managerial technique, has potentially moved toward cooperation, partnerships,
and strategic alliances. It is not normally a question of formal companies, because
the capital and patrimonies are independent and the symbiosis is such that it is
not necessary to have a mercantile corporate relationship to be considered business
partners.
In recent years, in the public sector, the public–private cooperation has become
an international reference model that allows effective, efficient, and synergistic inte-
gration of entrepreneurship with the management capacity of public services of
the administration. Through its mediation, certain collaboration formulas are estab-
lished through which the public sector interacts with the private sector in the search
for greater involvement of the latter in the development of certain complex projects,
setting out a scheme of joint objectives, as well as sharing risks, costs, and benefits.
It is a new conception of public management through which social needs are met in
a different way, through a new frontier of relationship between the public sector and
the private sector. The trend toward its use has been generalized in recent years as a
management model for large projects, particularly in the innovation and high tech-
nology sectors, in internationalization initiatives and, in general, in all those in which
it is necessary to guarantee high levels of competitiveness in advance. It is concep-
tualized as an economic-organizational model of dynamic and plural relationship,
which groups a wide range of figures that maintain a link in time between two or more
public and private organizations. It signifies a new conception by means of which
a transition of a relationship between administration and industry of competitive
character which is generated to another one of collaborative character.
On the other hand, the application of public–private cooperation in the defense and
security sector entails a change of philosophy and mentality in the bodies involved
in the public administration that will go on to manage certain essential services
through a strategic partnership with private companies, assuming these shared risks
and responsibilities with the armed forces and police forces. The advantages for both
parties of the application of these models of public–private cooperation are clear.
It can mean higher levels of rationalization and efficiency for the armed forces and
police forces, as well as important savings via, among other factors, the reduction
of military and police personnel and the technical assistance involved, as well as
the reduction of costs of investment in facilities and equipment (avoiding duplicated
expenses, heavy investments in spare parts stocks that must be maintained during
long life cycles of the systems, reducing the costs of continuous training, technical
documentation and its corresponding update, etc.). On the other, the defense and
security industry can obtain a new field of business development induced by the
greater investment capacity derived from public savings, addressing the development
of new defense and security systems that will make the strengthening of the industrial
base and technology of that sector feasible.
Moreover, and perhaps most significantly, the functioning of the public–private co-
management models implies the permanent and bidirectional transfer of knowledge
between the public and private sectors, as well as the adaptation and the dynamic
and permanent adjustment to major changes, that the exponential digital evolution
will imply that “unique entity” formed by the agencies involved in the armed and
police forces with the companies participating in the defense and security industry,
in a contemporary way.
As a development of these models, the Spanish Air Force carried out the FENIX
Project at the end of the previous decade. It was a series of applied research stud-
ies to improve the operational efficiency and profitability of surplus capacities of
its Aerospace Maintenance Centers MRO (Maestranzas Aéreas) through the imple-
mentation of public–private cooperation models with the aerospace industry. The aim
was to achieve improvements in the level of generation of air support services, with
the lowest possible cost and risk. Previously, an international benchmarking study
of the solutions applied by the air forces of the USA, the UK, France, Germany, and
Italy was carried out [6, 7].
5 Conclusions
The work carried out in this paper has first of all analyzed the logistical needs of
development and maintenance regarding weapon systems and support material for
both the military and police. Nowadays, they are based on technology and logis-
tics capacity. In this sense, the new logistics have experienced a great catalyze with
the digital transformation, which requires a complete organizational, cultural, and
strategic reinvention of both the armed and police forces and the defense and secu-
rity industry. Secondly, the qualitative research conducted has detected a gap in the
reactive adaptation of the armed and police forces to the new logistics 4.0, in relation
to its proactive implementation by the main companies of this sector. In order to
reduce this gap, the implementation of public–private cooperation models, in addi-
tion to involving advantages in the improvement of military and police efficiency,
and business development for the industrial sector, will involve the permanent and
bidirectional transfer of knowledge among the structures involved, as well as the
adaptation and permanent dynamic adjustment to the vast changes that the exponen-
tial digital evolution will imply.
References
1. Carter, C.R., Rogers, D.S.: A framework of sustainable chain management: moving toward
new theory. Int. J. Phys. Distrib. Logist. Manag. 38(5) (Emerald Group Publishing Limited)
(2008)
2. Fernández-Villacañas, M.A.: Externalización del Sostenimiento, Revista de Aeronáutica y
Astronáutica. Spanish Air Force, Madrid (2007)
3. Fernández-Villacañas, M.A.: Externalización estratégica Vs. Externalización operativa: Desar-
rollos potenciales del modelo de Partenariado Público-Privado (PPP) para el sostenimiento de
Armamento y Material de las Fuerzas Armadas. Universidad Comercial de Deusto, Bilbao,
Spain, 28th and 29th Feb 2008
4. Fernández-Villacañas, M.A.: El outsourcing estratégico como modelo de integración de las
cadenas de suministro para una logística sustentable. V Fórum Mundial de Logística, Monter-
rey, México, 6th and 7th Mar 2008
5. Fernández-Villacañas, M.A.: Reflexiones en torno al logro de una Logística Global Sostenible.
V Congreso Internacional ACOLOG - Asociación Colombiana de Logística, Bogotá, Colom-
bia, 28–30 Aug 2008
6. Fernández-Villacañas, M.A.: Modelos de cooperación industrial para el sostenimiento de arma-
mento y material: El Partenariado público-privado y la rentabilización de capacidades exce-
dentes. In: Different authors (eds.) Gobierno en la Industria de Defensa, Aula Abierta y Foro
de Estudios de Seguridad y Defensa, Universidad Politécnica de Cartagena, Cartagena, Spain
(2009)
7. Fernández-Villacañas, M.A.: Cooperación industrial público-privada para el sostenimiento
profundo de las aeronaves del Ejército del Aire y la rentabilización de capacidades excedentes.
5º Congreso Español de Mantenimiento y 16º Congreso Iberoamericano de Mantenimiento,
Asociación Española de Mantenimiento, Barcelona, Spain, 15–17 Nov 2011
8. Fernández-Villacañas, M.A.: La Nueva Economía de la Defensa en un Nuevo Orden Mundial.
Revista de Aeronáutica y Astronáutica. Spanish Air Force, Madrid, Spain (2017)
9. Fernández-Villacañas, M.A.: Las plataformas logística 4.0 y la mejora del comercio global:
Creando ventaja competitiva logística y desarrollo sostenible. VI Simposio Internacional Online
de Logística y Competitividad, High Logistics Simposios, Medellín, Colombia, 3rd and 4th
Oct 2018
10. García Arnaiz, J: Discurso de Inauguración del Jefe de Estado Mayor de las Jornadas Aeroes-
paciales de Economía de la Defensa del Ejército del Aire “La Nueva Economía de la Defensa
en un Nuevo Orden Mundial”. ACGEA y UNED, Madrid, Spain, 7–9 Mar 2017
11. Gidrón, G.: Nuevos modelos de gestión empresarial: El outsourcing de procesos de negocios.
Círculo de Empresarios, Madrid, Spain (1998)
12. Más Sabaté, J.: Gestión privada de servicios públicos. La externalización (outsourcing) en la
Administración Pública. V Congreso Internacional del CLAD sobre la reforma del Estado y
de la Administración Pública, Dominican Republic, 24–27 Oct 2000
13. Montreuil, B., Meller, R.D., Ballot, E.: Towards a physical internet: the impact on logistics
facilities and material handling systems design and innovation. In: Gue, K. (ed.) Progress in
Material Handling Research. Material Handling Industry of America, USA (2010)
14. Montreuil, B.: Toward a physical internet: meeting the global logistics sustainability grand
challenge. Logist. Res. 3(2–3), 71–87 (2011)
15. Romero Garat, F.P.: Adaptación de la Defensa y las Fuerzas Armadas al concepto industria
4.0 (1 y 2). Infodefensa.com homepage, http://www.infodefensa.com/es/2018/01/29/opinion-
adaptacion-defensa-fuerzas-armadas-concepto-industria.php, http://www.infodefensa.com/
es/2018/02/01/opinion-adaptacion-defensa-fuerzas-armadas-concepto-industria.php. Last
accessed 27 Dec 2018
Part IX
Safety and Maritime Protection
An Autonomous Airship Swarm
for Maritime Patrol
Constantino G. Ribeiro, Luciano Santos Constantin Raptopoulos
and Max Suell Dutra
Abstract The studies of swarm of unmanned aerial vehicles (UAVs) have been
increasing, and swarm of UAVs can become part of many daily tasks. But, as mat-
ter of fact, even the use of a UAV does not mean the decreasing of operational
complexities and, consequently, the costs of performing its tasks, because of high
costs of trained operators and remote control facilities to operate state-of-the-art
UAVs. So, in order to support the operation of swarm, this work proposes a paral-
lel/distributed framework to control mission of each UAV. These unmanned crafts
will less dependent on remote control facilities. Embedding the mission control run-
ning in an embedded parallel/distributed computer system will be able to carry on
basic mission control tasks. In order to prove this concept, the following items were
developed: (i) a prototype of an embedded parallel/distributed computer cluster using
low-cost components; (ii) new procedures to resolve navigation and collision evasion
issues; and (iii) a parallel/distributed path discover program. The tests carried out in
the embedded parallel/distributed computer cluster prototype and with new evasion
procedures proved the viability of proposed framework.
Keywords Autonomous airship · Swarm of robots · Embedded computer cluster ·

Parallel/distributed programs
C. G. Ribeiro (B)
CEFET-RJ, UnED Itaguaí, Rodovia Mário Covas, lote J2, quadra J, CEP: 23810-000, Distrito
Industrial de Itaguaí, Rio de Janeiro, Brazil
e-mail: constantino.ribeiro@cefet-rj.br
L. S. C. Raptopoulos
UnED Nova Iguaçu, Estrada de Adrianópolis, 1317, CEP: 26041-271, Nova Iguaçu, Rio de
Janeiro, Brazil
e-mail: luciano.raptopoulos@cefet-rj.br
M. S. Dutra
COPPE - UFRJ, Centro de Tecnologia, bloco G, salas 202/203/204, Universidade Federal do Rio
de Janeiro, Cidade Universitária - Ilha do Fundão, Caixa Postal 68.503, CEP: 21945-970, Rio de
Janeiro, Brazil
e-mail: max@mecanica.coppe.ufrj.br

https://doi.org/10.1007/978-981-13-9155-2_25
308 C. G. Ribeiro et al.
1 Introduction
There is a set of long-range aircrafts tasks, many of them performed over in sensitive
geographic places, that demand specialized equipment and well-trained crew. The
examples of these tasks are surveillance [1], patrol [2], data collection, search, and
rescue. These tasks are being substituted by drones, which eliminate crew fatigue and
improve performance. Although there have been many improvements in electronics,
robotics, and computing, the costs of using the so-called drones, especially in defense
and surveillance, have increased a lot as stated in Table 1.
The cost of using drones is high because they need a staff working in remote control
station. So, motivated by a scenario of new applications for autonomous unmanned
aerial vehicles (UAV), especially using airships, this work proposes a swarm of
UAVs (airships) to perform maritime surveillance/patrol. In this propose, a new UAV
mission control framework was developed. This framework is composed of a low-cost
embedded computer cluster; a new autonomous collision evasion system; a new on fly
parallel/distributed mission route generating system; and a set of instruments/sensors
for UAV orientation (GPS, electronic compass, accelerometers, among others).
2 Related Work
Vehicles path planning is one of the most studied subjects. It is considered a com-
putational complex problem or with exponential time for their resolution. In this
section, the most relevant and related works to this subject will be briefly described.
Reference [3] proposes an infinity or achievable horizon optimization method
to unknown environments. The work is based on a finite optimization time and is
dependent on computational performance of the used equipment. It uses a finite state
model of the movement of a mini robotic helicopter vehicle. The proposed work uses
cost function called tail discharge or path horizon.
Table 1 Operational costs of surveillance crafts

Platforms Costs ($ per hour)
71 m balloon land based (http://www.navsea.navy.mil) 610
MQ-1 Predator (robotic aircraft)—low operational ceiling; long range 5.000
(http://www.af.mil/information/factsheets/factsheet.asp?fsID=122)
Grumman E-2C “Hawkeye”—AWACS (airborne warning and control 18.000
system)—seagoing aircraft (http://www.navy.mil/view_image.asp?id=
140671)
RQ-4 Global Hawk (robotic aircraft)—high operational ceiling; long 31.000
range (https://www.americansecurityproject.org/the-us-and-its-uavs-a-
cost-benefit-analysis/#_edn4)
An Autonomous Airship Swarm for Maritime Patrol 309
The task of controlling a robotic blimp in a strong wind environment is studied

in [4]. The work proposes a stability control of the aircraft using following path
technic and flying against the wind. The stability control uses a Lyapunov function
that follows pre-established rules. An adapted remote control blimp of 12 m of length
was used in satisfactory tests.
A path optimization that uses finite retreaded horizon is proposed in [5]. The
work uses a Lyapunov as a cost-to-go function to construct a viability path graph
to be optimized. After that, the optimization problem is solved as sequential finite
state control and sequential decision problem and is used as global planner fed by a
finite retreaded horizon method. The work compares the performance and computa-
tion load of other works and the proposed model. The tests were conducted in real
environment using a Blade CX robotic helicopter.
A robust and simple navigation system using fuzzy logic is the subject of [6] that
uses a Plantaraco robotic blimp. The work uses ultrasonic sensors to get data about
the surrounding environment. These data feed a collision evasion fuzzy system and
that generates quick turns of 180°. The test results show a well-balanced behavior of
the fuzzy controller that can be even trained.
The control and surveillance of natural disaster areas by a robotic blimp are
proposed in [7]. The blimp uses a path control system with speed field method
and an optimal inverted path controller as control strategy. The used strategy was
efficient in strong wind environment to avoid complex time corrections in the path.
The optimal inverted path controller uses y and x coordinates, a nonlinear looping
based in a Hamilton–Jacobi–Bellman equation, and a Lyapunov filter to horizontal
vehicle control. The practical tests used a 12.2 m length blimp and payload of 15 kg,
with wind sensors and stereo cameras.
Reference [8] proposes a path control using a reinforced learning method. The
method keeps the blimp high using previous environment conditions and agents
(autonomous computational programs). The agent programs get bonus when they
execute right actions that maximize a Monte Carlo search method used to search
solutions. This method leads a direct learning, with no need of previous data storage
by agent programs. A Gaussian function is used to avoid problems with search space
size due to learning search of state-action tuples. Ultrasonic sensors capture the state
components of these tuples during blimp navigation, and a Kalman filter removes
any noise. The tests used an 18 m length blimp and were conducted indoors in a
5-m-high old factory shed.
The overall robotic blimp dynamic is described in a mathematical model by [9].
This model is used to develop a predictive and adaptive robust nonlinear control that
is used to navigate and guide a robotic blimp over environment disturbs. The blimp
data (like longitudinal flight, climb rate, turns, etc.) were collected, and the blimp
dynamic mathematical model was generated by MATLAB suite.
Reference [10] AURORA project proposes an image capture system that models
a set of elementary signals. This set establishes a relationship between the blimp
speed and the land targets. It uses a similar model of the pendulum problem but in
association with the movement to a zero distance with the blimp. The work takes into
account the wind effects in three possible situations: without wind and no environ-
mental disturbance; with wind and weak environmental disturbance; and with wind
and strong environmental disturbance.
3 The Proposed Swarm of UAVs
Airships were extensively used during World War II by the US Navy in coastal
patrol operations and escort of sea trains [11]. These operations were carried out
satisfactorily for escort of sea trains by Goodyear type K airship. Goodyear built 170
type K airships which served on four continents, escorting 89,000 ships. None of the
escorted ships were lost in combat [11] (Fig. 1).
Today, there is a set of applications that demand long periods of use, stealth ability,
low cost, and reliability of the platform to be used. A good example of these tasks is
anti-submarine warfare (Fig. 2).
Fig. 1 World War II US

Navy airship (http://www.
pbs.org/wgbh/roadshow/fts/
raleigh_200902A15_ss.
html#1)
Fig. 2 Airships used in

anti-submarine warfare
(http://www.xconomy.com/
seattle/2008/07/11/boeing-
and-skyhooks-zeppelin-
copter-faces-safety-
challenges/attachment/
skyhook_arctic/)
A swarm of robots is a group of robots that work to perform a collective behavior

to carry out a task interacting with each other and with the environment [12]. So,
this work proposes a parallel/distributed framework to control mission of the whole
swarm of robotic airships performing maritime surveillance and search in sensitive
and strategic seashore areas of Brazilian coast as offshore petroleum platforms. This is
a low cost, less dependable of human interaction, long range, and endurance solution
for the task. Next sub-sessions just explain where movement and force equations
came from and how they were used.
3.1 UAV Kinematics
The degree of freedom (DOF) is a set of independent movements or rotations that

can define the position or orientation of a mechanical system. They can be defined
by Euler angles and position coordinates (x, y, and z) [13]. The Euler angles and
position coordinates can give a total of six DOFs to a rigid body as stated in Table 2.
The coordinate vector η totally describes the airship orientation. It uses Euler
angles vector θ plus the three-position coordinated vector p shown in (1), (2), and
(3) equations. An airship is a sub-actuated vehicle. So, it demands less control data
than degrees of freedom (DOFs) [14] and stands for all environmental situations.
The minimal representation to describe the control data to aerial vehicle is the roll
(), pitch (θ ), and yaw () plus the position vector coordinate p, and they describe
the airship behavior. Equations (1)–(9) represent the position, attitude, forces, and
moments actuating in the airship.
η = [ p, θ ]T (1)
Table 2 Euler angles

Degree of freedom Forces and moments Linear and angular Euler angles and
(DOF) speeds position
x-axis movement X u x
(surge)
y-axis movement Y v y
(sway)
z-axis movement Z w z
(heave)
x-axis rotation (roll) K p ϕ
y-axis movement M q θ
(pitch)
z-axis movement N r
(yaw)
p = [x, y, z]T (2)
θ = [ϕ, θ, ]T (3)
v = [V, ]T (4)
V = [u, v, w]T (5)
= [ p, q, r ]T (6)
τ = [ f, m]T (7)
f = [X, Y, Z ]T (8)
m = [K , M, N ]T (9)
The airship position is described in relation to inertial coordinates and its speed by
a body fixed referential. The η vector (1) has the referential coordinates in a specific
referential and v vector (4) the speed related to inertial referential. So, the relation
among variables of each referential where one referential is passed to each other [10]
can be described by:
ṗ = R(θ )V (10)
where R(θ ) is the transformation matrix with rotations and translations to right refer-
ential. And after the appropriated transformations, you can use the Jacobean matrix
notation (11) and (12) to describe the UAV kinematics.

ṗ R(θ ) 03×3 V
η̇ = J (n)v (11) ↔ = (12)
θ̇ 03×3 T (θ )
3.2 UAV Dynamic
The blimp dynamics, described in [15, 16, 17, 14], takes into account many data
about the effects of aerodynamics, structural issues, actuators, and propulsions. Using
Newton and Lagrange laws (forces and momentum applied to rigid bodies) [18]:
M v̇ + C(v)v + D(v) + g(η) = τ (13)

⎧
⎨ η = [x, y, z, ϕ, θ, ]T
v = [u, v, w, p, q, r ]T (14)
⎩
τ = [X, Y, Z , K , M, N ]T
where
M = M RB + M A rigid body system inertia matrix (with added masses).
C(v) = C RB + C A (v) coriolis force matrix and centripetal force matrix (with added
masses).
D(v) aerodynamic dumping matrix.
g(η) gravitational forces and momentum and static sustentation
vector.
τ control data vector.
The complete development of proposed airship (with dimensions, mass, and other
parameters) can be found in [15], as well the specific control system using computed
torque.
4 The New Framework
The framework proposed is composed of an embedded parallel/distributed com-

puter, a new collision evasion system, a parallel/distributed mission route generator
program, and a set of electronic sensors.
4.1 The Embedded Computer Cluster
The Beowulf-based computer cluster [19, 16] is a reliable and easy way to construct
a high-performance computer facility. It can provide a lot of benefits as, for example,
free parallel/distributed programing environment and high-performance programing
tools. This work proposes an embedded computer cluster made from low cost and
market (high-performance multicore board used in last generation cellular phones).
These boards are connected by a switch, and cluster main node will be connected
with Arduino [17] or similar interface to control airship movements and to radio
link connect the cluster to the mission supervision center (see basic computer cluster
layout in Fig. 3).
The speed of route calculations on fly is crucial to make the framework reliable
and effective. A brand-new platform is used as cluster nodes: four ODROID-x2 open
development platform, based on Exynos 4412 Prime 1.7 GHz ARM Cortex-A9 Quad
Core with 2 GB memory and two and 2 e ODROID-C1 [14, 20, 21, 22]. The cluster
will supply the airship with enough computer power to generate its routes on fly as
it needs, and it will run a new, vivid, and specific obstacle collision detection and
avoidance system. The main path generator task must have performed by a graph
path discover program. The Traveler Salesman (TS) algorithm was chosen to perform
it. A version of TS program was modified to be executed in a parallel/distributed
programing environment; using message passing interface (MPI) library, it can be
executed in the proposed embedded computer cluster. The parallel/distributed TS
Fig. 3 Embedded computer cluster of proposed framework
will be performed any time it is needed, because of any disturbance in previously

calculated route due to any obstacle avoidance performed procedure and every time
an estimated target is reached or any other significant disturbance that affected the
airship attitude or flight.
4.2 The New Path Control System
The path mission control system (PMCS) uses a set of pre-selected coordinates
to be reached by the proposed autonomous managed blimp. The set is used by
the parallel/distributed TS module of the PMCS to establish an optimal route to
complete the mission. During the execution of planned route, the laser detection and
ranging (LIDAR) [23, 24] and normal radar are used to scan possible unpredicted
and unexpected obstacles. As soon as an obstacle is detected, it is analyzed and the
proper avoidance procedures are taken. Figure 4 shows the flowchart of path mission
control system.
Fig. 4 Flowchart of path mission control system
4.3 New Obstacle Avoidance Systems
As a part of PMCS, a new collision avoidance system is proposed. As described,

the framework is equipped with a LIDAR and radar to detect obstacles. The radar
will make a long-range search and detection of obstacles, while the LIDAR will sort
range detection and collect precise data of obstacle like distance, speed, and attitude.
The system identifies two classes of obstacles: fixed and mobile ones.
Fig. 5 Flowchart of new

procedure to detect mobile
obstacles
The evasion of fixed obstacles uses fallow the wall procedure [25] using LIDAR
data to decide which side to turn and contour the fixed obstacle as a wall. A new
and innovative procedure to detect mobile obstacles is proposed using data from
LIDAR and COLREGS 72 (Convention on the International Regulations for Pre-
venting Collisions at Sea) [26] navigation rules for cross routes to avoid possible
collision situations. The reason to use navigation rules is because the airship has
similar behavior (dynamic forces) of a ship, and as COLREGS 72 rules are widely
used and tested in control ship and their movements and routes, they must work well
with the path control of blimps. Figure 5 shows the flowchart of new procedure to
detect mobile obstacles.
4.4 UAV Sensors and Navigation Framework Layout
A basic set of sensors is necessary to support airship autonomous operation. The basic
sensors are electronic compass, speed sensor, wind speed sensor, accelerometer,
altimeter, frontal radar, LIDAR with 10 km of range and 90° of aperture, GPS,
Fig. 6 Prototype of
embedded computer cluster
electronic gyroscopic, six ultrasonic close-range sensors, and four digital cameras.
These should be basic and always present sensors that will supply data for airship
navigation and attitude procedures. Depending on the kind of mission, some extra and
specific equipment will be necessary such as night vision high-resolution cameras,
electronic surveillance devices for law and enforcement and patrolling tasks.
5 Tests and Results
The embedded computer cluster is the core of this propose, so a proof of concept
was carried on. A prototype of an embedded computer cluster with six nodes (see
Fig. 6) was constructed, and performance and fault tolerance were conducted using
prototype parallel/distributed programs.
5.1 Results of Speed Up Execution Tests
The basic speed up test was to execute a sequential version of TS program (devel-
oped in C language) for: 8, 16, 32, 64, and 128 targets and then to execute a paral-
lel/distributed version of TS program (developed in C language with message passag-
ing interface—MPI—library) for the same set of targets and compare the execution
time (see Fig. 7).
5.2 Resilience and Fault Tolerance Tests
In these tests, all prototype programs for mission control were executed. During the
execution of all mission control programs, the processor parameters (CPU, memory,
Fig. 7 Speed up test times
Fig. 8 Single processor parameters usage test
Fig. 9 Embedded computer cluster parameters usage test
Fig. 10 Embedded computer cluster fault tolerance test
network usage) of a single processor environment and the same parameters for the
embedded computer cluster were compared (see Figs. 8 and 9).
Another test was a fault tolerance test where the parameters (CPU, memory,
network usage) were compared for the complete cluster (with six nodes—left side of
the figure) and a damaged cluster with less one node (with three nodes—right side
of the figure) (see Fig. 10).
6 Conclusions
A prototype of an embedded computer cluster with market computer elements used

in mobile phones was constructed and operated as any other computer cluster with
the advantage of being small and with very low energy consumption. The basic tests
of speed up for embedded distributed/parallel application showed the improvement
in the performance of these applications compared with the embedded sequential
ones, especially with big set of targets. The usage tests proved that the prototype-
embedded computer cluster has smaller consumption of computer resources in each
node of the cluster, so other missions’ specific tasks can run into the cluster. The fault
tolerance test showed that the embedded computer cluster can operate properly even
when there is a damage in one of some of its nodes. All the tasks described above
proved the viability of proposed framework.
References
1. Defense Industry Daily staff: JLENS: Coordinating Cruise Missile Defense – And More,
30 December 2012. http://www.defenseindustrydaily.com/jlens-coordinating-cruise-missile-
defense-and-more-02921/ (2012)
2. Corporation, A.P.: Aerial Products, 28 December 2012. http://www.aerialproducts.com/
surveillance-systems/aerial-surveillance.html (2008)
3. Dadkhah, N., Korukanti, V.R., Kong, Z., Mettler, B.: Experimental demonstration of an online
trajectory optimization scheme using approximate spatial value functions. In: Proceedings of
the 48th IEEE Conference on Decision and Control, pp. 2978–2983 (2009)
4. Saiki, H., Fukao, T., Urakubo, T., Khno, T.: Hovering control of outdoor blimp robots based
on path following. In: 2010 IEEE International Conference on Control Applications (CCA),
pp. 2124–2129 (2010)
5. Mettler, B., Kong, Z.: Receding horizon trajectory optimization with a finite-state value function
approximation. In: Proceeding of 2008 American Control Conference, Seattle, Washington,
pp. 3810–3816 (2008)
6. González, P., Burgard, W., Sanz, R., Fernadez, J.L.: Developing a low-cost autonomous indoor
blimp. J. Phys. Agents 3(1), 43–51 (2009)
7. Fukao, T., et al.: Inverse optimal velocity field control of an outdoor blimp robot. In: Proceedings
of the 17th World Congress the International Federation of Automatic Control, Seoul, Korea,
pp. 4374–4379 (2008)
8. Rottmann, A., Plagemann, C., Hilgers, P., Burgard, W.: Autonomous blimp control using model-
free reinforcement learning in a continuous state and action space. In: Proceeding of Interna-
tional Conference on Intelligent Robots and Systems, IROS 2007, pp. 1895–1900 (2007)
9. Gammon, S.M., Fryr, M.T., Qian, C.: The mathematical model of the tri-turbofan airship for
autonomous formation control research. In: IEEE Region 5 Technical, Professional, and Student
Conference, San Antonio, Texas (2006)
10. Azinheira, J.R., et al.: Visual servo control for the hovering of an outdoor robotic airship. In:
Proceedings of the 2002 IEEE International Conference on Robotics & Automation, Washing-
ton, DC, pp. 2787–2792 (2002)
11. Machry, C.R.: As Potencialidades de Emprego dos Dirigíveis no Transporte de Carga. Reista
da UNIFA (2003)
12. Navarro, I., Matía, F.: An introduction to swarm robotics. ISRN Robot. 2013, Article ID 608164.
http://dx.doi.org/10.5402/2013/608164 (2013)
13. Fossen, T.I.: Guidance and Control of Ocean Vehicles. Wiley, West Susses, England (1994)
14. Hardkernel Co., Ltd.: Products, 24 January 2013. http://www.hardkernel.com/renewal_2011/
products/prdt_info.php (2013)
15. Ribeiro, C.G., et al.: A platform for autonomous path control of unmanned airship. J. Braz.
Soc. Mech. Sci. Eng. (2017). https://doi.org/10.1007/s40430-017-0891-9. ISSN 1678-5878,
Rio de Janeiro, Brazil
16. Becker, D.J., Sterling, T.: Beowulf: a parallel workstation for scientific computation. In: Pro-
ceedings of the International Conference on Parallel Processing (1995)
17. ARDUINO: Arduino, 25 March 2013. http://www.arduino.cc/ (2013)
18. Fossen, T.I.: Marine control systems, guidance, navigation, and control of ships, rigs and
underwater vehicles. In: Marine Cybernetics, Trondheim, Norway (2002). ISBN 82-92356-
00-2
19. Reschke, C., Sterling, T., Ridge, D.: A design study of alternative network topologies for the
Beowulf parallel workstations. In: Proceedings of the IEEE International Symposium on High
Performance Distributed Computing (1996)
20. Ralph, N.: Odroid-x development board brings quad-core Exynos 4 Quad processor to budding
Android hackers for $129, 24 January 2013. http://www.theverge.com/2012/7/13/3156032/
odroid-x-development-board-exynos-4412-quad (2012)
21. Larabel, M.: Quad-Core ODROID-X Battles NVIDIA Tegra 3, 24 January 2013. http://www.
phoronix.com/scan.php?page=article&item=samsung_odroidx&num=1 (2012)
22. redOrbit Staff & Wire Reports - Your Universe Online: Miniature Quad-Core Computer
For Under $130, 24 January 2013. http://www.redorbit.com/news/technology/1112656695/
miniature-quad-core-computer-for-under-130/ (2012)
23. Lidar USA 3D Documentation and Beyond: Lidar USA 3D Documentation and Beyond. http://
www.lidarusa.com/page.php?cat=3, 7 July 2013 (15 de December de 2013)
24. Texas Instruments: LIDAR System Design for Automotive/Industrial/Military Applications,
SIGNAL PATH designer SM. Tips, tricks, and techniques from the analog signal-path experts
Texas Instruments, Literature Number: SNAA 123 (2013)
25. i Badia, S.B., Pyk, P., Verschure, P.F.: A biologically inspired flight control system for a
blimp-based UAV. In: Proceedings of the 2005 IEEE International Conference on Robotics
and Automation, pp. 3053–3059 (2005)
26. U.S. Department of Homeland Security: NAVIGATION RULES ONLINE, 24 January 2013.
http://www.navcen.uscg.gov/?pageName=navRulesContent (2013)
Assessing the Location of Search
and Rescue Stations on the Portuguese
Coast
Anacleto Correia , Ricardo Moura and Miguel Fonseca
Abstract The compliance, by a signatory country, with the International Convention

on Maritime Search and Rescue (SAR) requires maintaining adequate means of
search and rescue at sea. An important component of such infrastructure is the life-
saving stations, distributed along the coast and equipped with human resources and
equipment, such as lifeboats. The assessment of the best geographical location of life-
saving stations (LSS) can be carried out considering multiple criteria. Among them,
one can mention: the proximity to areas historically associated with the occurrence of
accidents; the maritime traffic pattern in the waters under surveillance; the severity of
accidents (in terms of loss of human life and environmental impact of accidents); the
typology and cause of the recorded accidents, as well as, the technical characteristics
of the lifeboats assigned to life-saving stations (e.g., autonomy, speed). This work
aimed to assess the adequacy of the current geographic distribution of the resources of
the maritime rescue system on the Portuguese coast, and specifically the location of
the life-saving stations, considering the history of the number of accidents at sea. For
this purpose, spatial analysis tools were used to compare georeferenced information
on the location of the LSS, as well as the autonomy of the lifeboats assigned to the
stations, with maritime accidents occurred in their proximity in recent years. The aim
was to assess the lifeboats’ degree of coverage compared to the location of maritime
accidents registered. In addition, the work also aimed at eliciting the functional
requirements of a decision support system for maritime search and rescue.
Keywords Geographic information systems · Spatial analysis · Study of

location · Maritime accidents · Search and rescue
A. Correia (B) · R. Moura

CINAV—Alfeite, 2810-001 Almada, Portugal
e-mail: cortez.correia@marinha.pt
R. Moura · M. Fonseca
CMA, Centro de Matemática e Aplicações, 2825-149 Caparica, Portugal

https://doi.org/10.1007/978-981-13-9155-2_26
322 A. Correia et al.
1 Introduction
The ocean is a development factor that makes possible different traditional and emerg-
ing economic activities, such as fishing, shipping transportation, tourism, and recre-
ative sailing, among others. Accidents caused by maritime activities result in heavy
losses of human lives every year; hence, the need to effectively prevent and respond
to emergencies at sea. Following a maritime accident, response time is critical, as the
delay of a rescue is often a factor that makes the difference between life and death
[1].
The first edition of the SOLAS Convention, in 1914, established a set of responsi-
bilities related to safety at sea, with an emphasis on safety standards and obligations
of shipowners and crews. Later, when many coastal countries already had services
for the rescue of castaways, the Convention on the High Seas, in 1958 at Geneva,
assigned greater responsibility to coastal countries on safety of ships crossing waters
under their jurisdiction, especially near the coastal line, due to the higher propensity
to the occurrence of maritime accidents and, not infrequently, given the greater dif-
ficulty usually found to give help. Therefore, more maritime safety responsibilities
have been assigned to maritime countries, in particular on safeguard human life at
sea.
Nowadays, more than one hundred countries have ratified the Search and Rescue
(SAR) Convention of 1979, in particular, almost all coastal countries, accepted their
responsibilities regarding the international maritime community. However, it was the
SAR Convention, established since 1985, which structured the international organi-
zation for maritime search and rescue. As a result of the convention, the oceans were
divided into regions (Search and Rescue Regions—SRR). For the coastal countries,
new responsibilities were assigned regarding the coordination of maritime safety
actions in their waters (see Fig. 1a). This responsibility entails the existence of emer-
gency service at sea, with immediate readiness (24 h a day), able to respond to all
kinds of accidents related to the endanger of human life at sea. The size of the regions
was defined according to the capacity of response of coastal countries to distress calls.
Portugal is a country on the periphery of Europe with an extensive maritime area,
resulting from the continental coastlines and the archipelagos of Madeira and the
Azores. The maritime space under sovereignty or jurisdiction of the country includes
regions with responsibility for maritime search and rescue (see Fig. 1b), which are
crossed by intense maritime, commercial, and recreative traffic (see Fig. 1c).
As one of the signatories of the SAR Convention, Portugal has an accident
response system at sea to assist ships in distress in the maritime areas of the con-
tinent and the North Atlantic. The execution of this mission is carried out by a
national maritime search and rescue service, responsible for responding to the emer-
gency from ships or vessels, in the areas of national responsibility. The search and
rescue actions are addressed to one of the two centers responsible for coordinating
the two Portuguese SRR: (1) the Maritime Rescue Coordination Center (MRCC)
Lisbon—supported by the Maritime Rescue Sub-Center (MRSC) Funchal, respon-
sible for SRR Lisbon, which covers the coast of the continent and the archipelago of
Assessing the Location of Search and Rescue Stations … 323
Fig. 1 a SRRs around the world (Source IMO); b Portuguese SRRs superimposed on maritime
spaces under Portuguese sovereignty or jurisdiction; and c heat map of maritime traffic density in
Portuguese SRRs in the year 2017 (Source Marine Traffic)
Madeira and (2) the MRCC Delgada, which has under its jurisdiction the SRR Santa
Maria, in the North Atlantic.
Important infrastructures, to support of SAR actions, are life-saving stations
(LSS), located along the coast. These means carry out annually numerous actions of
maritime search and rescue and relief to castaways, as happened in 2017, the year in
which there were recorded 164 accidents with vessels into the two Portuguese SRR.
The proper location of these resources of the national maritime search and rescue
system is, therefore, an important factor in the effectiveness of SAR missions.
The objective of the present work is to assess the adequacy of the configuration
of the life-saving stations on the Portuguese coast, given the historical occurrence
of maritime accidents in the SRRs under Portuguese responsibility. The article was
developed in five sections. In this section, the motivation of study was presented. The
second section presents a literature survey on maritime SAR studies. The third section
characterizes the profile of the maritime accidents and the maritime rescue stations
on the Portuguese coast. The assessment of the location adequacy of the set of LSS,
compared to the registry of marine accidents, is carried out in the fourth section, as
well as the elicitation of the requirements for a decision support system for maritime
SAR activity, to be developed. In the last section, we present the conclusions of the
work carried out, as well as the future work.
2 Literature Review
Plenty of location studies regarding maritime SAR have been carried out to determine
or optimize the location of infrastructures or resources of the emergency services.
Basdemir conducted a study, in 2004, for the Turkish Air Force [2] with the pur-
pose of distributing new SAR facilities across the Western Mediterranean region and
the Aegean Sea. With this purpose, he used the Maximal Coverage Location Prob-
lem (MCLP) model to define the locations for the required number of SAR stations,
in order to obtain full coverage of the accident zones considering the operational
capabilities of SAR units, in this case, helicopters.
In another work, Afshartous et al., studied the best location for United States
Coast Guard air stations (USCG), in Miami, to respond more adequately to emer-
gency calls [1]. A simulation/optimization methodology was used which, based on
real emergency call data, simulated the variation of the location and occurrence of
emergency calls, in space and time. Later on, by using the pUncapacitaded Facility
Location Problem (p-UFLP), the pMedian Problem (p-MP) and the Uncapacitaded
Facility Location Problem (UFLP), they achieved the best location for the stations be
able to respond to emergency calls from an uncertain location. However, the model
was a simplification that considered only few criteria in the analysis.
In the 1990s, a study from Brown et al. resulted in the mixed Integer Linear Pro-
gramming (ILP) model used for plan the assignment of patrol missions to vessels, of
United States Coast Guard (USCG), in several districts [3]. The study was intended
to meet the needs of the patrol, considering the weekly maintenance schedule of
the vessels, while reducing, at the same time, the related costs. Azofra et al. pro-
posed a tool to objectively distribute the resources assigned to the maritime SAR [4].
To achieve this aim, they developed a methodology based on gravitational models.
Firstly, they applied the Zonal Distribution Model (ZDM), to determine the areas
with the highest concentration of accidents from the registry of marine accidents in
the area. Each zone was represented by a super-accident centroid. Later they defined
an Individual Distribution Model (IDM), which were able to provide the distribution
of individualized resources to the maritime rescue actions, directing those resources
to the locations of the accident, taking in account the distance to the super accident
and their range of action.
Radovilsky and Koermer built another ILP model to optimize the distribution
of small USCG vessels through existing stations in the Pacific area [5]. The main
objective of this model was to provide the best distribution of the vessels, minimiz-
ing the shortage or overload on stations’ capability, and simultaneously reduce the
operational costs. Later, Wagner and Radovilsky improved the previous work by
developing the ILP model called the Boat Allocation Tool (BAT) model, considering
as the main variable the total of annual hours available for each vessel, i.e., the hours
available until each vessel needed maintenance, instead of the number of vessels
only [6]. They also introduced risk management, considering a possible shortage of
vessels, as well as, the uncertainty in the future needs of the stations. The model also
considered some logistic restrictions of the USCG, such as the number and type of
vessels available per station.
Another interesting research, developed by Pelot el al., was a location study for
the Canadian Coast Guard Search and Rescue vessels in the Atlantic region in order
to ensure maximum coverage of the accidents possible to occur in the area [7].
Based on the Maximal Coverage Location Problem (MCLP), they proposed three
extensions: (1) the MCLP model, which assigned weights to different classes of
maritime accidents in order to compute the best location of the vessels to provide
the maximum coverage for the required areas; (2) the MCLP that computes the best
location of the vessels in order to allow maximum coverage by equally distributing
the workload through the stations; and (3) the MCLP which considers the uncertainty
on vessels availability. The objective of this model was to maximize the coverage
considering the probability of boats’ absent, as happens when, for example, they are
on maintenance period.
More recently, some studies with new perspectives of the localization problem
regarding the maritime SAR domain have been presented. They considered simul-
taneously multiple aspects of the problem. In this context, Razi and Karatas [8]
developed the Incident Based-Boat Allocation Model (IB-BAM), a multi-objective
model to assign vessels to maritime SAR activities in Turkey [8]. In this model,
the Analytical Hierarchy Process (AHP) was used to assign different weights to
each type of maritime incident in the area of interest. Subsequently, a ZDM was
used to determine highest probability zones for occurrence of maritime incidents.
Finally, the Multi-Objective Mixed Integer Program (moMIP) was applied to pro-
vide low response times to maritime incidents, while operating costs minimization
is imposed, as well as the compatibility between workload and available hours for
operations.
In 2017, Akbari et al. presented a couple of studies focused on the develop-
ment of structures to carry out analyzes considering multiple criteria in the location
and distribution of resources. The studies targeted the Canadian Coast Guard SAR
operations in the Atlantic region, aiming the improvement of quality service and
operational costs reduction. In the first study, they presented a multi-criteria analysis
on the performance of solutions provided by two well-known models, the MCLP and
p-Median Problem (p-MP), for the case of maritime SAR location [9]. The optimal
solution of the two models was obtained based on five decision criteria relevant for
CCG: (1) access time—criterion that evaluates the average time that the nearest ship
takes to arrive at the incident local (time spent in transit); (2) primary—criterion
that measures the percentage of incidents that must be covered within a given access
time by, at least, one vessel; (3) backup coverage—criterion that expresses the per-
centage of incidents that are within a given coverage region of at least two vessels;
(4) Gini index—criterion that is measured by the level of deviation of access time to
all accidents; and (5) Maximum access time. To model the problem, they considered
the characteristics of vessels types, the existing SAR stations, and, based on the his-
tory of maritime incidents in the area, they forecast the areas where future maritime
incidents could occur, in order to anticipate future SAR needs. Subsequently, Akbari
et al. developed a Goal Programming Multi-Objective model to optimize the location
of CCG vessels in order to achieve high coverage and responsiveness of probable

future incidents [10]. The objectives were established on a basis of information and
needs provided by the CCG. In order to reach the proposed objectives, they con-
structed the Mixed Linear Problem model (MILP) where three criteria were used:
primary, backup coverage, and mean access time. The model considered different
classes and types of vessels with different speeds, capacities, potential locations, and
also included operational constraints to simulate possible ship unavailability. As part
of the model, it was also considered the spatial distribution of the history of maritime
incidents to simulate new occurrences of accidents.
In this exploratory work, rather than use algorithmic models, geospatial tools are
used to compare the areas where maritime accidents have been registered, with the
range of the rescue means assign to stations, in order to assess the degree of coverage
of current stations’ configuration regarding previous incidents.
3 Maritime Accidents on the Portuguese Coast
By exploring data through statistics collecting and georeferencing of maritime acci-

dents, the aim was to understand the phenomenon and get some clues on how to
plan and improve SAR response. The analysis of the history of offshore accidents,
revealed their frequency and the most vulnerable areas, to be compared with the
arrangement of the overall rescue facilities.
To understand the type of maritime accidents in Portuguese waters, the annual
trend of the accidents (a total of 791) was analyzed over the period from 2011 to
2016. During that period, the year of 2013 was the one that registered the greatest
number of accidents (164), while the year 2015 had the least occurrences (111), with
a slight decrease in the number of occurrences at the end of the period (see Fig. 2a).
A monthly analysis shows that July is the month with the most accidents (see
Fig. 2b), during the summer, in which, although the meteo-oceanographic conditions
are favorable to maritime activities, there is an increase in the number of recreational
boats at sea. The type of vessels that contributed most for the number of accidents is,
precisely, the recreational craft (444 in the total of the period, as depicted in Fig. 2c).
However, the number of accidents shows a large variation over the period, with a
maximum of 103 accidents recorded in 2013, to a minimum of 50 recorded in 2015.
On the other hand, the number of accidents (254) of fishing vessels (the second
type of vessel with more registered accidents) is relatively constant throughout the
period. Regarding the types of accidents recorded, the most frequent was vessels’
breakdown, followed by sinking and collision.
By analyzing the locations where the accidents occurred, the region of Lisbon has
the highest number, followed by Funchal in Madeira. With an expressive contribu-
tion, there are also the Viana do Castelo, Setubal, and Aveiro regions where fishing
activity is significant. Overall, 84.5% of accidents (668) occurred in continental
waters, 8.1% (64) in the Madeira region and 7.4% (59) in the Azores region.
(a) (b)
(c)
Fig. 2 Distribution of the number of accidents (2011–2016): a per year, b per month, and c by
type of vessel
Fig. 3 Distribution of marine accidents between 2011 and 2016 (#791): a overall; b with fishing
boats; c with recreational boats; and d SAR actions between 2015 and 2017
Fig. 4 Distribution of maritime rescue stations in the Azores, Madeira, and the Continent in 2017
Figure 3a shows the global picture of accidents in Portuguese waters, allowing

to perceive the occurrence of accidents in inland waters. Figure 3b and c depict,
respectively, the locations of accidents involving fishing and recreational vessels,
where can be noted a close proximity of the coastline. As can be seen in Fig. 3d,
there is a notorious geographical coincidence between the locations with mapped
accidents and the SAR occurrences between the years 2015 and 2017.
Regarding the number of rescue stations, there are 28 LSS. Their geographic
distribution is depicted in Fig. 4. The LSS is of four types (types-A to D) with each
typology having different kinds of infrastructures, staff number, as well as operational
and rescue means. Each LSS of type-A has lifeboats with greater autonomy, as well
as garrisons capable of operating them. The LSS of type-D, on the other hand, has
small rubber boats and water scooters, for use along the seafront.
4 Assessment of the Maritime Rescue Stations

Configuration
This section describes the process of assessment of current configuration of life-

saving stations, spread by Portuguese coast, compared to the areas in which there
have been a greater number of maritime accidents in Portuguese waters.
The criteria followed by the maritime rescue authorities to distribute the life-saving
stations through the Portuguese coast have been the following: (a) the number and
type of previous occurrences of maritime accidents; (b) neighborhood areas of major
fishing communities; (c) neighborhood areas of major commercial ports; (d) quick
access to areas of greater maritime traffic density; (e) type of the maritime activity
developed in the area; and (f) distance between life-saving stations and the range of
the lifeboats in each LSS.
After the data gathering process, the information was stored in a geographic
database (PostgreSQL with PostGIS extension) [11], after that the maritime accidents
Fig. 5 a Heat map of major density areas of maritime accidents, compared with the location of
life-saving stations b Maximum coverage of the lifeboats in each LSS, compared with locations of
maritime accidents
as well as the lifeboats coverage were depicted in maps (see Fig. 5), using a geographic
information system (QGIS) [12]. The produced maps were built to help perform the
spatial analysis in order to assess the current locations and the coverage of the rescue
means assigned to the life-saving stations.
The analysis of the heat map of Fig. 5a allows to conclude that most of the zones
with greater density of accidents have a life-saving station assigned. The exception
is the extreme north of the continent (Caminha), where it would be pertinent the
implantation of a life-saving station, given the difficult existing traffic in the waters
of the Minho’s river mouth. The remaining life-saving stations are located at enough
proximity of areas of higher accident density to access them in emergency situations.
In Fig. 5b, one can observe the depiction of the sea area covered by the rescue means
currently available in the life-saving stations. The depicted buffer from each one of
the life-saving station show the limits of intervention (i.e., autonomy) of the means
assigned for each LSS, which are enough to ensure the coverage of the distances for
the zones of greater density of accidents.
This work also sought for the requirements elicitation for future development of
a decision support system for maritime rescue and management of the life-saving
stations. The overall objectives, considered relevant for the system were: (a) maxi-
mization of the quality of service (measured in response time) of rescue means; (b)
minimization of operating costs; (c) minimization of the transit time/distance trav-
eled by the rescue means; and (d) minimization of the number of life-saving stations,
without compromise the fulfillment of the search and rescue mission.
In addition to the criteria currently used for stations’ location assignment, referred
at the beginning of this section, it was considered of paramount importance that the
decision support system should always prioritize the rescue of people when allocat-
ing the means for rescue. Furthermore, were also considered as requirements for the
system, the following: (a) classification of historical data on maritime accidents by
areas, envisioning future occurrence of accidents, considering assignment of weights
based on, inter alia, the type of vessels, the location, and severity of previous acci-
dents; (b) counseling on the assignment of existing marine rescue means, given their
maximum speed, maximum range with maximum load, as well as operational limits
based on the sea turbulence at the rescue area; (c) minimization of the average time
to respond to an accident (considering the sum of the time elapsed from the dis-
tress call for the life-saving station, the time spent in transit and the time needed for
on-the-spot searches); (d) incorporation of meteo-oceanographic conditions at the
location and time of the accident and subsequent evolution in the search and rescue
zone; (e) real-time monitoring of risky areas for the occurrence of accidents due
to the existence of fishing communities and greater density of maritime traffic; (f)
consideration of constraints on the availability of life-saving stations means due, for
instance, to vacation period of staff, maintenance of the lifeboats or their occupancy
in another rescue mission; and (g) incorporation of seasonal factors in the analysis.
5 Conclusion
The readiness of the maritime search and rescue system, on the Portuguese coast,
is based on a set of life-saving stations of different types located and equipped
with rescue means with different capabilities. Using geospatial information, it was
assessed the degree of adequacy of the rescue stations locations regarding the actual
needs, given the history of accidents in recent years in Portuguese waters. From the
work carried out, it was concluded that the current arrangement of life-saving stations
is generally adjusted for the location of previous registered maritime accidents. There
is, however, one adjustment that must be considered, which involves the setup of an
LSS at the north of the continent, where there is a significant fishing and recreational
local community and also a significant number of accidents the sea.
The work also led to the elicitation of requirements for the development of a
decision support system to advise on how to use existing rescue means by applying
predictive techniques to the historical data collected, taking into account the technical
characteristics of the rescue means, the typology of accidents, as well as, meteo-
oceanographic conditions in the risk zones.
Acknowledgements The work was funded by the Portuguese Ministry of Defense and by the
Portuguese Navy/CINAV.
References
1. Afshartous, D., Guan, Y., Mehrotra, A.: US Coast Guard air station location with respect to
distress calls: a spatial statistics and optimization based methodology. Eur. J. Oper. Res. 196(3),
1086–1096 (2009)
2. Basdemir, M.: Locating search and rescue stations in the Aegean and Western Mediterranean
Regions of Turkey. J. Aeronauti. Space Technol. 1(3), 63–76 (2004)
3. Brown, G.G., Dell, R.F., Farmer, R.A.: Scheduling coast guard district cutters. Interfaces 26(2),
59–72 (1996). https://doi.org/10.1287/inte.26.2.59
4. Azofra, M., Pérez-Labajos, C.A., Blanco, B., Achútegui, J.: Optimum placement of sea rescue
resources. Saf. Sci. 45(9), 941–951 (2007)
5. Radovilsky, Z., Koermer, T.: Allocation of US Coast Guard boats utilizing integer programming.
J. Academy Business Econ. 7(2), 130–135 (2007)
6. Wagner, M.R., Radovilsky, Z.: Optimizing boat resources at the US Coast Guard: deterministic
and stochastic models. Oper. Res. 60(5), 1035–1049 (2012)
7. Pelot, R., Akbari, A., Li, L.: Vessel location modeling for maritime search and rescue. In:
Eiselt, H.A. Marianov, V. (eds). Applications of Location Analysis (2015)
8. Razi, N., Karatas, M.: A multi-objective model for locating search and rescue boats. Eur. J.
Oper. Res. 254(1), 279–293 (2016)
9. Akbari, A., Eiselt, H.A., Pelot, R.: A maritime search and rescue location analysis considering
multiple criteria, with simulated demand. INFOR: Info. Syst. Oper. Res. 56(1), 92–114 (2017).
https://doi.org/10.1080/03155986.2017.1334322
10. Akbari, A., Pelot, R., Eiselt, H.A.: A modular capacitated multi-objective model for locating
maritime search and rescue vessels. Ann. Oper. Res. 1–26 (2017)
11. PostgreSQL.: https://www.postgresql.org/. Accessed on Dec 2018
12. QGIS.: https://www.qgis.org/. Accessed on Dec 2018
Operational Scenarios Identification
and Prioritization in Systems of Systems:
A Method and an Experience Report
in the Defense Domain
Carlos Eduardo de Barros Paes and Valdemar Vicente Graciano Neto
Abstract Military systems play a crucial role in the defense of a country. They
are often interconnected to monitor the frontiers, increase the national sovereignty,
and interoperate to offer a unified control of the entire territory. These alliances
of systems are currently known as systems of systems (SoS). During the last years,
software has been increasingly embedded in those systems to increase their precision.
Besides, multiple cyber-physical systems have also been associated to interoperate
in the context of SoS, bringing novel challenges related to the conception of those
systems, in particular (i) how to elicit systems’ requirements and make them coexist
in the same large-scale system, (ii) how to coexist competing requirements, (iii) how
to support the distributed development of constituent systems and interoperability
links, (iv) how to capture their conceptual operation in a unified document to exploit
their functionalities in a fruitful manner, and (v) how to make the SoS cohesive,
exploiting their functionalities and creating emergent behaviors to foster the national
security. Given those gaps and challenges, this paper reports an experience of a real
project carried out in a real Department of Defense, particularly in the navy’s context,
besides externalizing a method called OSW to support the ConOps documents in SoS
engineering projects. We report the conception of a SoS for defense and maritime
surveillance, focusing on the concept of operations (ConOps) of that systems, the
challenges faced, and the method established. We claim that this report contributes
to the engineering of defense SoS since it provides a panorama of SoS conception,
besides presenting solutions that can be replicated in other projects to be carried out
worldwide, contributing thus for the advance of SoS research in that domain.
Keywords Systems-of-systems engineering · Experience report · National

defense and security
C. E. de Barros Paes (B)

Pontifical Catholic University of São Paulo, São Paulo, Brazil
e-mail: carlosp@pucsp.br
V. V. G. Neto
Federal University of Goiás, Goiânia, Brazil
e-mail: valdemarneto@inf.ufg.br

https://doi.org/10.1007/978-981-13-9155-2_27
334 C. E. de Barros Paes and V. V. G. Neto
1 Introduction
Defense systems of systems (D-SoS)1 are a cutting-edge technology. They comprise

a set of multiple military constituent systems connected through interoperability links
to (i) unify functionalities offered by those systems, (ii) increase the robustness and
reliability of the entire military technology set, and (iii) enable the integration of navy,
army, and aeronautics forces, obtaining, consequently, a reinforcement of national
security and sovereignty. Crucial emergent behaviors have been achieved by them,
such as the defense of a country itself. In this sense, SoS engineering (SoSE) has
become strategic, guiding how to develop constituent systems and offering insights
on how to interoperate them to accomplish complex missions.
Despite the existence of D-SoS since the 1960s, many challenges still remain. D-
SoS have become software-intensive, i.e., systems in which software development
is a dominant element that crosscuts the entire system life cycle [1]. Therefore, they
have imposed important challenges. Specifically regarding to requirements engi-
neering issues, we can mention the need for (i) dealing with distributed teams of
different defense branches (army, navy, and aeronautics) with potentially competing
requirements, (ii) establishing and overcoming communication issues due to the dis-
tributed nature of such development, (iii) capturing those potentially conflicting and
competing requirements in a unified document, and (iv) maintaining them to achieve
reasonable quality in the final obtained D-SoS.
In this direction, the main goal and contribution of this paper are (i) to present a
novel method that was used in a real project to support the elaboration of a concept
of operations (ConOps) document [2, 3] and (ii) to share an experience report that
captures expertise knowledge that could be shared and reused to aid the conduction
of other similar SoSE development projects. Such contributions are of prominent
importance, since this knowledge is often restricted to confidential documents that
are not broadly available. In this paper, we analyze a real Navy Management SoS
project composed of a set of interoperable systems already in use to collect, share,
analyze, display operational information, and provide decision support regarding
the national security of a given country [4, 5, 6]. Many of these systems are by
the Brazilian Navy and others will be developed and integrated during the systems
development. The remainder of this paper is organized as follows. Section 2 presents
a background. Section 3 presents an overview about the SisGAAz (Blue Amazon
Management System). Section 4 outlines the method used for elaboration of ConOps
document in the SoS projects. Section 5 presents a case study of operational scenarios
elaboration in the SisGAAz. Section 6 discusses the analysis of the results obtained
during the identification and prioritization of the operating scenarios. Finally, Sect. 7
presents final remarks.
1 For sake of simplicity, D-SoS and SoS will be herein used interchangeably to express both singular
and plural forms of SoS and to denote military and non-military SoS.
Operational Scenarios Identification and Prioritization … 335
2 Background
SoS result from multiple operationally and managerially independent systems (called
constituents) that work together to reach common goals. Each constituent performs
its individual mission contributing to the success of the global missions [7]. Each
different architectural arrangement that results from different combinations of con-
stituent systems is named a coalition [8]. SoS have been distinguished from single
systems by a set of characteristics, as follows [9]: (i) operational independence of
the constituent systems, since constituent systems share their resources to contribute
to the accomplishment of goals at SoS level, but also hold their local activities being
performed; (ii) managerial independence of the constituent systems, which are man-
aged by independent organizations and stakeholders (examples include army, navy,
and aeronautics managing constituents of a same D-SoS); (iii) evolutionary develop-
ment of the SoS, since the SoS evolve as a consequence of the individual evolution
of its constituents, their functions, and purposes; (iv) distribution, since constituents
interoperability is necessarily supported by a communication technology; and (v)
emergent behavior, which comprises novel functionalities obtained from the coop-
eration among constituent systems that could not be offered by a single constituent,
and that raises as a result of a coordinated interoperability among them. In particular,
missions are also an important concept, especially in D-SoS domain.
SoS are often developed under the umbrella of systems engineering (SE), an
interdisciplinary approach to enable the realization of successful system under a
broad meaning of the word [3]. Several SE processes have been proposed and adapted
by different international standard organizations for construction of SoS, including
(i) those evolved from the DoD [10], proposed in the 1970s by the US Department of
Defense, (ii) other standard systems engineering processes proposed by industry in
the 1980s, and (iii) those based on the ISO/IEC 15288 standard, which was adopted
by International Council on Systems Engineering (INCOSE) and the IEEE to form
the basis of the system-engineering handbook [11, 12]. Most of these standards and
systems engineering process are widely used in the technology industry, aerospace
and defense departments of several countries.
One of the important artifacts developed during the early phases of the systems
engineering process is SoS operational concept (ConOps). The ConOps document
describes how functionalities will be employed in an operational coalition to achieve
the missions established for SoS [13]. It is developed in parallel with the capability
objectives. Capability objectives are a statement of top-level objectives for the SoS.
They describe the capabilities needed by the user, ideally based on some definitive
or authoritative materials. The capability objectives provide a basis for translating
operational needs into high-level requirements, assessing performance to capability
objectives, and developing an architecture and solution options [13]. As the capability
objectives evolve, the ConOps should evolve in detail, as well. ConOps is used to
define the SoS requirements space to identify aspects of systems that can affect
the SoS architecture and to select performance metrics and test environments [14].
However, SE guides do not recommend which techniques should be used or adopted
to support the ConOps document in the context of SoS projects. Since (i) the scope of
SoS can cover the entire territory of a country, (ii) the teams are distributed, and (iii)
involve multiple actors, it is necessary to use a technique to identify the operational
scenarios of each set of stakeholders of each constituent involved or impacted by the
SoS. Section 4 details the technique proposed for SoS context. Next section brings
an overview of the SoS engineering project used as input to conceive and apply the
method to support the elaboration of the ConOps document for a real military SoS.
3 SisGAAz Overview
The Brazilian Navy has developed a system called SisGAAz to meet the guide-
lines of the National Defense Strategy for managing an important maritime area that
surrounds the territory and corresponds to the international areas of responsibility
for the search and rescue operations (SAR—search and rescue) and areas of spe-
cific interest that go beyond the Brazilian Jurisdictional Waters (BJW) and the SAR
area [5, 6]. SisGAAz supports navy’s activities in its constitutional allocations and
subsidiaries’ assignments, such as protecting the national frontiers and ensuring the
sovereignty. The project involves a diverse team with several professional roles such
as telecommunication engineers, systems engineering, software engineers, domain
experts, project managers, quality managers, and infrastructure technicians, totaliz-
ing around 30 people involved.
SisGAAz mission is monitoring and controlling, in an interoperable way, the
national maritime area (waters under the Brazil jurisdiction), international areas of
responsibility for the search and rescue operations, and areas of specific interest that
go beyond the previous ones, in order to contribute to the strategic mobility, repre-
sented by the ability to respond promptly to any threat, emergency, aggression, or
illegality. SisGAAz’s main purposes are: (i) information sharing and interoperabil-
ity among the institutions with interest in the sea (national and international, public
and private) and (ii) supporting network-centric operations and providing decision
support by a shared virtual environment.
SisGAAz development was partitioned into two phases: conceptualization and
development. The conceptualization phase involved the system specification and
design (operational concept and requirements), elaboration of project management
plans, and architecture design. This phase was completed in two years and it was
developed in the following four tasks: (i) project management planning; (ii) elabora-
tion of concepts of operations (ConOps); (iii) elicitation, analysis, and documentation
of system requirements; and (iv) conception and design of SoS architecture.
The development of SisGAAz follows an evolutionary approach (iterative and
incremental). Initially, a small core of functionalities is already implemented and
deployed. Subsequently, the implementation activities are performed in a way that
new functionalities are added to constituents according to the local demands of an
operational area. For example, one of the navy districts is in State 1 (the operational
area) and another one in State 2. State 1 district hosts the first constituents imple-
mented according to the local demands. After that, the implementation will take place
in State 2. The new constituents developed there match their specific demands, and
after developed, they are joined to the first ones to form the preliminary body of the
SoS. Next, this work is iteratively carried out in other regions of the Brazil, creating
new constituents, connecting them to the others, realizing the SoS, and expanding
the SisGAAz SoS progressively.
The proposed SoS has five types of constituents, each one responsible for one of
the following assignments: logistics support, cybersecurity, intelligence, command
and control (C2), communications, and remote sensing. Logistics constituent aims to
provide support to logistics operations regarding navy’s equipment, such as acquisi-
tion and transport of material to national oceans supervision. Cybernetics constituent
provides capabilities related to SisGAAz safety and security aspects, supporting (i)
availability and assurance of confidentiality of information and services, (ii) defense
and protection of networks (internal and external), (iii) monitoring the full spec-
trum of cyber warfare operations, and (iv) safety and protection of services and
constituents that form the SisGAAz. The command and control (C2) constituents
are military systems that provide features to support planning of command opera-
tions and control of constituents in the accomplishment of the mission, playing a
type of central authority role. The communications constituents provide the techno-
logical infrastructure to integrate the various military organizations and SisGAAz
constituents. The intelligence constituent aims to support planning and execution of
operations, according to the ConOps. Finally, the remote sensing constituent pro-
vides functionality for collecting, processing, and distributing data from a variety
of sensors, new or legacy, which make up the SisGAAz and data received through
cooperation with other organizations. Each Brazilian geographic area can host a sub-
set of those types of constituents that are linked together to form the national-scale
SoS known as SisGAAz. Figure 1 illustrates the scenarios of interoperability among
Brazilian Navy’s constituents in the context of SisGAAz.
4 OSW: A Method for Elaboration of ConOps Documents

in the SoS Projects
Since SE guidelines do not recommend methods to support the ConOps conception

for SoS development projects, we have proposed and applied a technique so-named
Operational Scenarios Workshops (OSW) to support the elaboration of a ConOps
for a SoS project context.
Developed in two phases, the first phase was conducted in a three-day work-
shop exclusively run with Brazilian Navy organizations, focusing on their opera-
tional commands and naval districts. The second phase involved workshops with
stakeholders that were external to the Brazilian Navy, including agencies and actors
directly related to maritime activities. In addition, an approach for the identification
of critical scenarios proposed by the Naval War School of the Brazilian Navy was
Fig. 1 SisGAAz operational context and navy constituents
applied to drive the information collection in order to elaborate ConOps [6]. This
approach consisted in the identification of threats, critical activities and areas, and
a merging of these three sets to define priorities. That was fundamental and helped
to design the operational scenarios to which SisGAAz should respond. Based on
these operational scenarios, ConOps was specified following the IEEE standard for
information technology, the IEEE 1362 [15]. As a result, the Operational Concept
Validation Plan was produced, containing instructions to validate the ConOps. The
OSW method was structured in the following steps:
Step 1. SoS Operational Scenarios Identification and Prioritization: This step
consists in the establishment of the interactions between the critical activities
of the SoS and the threats considering the areas of action of each constituent.
Step 2. Identification of SoS Critical Activities: This step involves the identi-
fication of the activities of interest of each constituent and the targets of
monitoring and protection fulfilled by the SoS.
Step 3. Identification of threats and emergencies: This step comprises the iden-
tification of the threats in which critical activities need to be preserved by
the SoS. In addition, the main emergencies that may occur and that need to
be addressed by SoS are also identified.
Step 4. Constituent area identification: In this step, the geographical areas of
each SoS constituent are elicited. This information is important for the
elaboration of the scenarios, since it allows to identify the critical activities
carried out in each area and their respective threats and emergencies that
need to be considered.
Step 5. SoS critical activities confronting: This activity comprised the confronta-
tion of the identified critical activities with the perceived threats and emer-
gencies. The combination of the elements previously considered elements

was used to form the scenarios of interest of each constituent to act in the
SoS.
Step 6. SoS operational scenario prioritization: In this step, the priorities are
defined taking into account each previously identified operational scenario
and the meaning in terms of the use of the available capacities for the SoS.
A procedure to support this step of prioritizing operational scenarios can be
applied.
For exemplification purposes, in the next sections, we systematically follow each
one of the steps proposed in OSW method, showing how they were applied to conduct
a real elaboration of a ConOps for the SisGAAz.
5 Case Study: Elaboration of Operational Scenarios

in the SisGAAz Context
Operational scenarios were established to optimize the use of monitoring and protec-
tion resources available in the Brazilian Navy. They were elaborated by interactions
between critical activities and threats in each maritime area of interest. We present
how the OSW method steps were used to support the elaboration of the ConOps of
the SisGAAz, as follows:
Step 1. Operational scenarios identification and prioritization. During the work-
shop sessions in the SisGAAz, the SE team used an approach to identify
the operational scenarios for each naval district (ND) area. The activities
proposed by the approach allowed to delimit the operational scenarios of
each naval command. As a result, the compatible response was properly
delineated in the SisGAAz to meet the operational processes of each SoS
operational area. This approach consisted of the following four steps: (i)
identification of SoS critical activities; (ii) identification of threats and
emergencies; (iii) constituent area identification; (iv) SoS critical activ-
ities confronting; and (v) SoS operational scenario prioritization. The
combination of the previously considered elements was used to form the
scenarios of interest of each ND.
Step 2. Identification of SoS critical activities (CA). This step consisted of enu-
merating the critical activities, i.e., activities of interest in the area of
jurisdiction of the 1st ND that need monitoring and protection: (i) explo-
ration and exploitation of the oil basins (CA1); (ii) exploitation of living
resources (fishing areas) (CA2); (iii) LCM protection and security (CA3);
(iv) preservation of the environment (CA4); (v) safeguarding human life
at sea (CA5); (vi) protection of genetic heritage (CA6); (vii) protection
of scientific knowledge (CA7); (viii) protection of the submerged assets,
including the protection and inspection of the research, removal, demoli-
tion, or exploitation of buried goods belonging to third parties or to the
Table 1 Maritime areas of interest in the 1st ND jurisdiction

Occurrence Physical areas Acronym
Oil basins Basin of Espírito Santo BES
Basin of Campos BCA
Basin of Santos BSA
MCL navigation focal points Surroundings of Cabo Frio SCF
Focal points for access to the Port of Vitória and Tubarão PVT
main ports and maritime Port of Santos PSA
terminals Port of Rio de Janeiro PRJ
Terminal of TEBIG and Ilha TBM
da Madeira (Ilha Grande Bay) TUB
Terminal of Ponta do Ubu TSB
(ES)
Terminal of São Sebastião
(SP)
Reserves or incidences of 1st ND jurisdiction area 1DN
living and non-living
resources (excluding oil and
natural gas)
Maritime environmental 1st ND jurisdiction area 1DN
protection sites
Brazil, in addition to activities in archaeological sites incorporated to Brazil

domain (CA8); (ix) exploration and exploitation of non-living resources
(except oil and natural gas) (CA9); and (x) protection of strategic areas of
interest (CA10).
Step 3. Identification of threats and emergencies. This step consists of enumer-
ating possible threats or possible perceived emergencies. In the 1st ND area
were identified the following threats and emergencies: (i) maritime terror-
ism; (ii) illegal fishing; (iii) transnational illicit acts; (iv) environmental
crimes; (v) national illicit acts; and (vi) SAR emergency.
Step 4. Constituent area identification. This step consisted of identifying the exis-
tence of maritime areas of interest for each SoS constituent, geographically
delimited. In the 1st ND area were identified the following maritime areas:
(i) oil basins; (ii) strategic interests; (iii) MCL navigation focal points; (iv)
focal points for access to the main ports and maritime terminals; (v) reserves
or incidences of living and non-living resources (excluding oil and natural
gas); (vi) maritime environmental protection sites; (vii) archaeological sites
or other areas containing submerged goods; and (viii) objectives or areas
defined by the 1st ND commander. Table 1 organizes the 1st ND maritime
areas with respective physical area and acronym.
Step 5. SoS critical activities confronting. This step consisted of confronting the
critical activities identified for the SoS with the envisaged threats and emer-
gencies, indicating the marine areas of interest evaluated as capable of inter-
acting between critical activities and threats and emergencies. Table 2 shows
Table 2 Intersection of physical areas with threats and critical activities

CA1 CA2 CA3 CA4 CA5 CA6 CA7 CA8 CA9
Maritime BES X PCF 1DN X X X X 1stND
terrorism BCA PVT
BSA PSA
PRJ
TBM
TUB
TSB
the confrontation of critical activities with the threat of maritime terrorism

for 1st ND.
Step 6. SoS operational scenario prioritization. At the end of the aforementioned
steps, we consolidated the threats, emergencies, critical activities, and areas
of interest for the SoS. After the completion of the last step, the scenarios
were prioritized. The need to prioritize the scenarios arises from the fact
that there is hardly any naval, air, and aircraft resource in enough quantity
to simultaneously and uniformly meet all the operational scenarios. There-
fore, prioritization is not only desirable, but also essential if the available
resources are to be effectively used, maximizing the system’s prevention and
response capabilities. These steps could be used as guidelines to be adopted
in other SoS development projects, not only in other military domains (aero-
nautics and war), but also in non-military domains, such as public govern-
ments of smart cities. We then finished that iterations of the SisGAAz and
concluded that the use of OSW method enabled us to obtain a more precise
specification of operations in the resulting ConOps.
6 Analysis of Results and Discussion
The conception of SoS in the defense area is a major technological challenge, not
only because of the complexity of the constituents and equipment that will be part
of it, but also because of the complexity of the environment in which it will operate.
During the systems engineering of the SisGAAz, the identification and prioritization
of the operational scenarios were of extreme importance. The task of SoS conception
that can deal with a diverse reality and its many constituents would not be possible
without first identifying the scenarios in which these constituents would be inserted.
All operational scenarios identified during the OSW were also registered in the
Operational Scenarios Report (OSR). The final version of OSR has approximately
100 pages and considers the scenarios from nine naval districts. During the design
of the SisGAAz, only the operational scenarios of the internal actors of the Brazilian
Navy were considered. These scenarios were analyzed during the elaboration of the
operational concept and helped to identify the necessary capabilities for the SisGAAz.
The operational scenarios served to establish the situations in which the SisGAAz will
act to fulfill the established mission in the operating concept document of the system.
The SisGAAz center of gravity resides in areas of concern where critical activities
occur that must be protected from threats or where there may be emergencies.
During the OSW in the SisGAAz was observed that most of the information
obtained becomes repetitive, with only a few important variations, which make up
the nuances that distinguish one operating command from another. The critical activ-
ities identification evidenced differences between the regions involved, which were
adherent with the economic vocations of each of them. However, despite this differ-
entiation, a high degree of convergence was observed regarding the understanding
of the concepts and definitions used. Regarding threats, it can be observed that most
of them were identified in a common way in all the places visited, demonstrating a
convergence of understandings on this subject. In contrast to the information about
the activities and threats, which showed a convergence between the collected data,
the comparison between the information on the areas of interest obtained from the
various operational commands was more dispersed. This heterogeneity between the
commands can be attributed to the differences between the nature and intensity of
the maritime activities developed in each region. Each operational command, with
its regional characteristics, has a more intense group of activities, concentrated in a
given area, justifying differences in the data collected.
In addition to the scenario definition information, the OSW was also used to obtain
additional information needed by the SisGAAz. This information consisted of four
groups of data: the stakeholders, the SoS requirements, the publications and legal
standards employed, and the SoS needs and characteristics identified by the users.
Finally, it is important to point out that the operational scenarios of a SoS represent
a specific conjuncture of a defined moment, that is, the operational scenarios raised
refer to a given period. If there is a very long time between scenario-gathering and
the deployment of SoS, the reality will have undergone changes that, if relevant,
could invalidate SoS operational concepts, claiming for a new application of OSW
method. Next section brings final remarks.
7 Final Remarks
This paper presented OSW, a method specially tailored to support the ConOps docu-
ments for SoS projects. Among several challenges brought by SoS to the SE projects,
the ConOps is one of them, given the distributed nature of the teams, the conflicting
requirements, the multiple constituents, institutions, and stakeholders involved, and
the operational and managerial independence of the constituents. OSW method was
applied to a real project in the context of the navy of a given country. The contribution
of this paper is twofold: One contribution comprises the OSW method itself, which
can be replicated in any SoS project developed by distributed teams, certainly adopt-
able for defense domain, but potentially extensible for application in SE projects
for SoS of other domains; the another contribution consists of the experience report
presented according to the steps prescribed by OSW method, which provides a proof
of concept for the method (since it has been applied to a real project), and a report
that can be valuable for other teams that also intend to adopt the method.
Future work includes the application of OSW method in other SoS development
projects and its adaptation for reevaluation of SisGAAz. We expect that OSW method
can broaden the state of the art and practice in SoS engineering, consolidating the
SoS body of knowledge, enriching the collection of available techniques to be sys-
tematically applied in the development of SoS, increasing the quality of the delivered
product, and ensuring the correct operation of the system, resulting in a guarantee
of national security and sovereignty as a result of emergent behaviors obtained from
the SoS operation in the defense domain. We also hope these techniques can be
evaluated and incorporated as novel practices and approaches to SoS engineering in
forthcoming SoS development projects around the world.
References
1. ISO/IEC/IEEE Systems and Software Engineering—Architecture Description: ISO/IEC/IEEE

42010:2011(E) (Revision of ISO/IEC 42010:2007 and IEEE Std 1471-2000), pp. 1–46 (2011)
2. Department of Defense (DoD): Systems Engineering Guide for Systems of Systems, V 1.0.
(2008)
3. Walden, D.D., Roedler, G.J., Forsberg, K., Hamelin, R.D., Shortell, T.M.: International Council
on Systems Engineering: Systems Engineering Handbook: A Guide for System Life Cycle
Processes and Activities. Wiley, Hoboken, NJ
4. Paes, C.E.B., Graciano Neto, V.V., Oquendo, F., Nakagawa, E.Y.: Experience report and chal-
lenges for systems-of-systems engineering: a real case in the Brazilian defense domain. In:
Proceedings of Workshop on Distributed Development, Software Ecosystems and Systems of
Systems, pp. 41–50 (2016)
5. de Barros Paes, C.E., Neto, V.V.G., Moreira, T., Nakagawa, E.Y.: Conceptualization of a system-
of-systems in the defense domain: an experience report in the Brazilian Scenario. IEEE Syst.
J., 1–10 (2018)
6. de A. Chaves, S.F.: Sistema de Gerenciamento da Amazônia Azul (SisGAAz): o passo inicial
para o efetivo controle da área marítima brasileira. Monografia apresentada ao Departamento
de Estudos da Escola Superior de Guerra como requisito à obtenção do diploma do Curso de
Altos Estudos de Política e Estratégia (CAEPE), Rio de Janeiro (2013)
7. Silva, E., Cavalcante, E., Batista, T., Oquendo, F., Delicato, F.C., Pires, P.F.: On the character-
ization of missions of systems-of-systems. In: Proceedings of the 2014 European Conference
on Software Architecture Workshops—ECSAW’14, pp. 1–8. ACM Press, New York, NY, USA
(2007)
8. Oquendo, F.: Architecturally describing the emergent behavior of software-intensive system-
of-systems with SosADL. In: 2017 12th System of Systems Engineering Conference (SoSE),
pp. 1–6. IEEE (2017)
9. Maier, M.W.: Architecting principles for systems-of-systems. Syst. Engin. 1, 267–284 (1998)
10. Department of Defense (DoD): Software Development and Documentation—Military Standard
MIL-STD-498 (1994)
11. Hamelin, R.D., Walden, D.C., Krueger, M.E.: INCOSE Systems Engineering Handbook v3. 2:
Improving the Process for SE Practitioners. INCOSE International 11 (2010)
12. ISO:IEC 15288: 2008 Systems and software engineering-System life cycle processes. In:
2nd International Organization for Standardization/International Electrotechnical Commission,
18th March. (2008)
13. Dahmann, J., Rebovich, G., Lane, J.A., Lowry, R.: System engineering artifacts for SoS. In:
2010 IEEE International Systems Conference, pp. 13–17. IEEE (2010)
14. Dahmann, J.S., Baldwin, K.J.: Understanding the current state of US defense systems of systems
and the implications for systems engineering. In: 2008 2nd Annual IEEE Systems Conference,
pp. 1–7. IEEE (2008)
15. IEEE Computer Society. Software Engineering Standards Committee., Institute of Electrical
and Electronics Engineers., IEEE-SA Standards Board.: IEEE guide for information technolo-
gy– : system definition– concept of operations (ConOps) document. Institute of Electrical and
Electronics Engineers (1998)
Science Diplomacy: Strategic Initiative
to Create a Buffer Zone in the Caribbean
Colombian Marine Protected Area
Seaflower
Fabián Ramírez-Cabrales and Sergio Iván Rueda Forero
Abstract It recognized the science diplomacy as a path to address national and

common interest related to marine spaces declared as human heritage areas. Science
diplomacy is an instrument that allows scientific cooperation between states and inter-
national organizations, and stakeholders interested in to obtain scientific evidence to
the decision-making process at the political level in all senses. The Specially Pro-
tected Areas And Wildlife to the Convention for the Protection and Development
of the Marine Environment of the Wider Caribbean Region (SPAW) Protocol is the
main instrument to evaluate the possible course of action that allows us to propose
the implementation of this initiative. To do so, it is necessary to consider aspects as
sovereignty and maritime jurisdiction of state parties and non-state parties. Besides,
are considered threats related to maritime transport. It analyzed the geographical
conditions and the managerial institutions and stakeholders that participate in the
management. Also, the uses of this strategic marine space within the Colombian
Caribbean Sea are reviewed. The aim is to address the possibility to create a buffer
zone in the boundary of the Marine Protected Area (MPA) Seaflower considering
the borders of the Seaflower biosphere reserve.
Keywords Science diplomacy · Buffer zone · Caribbean · Colombian · Marine

protected area · Seaflower
1 Introduction
Science diplomacy (hereafter SD) is an international, interdisciplinary, and inclusive

process that allows us to understand and give a better solution, both to global problems
and to national strategic interests. It also incorporates the international dimension
F. Ramírez-Cabrales (B) · S. I. R. Forero

University Escuela Naval Almirante Padilla, Cartagena D.T.y C, Colombia
e-mail: rc.fabian@gmail.com
S. I. R. Forero
e-mail: comancheof@gmail.com

https://doi.org/10.1007/978-981-13-9155-2_28
346 F. Ramírez-Cabrales and S. I. R. Forero
of science, technology, and innovation policies in the foreign policy of the states.
It links the participation of scientific advisors and considers the value of scientific
evidence in international relations for informed decision-making by the state.
This study is an exploratory investigation originated from the analytical review
of the available literature on Marine Protected Areas (MPAs), and a proposal for
some governance and spatial management mechanisms in the Colombian Caribbean
Sea (2016), particularly in the Marine Protected Area (hereafter MPA) Seaflower.
Currently, this proposal is linked to the scientific diplomacy research project for the
protection of the marine environment as a contribution to the development strategy
for the archipelago of San Andrés and Providencia led by the Republic of Colombia
Navy. The purpose is to develop a strategic initiative to create a buffer zone in the
boundary of the MPA Seaflower considering the borders of the Seaflower biosphere
reserve.
The Seaflower MPA is recognized as possessing one of the highest levels of
marine biodiversity in the Caribbean region. Actually, over 72% of all-regional Cen-
tral and South American transshipments move through the southwestern Caribbean
area—wherein the Seaflower is located, a high amount of cargo, being containers
and petroleum products, the principal commodities moved through the area. These
factors are threatening the surrounding ecosystems by ship-source pollution. In con-
sequence, the Seaflower MPA requires a buffer zone to protect one of the highest
levels of marine biodiversity in the Caribbean region.
This study proposes as a research question, how science diplomacy can contribute
to create a buffer zone in the Caribbean Colombian MPA Seaflower? To answer this
question, the article is structured into five parts. The first part gives a brief introduc-
tion; the second part defines the science diplomacy and identifies its dimensions; the
third part describes international and regional scenarios that promote the sustainable
development and care of the MPAs within the framework of the United Nations Sus-
tainable Development Programme (UNEP) and the regional programs implemented
in the Wider Caribbean. As a fourth part, we describe the geographical aspects of
the MPA Seaflower highlighting the importance that this MPA represents the Wider
Caribbean Region and the threats that loom over it. The fifth part describes the three
dimensions of the SC and its possible contribution to the strategic initiative to create
a buffer zone in the Caribbean Colombian Marine Protected Area Seaflower.
2 Science Diplomacy
2.1 What Is Science Diplomacy?
According to Langenhove [1], diplomacy and science represent, in principle, two

fields of human activity that have little in common. However, scientific networks and
their constant activity play an increasingly determining role in diplomatic efforts
Science Diplomacy: Strategic Initiative to Create a Buffer … 347
initiated by state and non-state actors. The result of this conjugation of efforts results
in a practice that can be called “science diplomacy” (SD) [1].
The SD is a fundamental tool for international relations; it promotes science, tech-
nology, and innovative knowledge, which includes the deployment of scientific and
career diplomatic personnel, and the insertion of these topics in the development of
the foreign policy and the new global agenda. The SC demands new contributions
and primary expertise in international relations associated with science and scien-
tific and technological development. This discipline is a response to the new global
challenges imposed by the effects of climate change, energy efficiency, biodiversity
loss, food security, the impacts of pollution, and the conservation of water resources,
among others [2].
Considering this scenario, the exercise of the foreign policy requires tools, tech-
niques, and tactics adaptable to a world that is increasingly complex in technical and
scientific terms. This has created a new role for science in the field of policy formu-
lation and diplomacy. But nevertheless, how science can contribute to foreign policy
objectives? This question had a wide debate during a two-day meeting on “New
Frontiers in Science Diplomacy,” which was hosted by the Royal Society from June
1 to 2, 2009, in partnership with the American Association for the Advancement of
Science (AAAS). The result identified three dimensions in which science diplomacy
can contribute to policy objectives as well [3]:
• Science in diplomacy: Science providing advice to inform and support foreign
policy objectives.
• Diplomacy for science: Diplomacy facilitating international scientific cooperation.
• Science for diplomacy: Scientific cooperation improving international relations.
These three dimensions can contribute to the strategic initiative to create a buffer
zone for the MPA Seaflower. The structure and programs to implement these actions
are based on the following regional structure.
3 United Nations Environment Programme (UNEP)
At the regional level, the UN system recognizes the Caribbean Sea as one of the
most unique and complex areas of the world. Its extension is over 2.5 million km2
[4]. It is geographically and politically diverse as it embraces 44 states and territories
with over 100 maritime boundaries [5–7]. For that reason, to manage this vast area,
several initiatives have been developed. There are only two biggest environmental
protection programs concerning the entire area that today will be discussed. The first
one is the Caribbean Environment Programme (CEP), and the other is Caribbean
Large Marine Environment (CLME) project. For the purposes of this initiative, the
following section will focus on the CEP.
3.1 United Nations Environment Programme-Caribbean

Environment Programme—(UNEP-CEP)
The CEP was the second UNEP established in 1982 with all its institutional com-
ponents (convention, action plan, funds, and coordination unit), building one of the
most valuable programs of the UN system. In addition to this, the CEP is recog-
nized by the countries in the region for its high commitment to the development of
environmental management initiatives [8].
The UNEP’s main goal is to reduce and tackle the degradation of environments
by marine pollution [8]. In that sense, the Caribbean is considered the second most
complex marine route worldwide. This is a consequence of the traffic through the
Panama Canal, and the increase of seaborne trade in many nearby ports in the area
due to its economic attractiveness.
Ship-source pollution is the factor that is threatening the surrounding ecosystems,
including the issue of invasive species [9]. Likewise, tourism and fisheries are the
main activities in the zone and are also sources of marine pollution. Furthermore, the
Caribbean faces several political conflicts and presents diverse economies in terms
of development, with significant influence of developed countries exerting colonial
powers and hosting many different languages [8–10].
To address all the above issues and challenges, in 1982, countries unanimously
decided to work for the adoption of a cooperative and integrated approach through a
convention for the protection and sustainable development of the Caribbean region
[6].
3.2 Convention for the Protection and Development

of the Marine Environment in the WCR and Its Protocols
The Wider Caribbean Action Plan was the first step taken by CEP and countries to
start developing a binding agreement. As a result and through the commitment of the
countries involved, it was possible to move forward to the adoption of the convention
for the Wider Caribbean, which was agreed in Cartagena de Indias, Colombia, in 1983
(hereafter the Cartagena Convention). This regional treaty entered into force in 1986.
To date, 25 states have ratified it [6, 11].
The Cartagena Convention is the current binding regional environmental agree-
ment serving as an essential reference for both managers and policy-makers when
protection measures and management development of coastal and marine resources
are in place, either individually or jointly. Nevertheless, due to developing economies
surrounding the area, many issues are facing the Caribbean region. For that reason,
one of the main targets is the establishment of MPAs as a mechanism to ensure the
conservation of biodiversity, pollution reduction, education, and capacity building.
Following the adoption of the convention in 1983, two legal instruments were
subsequently adopted. The first one was the Protocol Concerning Cooperation in
Combating Oil Spills in the Wider Caribbean Region (Oil Spills Protocol). This
protocol and the convention itself were convened simultaneously. The second is
the Protocol Concerning Specially Protected Areas and Wildlife (SPAW Protocol),
which was adopted in the early 1990s and entered into force in 2000. The latter was
the third program driven by UNEP with the aim of preserving and improving the
marine ecosystem in the region [5, 6, 11].
Currently, the SPAW Protocol has been ratified by only 16 states, including Colom-
bia. The aim is “to take necessary measures to protect, preserve and manage, in a
sustainable way, zones that require protection to safeguard their particular value and
that threaten or endanger species of flora and fauna” (Art. 3) [6, pág. 40].
To date, nine coastal states in the Wider Caribbean encompass over 32 MPAs
under the SPAW Protocol and the CEP. The marine area covered is approximately
100,000 km2 corresponding to only 4% of the all-regional sea (CAR-SPAW-RAC
2018).
Although the number of MPAs is relatively small, in 2014, the Caribbean Regional
Programme (CRP), under the umbrella of the SPAW Protocol, established the
Caribbean Challenge Project [5]. The objective is to enlarge regional MPAs up to
20% by 2020 through a sustainable governance mechanism based on cooperation
between countries for the protection of the ecosystem in the larger Caribbean [7].
Colombia is the depository country for the convention and one of the first to ratify
it. Therefore, it is making efforts to fulfill the agreement. Currently, around nine
percent of its maritime jurisdiction is protected through MPAs [12]; one of 30 the
largest in the region is the Seaflower MPA [5, 13].
4 The Seaflower MPA
4.1 Background
Colombia’s constitutional mandate considers the preservation and conservation of

its unique ecosystems and areas of ecological importance, and is respectful of the
obligations prescribed in the international environmental treaties to which it is party
[14]. The government, in 1998, proposed the designation of the archipelago of San
Andres and Providencia and Santa Catalina, and the surrounded marine area, as a
biosphere reserve (Fig. 2) to be known as the Seaflower Biosphere Reserve [15].
Later in 2000, UNESCO included it within the MAB World Biosphere Network [13]
(Fig. 1).
Moreover, in January 2005, the Ministry of Environment and Sustainable Devel-
opment (MADS, Spanish acronym) through Resolution 107 declared an MPA, within
the biosphere reserve [15], the largest in the Caribbean Sea, and the first of its category
in Colombia [13, 16, 17, 18].
Fig. 1 Seaflower biosphere reserve
4.2 Geographical Position
The Seaflower MPA is located in the southwestern Caribbean region, surrounding

the San Andres Archipelago (i.e., three small inhabited oceanic islands and eight
unsettled cays and atolls) (Fig. 2). The largest island, San Andres, is 800 km (480
NM) northwest of the Colombian continental territory [17, 19]. The MPA covers a
maritime area of 65,000 km2 , with only 1% of the terrestrial surface (650 km2 ) [17].
The population of over 100,000 base their livelihood on artisanal fisheries and
tourism. The MPA is part of the Caribbean coral reef hotspot, which is among the
richest areas in marine species diversity, but also one of the most threatened [13]. For
instance, the area is considered to be of regional and national ecological significance.
The Seaflower MPA is recognized as possessing one of the highest levels of marine
biodiversity in the Caribbean region [19]. It representatively defines Colombia’s six
tropical marine ecosystems: coral reefs, seagrass beds, mangroves, rocky littorals,
sandy beaches, and soft bottoms, which influence the high productivity of the area
[13].
The coral reef is the key ecosystem for conservation in the MPA because as one of
the largest in the Caribbean, it represents about 14% of the world’s coral reefs [20],
the second in the western hemisphere, and most productive in the region [17]. The
reef extension covers approximately 2000 km2 [18] and is composed of two barrier
reefs, five atolls, reef lagoons, and coral banks [13], which is 78% of the national
total [16].
Fig. 2 Seaflower marine protected area
Colombia set, through the National Law 99/93 [21], overall targets for the MPAs
in the territory, based on the global IUCN objectives [11, 22, 23].
Moreover, the Seaflower MPA has defined its objectives with the participation of
islander communities and stakeholders based on the concept of community-based
management ([13]; De Pourcq et al. 2015) to ensure effective governance and protec-
tion of the ecosystems in the zone. Moreover, to succeed in achieving the activities
within the MPA, five core objectives were defined [24]:
1. Preservation, recovery, and long-term maintenance of species, biodiversity, ecosystems,
and other natural values including special habitats; 2. Promotion of sound management
practices to ensure long-term sustainable use of coastal and marine resources; 3. Equitable
distribution of economic and social benefits to enhance local development; 4. Protection of
rights concerning historical use; and 5. Education to promote stewardship and community
involvement in planning, and management.
Based on these, and using the EBM approach [7, 25], both the community and
authorities designed the zones within the MPA to ensure the protection of ecologi-
cally relevant areas. The zoning criteria took into consideration the ecological criteria
(Roberts et al. 2003), seeking an easy demarcation based on representativeness and
connectivity of key habitats (i.e., coral reefs, seagrass beds, algal beds, and man-
groves).
4.2.1 MPA Zoning
The MPA is divided into three administrative sections, established through Accords
021 and 025 [24], seeking a better implementation of the biosphere reserve. There-
fore, the areas are: northern (37,522 km2 ), central (12,716 km2 ), and southern
(14,780 km2 ) (Figs. 3 and 4).
Moreover, within the administrative divisions, there are five zone types for in situ
conservation and sustainable use [13, 18]:
1. No-entry (116 km2 ), preservation zones restricted only for research and monitoring activi-
ties; 2. No-take (2,214 km2 ), conservation zones that incorporate and allowing non extractive
uses; 3. Recovering and sustainable use of marine resources (2,015 km2 ), allows traditional
artisanal fishing activities, and artisanal sport fishing and all the activities allowed in the
special use, no-entry and no-take zones; 4. Special use (68 km2 ), shipping related, leisure,
and waters sports; 5. General use (60,587 km2 ), minimal restrictions apply, seeking and
maintaining the MPA objectives to promote marine conservation.
4.2.2 Threats to the MPA in the Region
Vallega [26], UNEP [7, 20], Lopez and Krauss [27], Fanning et al. [25], Biggs
[28], and Morris [29] have described the threats and issues concerning the marine
ecosystems in the Caribbean region, and thus, they are not different for this MPA.
Moreover, today, new challenges concerning shipping activities have emerged as
Fig. 3 Seaflower marine protected area, administrative division

Fig. 4 Caribbean shipping density
a consequence of ongoing and short-term developments in the region. About 90%

of global trade is moved by seagoing vessels [30], and keeping pace with global
economic development, shipping has grown by 30% over a period of four to five
years [31], increasing routes and connectivity. For instance, according to UNCTAD
[32], Latin America and the Caribbean have the highest liner shipping connectivity
with Panama. At least, 21 vessel operators connect with the USA directly through the
Panama Canal, which is the crossroad between east–west and north–south routes,
followed by Mexico, Jamaica, and Colombia. Thus, the Panama Canal expansion
allows regional expansion of ports, and therefore, larger vessels transit through the
Caribbean (Fig. 4).
The opening, on June 26, 2016, of the expanded new Panama Canal [33] has
enabled the transit of approximately 4750 additional ships per year, handling over five
percent of global goods and about eight percent of all transshipments worldwide [34].
Furthermore, over 72% of all-regional Central and South American transshipments
move through the southwestern Caribbean area—wherein the Seaflower is located,
a high amount of cargo, being containers and petroleum products, the principal
commodities moved through the area [31]. Approximately, 60 voyages per week
(14,000 annually) to over 100 maritime routes pass through the Caribbean Sea [35,
36].
Thus, the risk of an incident occurring during transit through the routes in the
Caribbean is much greater. Further, it will cause an increase of ship-source pollution
such as CO2 and greenhouse gas (GHG) emissions [32], which contribute to ocean
acidification [37, 38], affecting the coral reefs in the Wider Caribbean Area, and
those within the Seaflower MPA.
5 Strategic Initiative
5.1 Science for Diplomacy
The SPAW Protocol establishes in its Article 8 that state parties may, as necessary,
“strengthen the protection of a protected area by establishing, within areas in which it
exercises sovereignty, or sovereign rights or jurisdiction, one or more buffer zones.”
For this, it sets the following guidelines:
1. If a Party intends to establish a protected area or a buffer zone contiguous to the frontier
or to the limits of the zone of national jurisdiction of another Party, the two Parties shall
consult each other with a view to reaching agreement on the measures to be taken and
shall, inter alia, examine the possibility of the establishment by the other Party of a
corresponding contiguous protected area or buffer zone or the adoption by it of any other
appropriate measures including co-operative management programmes.
2. If a Party intends to establish a protected area or a buffer zone contiguous to the frontier
or to the limits of the zone of national jurisdiction of a State that is not a Party to this
Protocol, the Party shall endeavour to work together with the competent authorities of
that State with a view to holding the consultations referred to in paragraph 1.
In this regard, countries such as Costa Rica, Jamaica, Haiti, Honduras, and
Nicaragua share the border with the Republic of Colombia, but none of them is a state
party to the SPAW Protocol. Besides, in late 2001, Nicaragua instituted proceedings
against Colombia at the International Court of Justice (ICJ), claiming rights to the
islands and maritime waters [39]. Also, Nicaragua requested to the Seabed Authority
an extension of their continental shelf with the pretension of granted in concession the
area claimed for exploration and exploitation of oil and gas, jeopardizing the marine
biodiversity, in particular, the coral reef areas. The ICJ ruled, based on customary
law, that Colombia has the sovereignty over the islands in question. However, the
ICJ decision is a disaster for the pristine marine ecosystems, dividing the integrity of
the Seaflower MPA and the biosphere reserve declared by UNESCO in early 2000
[13] and backed by many international environmental organizations such as UNEP
and GEF [40]. In consequence, political constraints and economic interest are affect-
ing and jeopardizing the protection of the marine ecosystems within the Seaflower
Biosphere Reserve and thus the MPA.
In that sense, to establish a buffer zone contiguous to the frontier of the zone
of national jurisdiction of a state that is not a party, it will be necessary to apply
science for diplomacy. This dimension can be driven by the scientific community.
The scientific community focuses on common problems beyond national borders.
Likewise, this facet is flexible and promotes the emergence of new forms of diplomacy
that require non-traditional alliances between states, sectors, and non-governmental
organizations. “These scientific exchange channels can contribute to the formation of

coalitions and the resolution of conflicts as long as they are aligned with the broader
objectives of foreign policy” [3]. Likewise, these actions can be beneficial to build a
better relationship with Costa Rica, Jamaica, Haiti, Honduras, and Panama, the latter
state party of the SPAW Protocol and pioneer in the practice of SD in Latin America.
These actions could contribute to the protection of a common interest such as the
Biosphere Reserve Seaflower.
5.2 Diplomacy for Science
The SPAW Protocol establishes a Scientific and Technical Advisory Committee

(STAC) to assist with the implementation of the protocol. The STAC is a Con-
sultative Body, Technical Adviser to the Meeting of the Parties, subordinate to the
Contracting Parties, which acts by express mandate of the Meeting of the Parties, and
whose function will be to advise the Meeting of the Parties on matters of a technical
nature.
Within its structure, it includes four committees: a. Subcommittee on Protected
Areas; b. Subcommittee on Species Protection; c. Education and Information for the
Public; and d. Legal and Regulatory Matters. The STAC promotes national projects
for conservation and identifies ways to help national governments to incorporate bio-
diversity conservation into their resource management activities. It also encourages
the coordination of regional activities related to protected areas.
Particularly, the role of the Secretariat for the development of projects for the
regional implementation as support for the objectives of the protocol is highlighted.
The STAC can develop cooperation with other international, regional, and national
organizations, under the direction of the parties, in order to generate experiences and
knowledge to promote partnerships for cooperation and collaboration, as appropriate.
In that sense, “multi-country and regional mechanisms can be effective for building
relationships between the scientific and the policy communities, as well as between
countries” [41]. Under this framework, the STAC is considered a strategic regional
actor to implement the strategic initiative for the creation of a buffer zone through
the diplomacy for science.
5.2.1 Science in Diplomacy
Currently, the Republic of Colombia has been planning and executing scientific
research cruises in the MPA Seaflower for almost four decades since before its dec-
laration as a Biosphere Reserve (BR). These actions are part of the integral strategy
for the exercise of sovereignty in the archipelago of San Andrés and Providencia and
Santa Catalina, led from the presidency of the Republic of Colombia to strengthen the
management and conservation of the Seaflower Biosphere Reserve through knowl-
edge management and scientific evidence (contributions to Seaflower knowledge).
During the last four years, Colombia has made four scientific expeditions in the
BR Seaflower: Cayos de Roncador Island 2015; Cayos de Serrana Island 2016; Cayos
de Serranilla Island 2017; and Cayo Alburquerque Island. The objective is to specify
a 2015–2023 monitoring plan and standardize sampling protocols. In addition, the
objective of scientific expeditions to the BR Seaflower is to demonstrate on a scientific
basis the importance of the Seaflower Biosphere Reserve for Colombia and as a
common heritage of humanity.
This national initiative has had an impact on the epistemic community, however,
and taking into account [42] it is necessary that the foreign policy of the Colom-
bian state “…should place greater emphasis on science within their strategies, and
draw more extensively on scientific advice in the formation and delivery of policy
objectives.” Particularly issues related with the RB Sea-flower.
6 Conclusions
The MPA Seaflower is facing several threats mostly due to shipping activities as
well as political conflicts. The zoning of this large area is not enough to protect these
pristine ecosystems located there. Due to its geographical location in the southwestern
Caribbean region, surrounding the San Andres Archipelago and within the borders of
the BR Seaflower, the outer limit of this MPA links maritime borders with countries
such as Honduras, Nicaragua, Costa Rica, and Jamaica. Nevertheless, none of these
countries is a state party of the SPAW Protocol. Thus, using the SPAW Protocol as a
tool to engage these countries to protect and to preserve the marine ecosystems within
the MPA it allows the interested party to initiate a broad process of consultations with
the authorities of the adjacent countries to the maritime borders to continue protecting
this large area considered as a human heritage site.
To this end, Colombia can use and promote the three dimensions of science
diplomacy. Each of these dimensions has to be institutionalized as a guideline for the
foreign policy. This action will allow a stronger regional projection supported by the
instruments and mechanisms of the SPAW Protocol. One of the main targets is the
Scientific and Technical Advisory Committee (STAC). This is a key actor to assist
the implementation of the SPAW Protocol. Through STAC, Colombia can influence
and provide advice to the parties on the scientific and technical aspects of MPA
Seaflower. Also, Colombia can promote this kind of initiatives for the conservation
of marine biodiversity and lead the coordination at the regional level.
To achieve this goal, Colombia must establish a multidisciplinary national sci-
entific committee to integrate each of the STAC Subcommittees, particularly the
Subcommittee on Regulatory and Legal Matters for the development of laws and/or
model guides for the implementation of SPAW obligations.
In terms of strategic alliance, Panama can play an important role given its status
as a state party because its political and economic influence on shipping represents
an important neighborhood to support this initiative. Particularly, because Panama
is one of the pioneer countries in Latin America that promotes Science Diplomacy
as a tool for the development of its foreign policy.
It is concluded that Colombian foreign policy requires more science to integrate
the value of scientific evidence in international relations that will allow to implement
this strategic initiative through SD.
References
1. Langenhove, L.V.: Global science diplomacy for multilaterlism 2.0. Sci. Diplomacy (2016)
2. Gorgone, H.R.: La Diplomacia Científico Tecnológica Como Herramienta De Internacional-
ización. Retrieved June 2017, from Repósitorio Institucional UFSC (2013, November 29):
https://repositorio.ufsc.br/handle/123456789/116162
3. The Royal Society.: New Frontiers in Science Diplomacy. Retrieved June 2018, from royalso-
ciety.org (2010): https://royalsociety.org/topics-policy/publications/
4. “The Caribbean Sea”: The Columbia Encyclopedia, 6th edn. Retrieved 11 Feb 2016, from
Encyclopedia.com: (2015) http://www.encyclopedia.com/doc/1E1-Caribbea.html
5. CEP.: The Caribbean Environment Programme—about Cartagena Convention. Retrieved 11
Feb 2016, from The Caribbean Environment Programme (2015): http://www.cep.unep.org/
6. UNEP: Convention for the Protection and Development of the Marine Environment of the
Wider Caribbean Region and its Protocols, 2nd edn. UNEP, Kingston, Jamaica (2012)
7. UNEP.: Setting a Course for Regional Seas. Retrieved 13 Mar 2016, from UNEP—Re-
gional Seas Programme (2014a): http://apps.unep.org/publications/index.php?option=com_
pub&task=download&file=011736_en
8. UNEP.: Measuring Success: Indicators for the Regional Seas Convention and Action Plans.
Retrieved 13 Mar 2016, from UNEP (2014b): http://apps.unep.org/publications/index.php?
option=com_pub&task=download&file=011400_en
9. Patrick, S., Storm, B.: Council on foreign relations. In: Brown, K., Thaler, F.F.
(eds.) Producers, & Media Strom. Retrieved 23 Jan 2016, from Global Governance
Monitor/oceans: http://www.cfr.org/global-governance/global-governance-monitor/p18985#!/
oceans?gclid=CN_QxJjLwMoCFSTecgodjDoHAw#Issue-Brief
10. Van Tatenhove, J.P.: How to turn the tide: developing legitimate marine governance arrange-
ments at the level of the regional seas. Ocean and Coastal Manage. J 71, 296–304.
Retrieved 23 Jan 2016, from (2013, January) http://www.sciencedirect.com/science/article/
pii/S0964569112003171: http://doi.org/10.1016/j.ocecoaman.2012.11.004
11. Lausche, B.: Guidelines for Protected Areas Legislation, vol. 26. IUCN, Gland, Switzerland
(2011)
12. PNNC-RUNAP.: Unique National Registry of Protected Areas—RUNAP. Retrieved 05 June
2016, from National Natural Parks of Colombia (2016): http://www.parquesnacionales.gov.
co/portal/es/sistema-nacional-de-areas-protegidas-sinap/registro-unico-nacional-de-areas-
protegias/
13. Howard, M.W.: Evaluation Report Seaflower Biosphere Reserve Implementation: The First
Five Years 2000–2005. Retrieved 2 Nov 2015, from UNESCO: http://www.unesco.org/csi/
smis/siv/Caribbean/San_actEnvEd_Seaflower2000-2005%20.pdf
14. Minambiente.: National Policy for the Integral Management of Biodiversity and its Ecosystemic
services (NPIMBES). (J. Weiskopf, Trans.) Bogotá, Colombia. Retrieved 10 Apr 2016, from
cbd.int (2011): https://www.cbd.int/doc/world/co/co-nbsap-v2-en.pdf
15. Minambiente.: Resolución número 107 (27 de Enero de 2005) Por el cual se declara un área
marina protegida y se dictan otras disposiciones. Retrieved 22 Nov 2015, from Ministerio
de Medio Ambiente y Desarrollo Sostenible (2005, January 27): http://www.coralina.gov.
co/coralina/rendiciondecuentas/monitoreo-plan-accion/normatividad-1/normatividad-por-
temas/areas-protegidas/238-resolucion-107-de-2005-por-la-cual-se-declara-un-area-marina-
protegida
16. Coralina-Invemar.: Atlas de la Reserva de Biósfera Seaflower. Archipiélago de San Andrés,
Providencia y Santa Catalina, vol. 28. In: Gómez-López, D., Segura-Quintero, C., Sierra-
Correa P., Garay-Tinoco, J. (eds.) Santa Marta, Colombia: Instituto de Investigaciones Marinas
y Costeras “José Benito Vives De Andréis”- INVEMAR- y Corporación para el Desarrollo
Sostenible del Archipiélago de San Andrés, Providencia y Santa Catalina –CORALINA (2012)
17. UNEP.: Seaflower Marine Protected Area—Colombia. Proposed areas for inclusion in the
SPAW list. Retrieved 11 Feb 2016, from Unite Nations Environment Program—Caribbean
Environment Program (2010, October 5): http://www.spaw-palisting.org/uploads/reports/
Report_31.pdf
18. Taylor, E., Baine, M., Killmer, A., Howard, M.: Seaflower marine protected area: governance
for sustainable development. Marine Policy 41, 57–64 (2013, January 23)
19. Murillo, I., Ortiz, R.: Analisis de los metadatos sobre proyectos de investigación científica
marina desarrollada en el area de la reserva de biosfera Seaflower. Secretaria Ejecutiva de la
Comision Colombiana del Océano- Area Asuntos Marinos y Costeros, Bogotá D.C (2013)
20. UNEP.: Wider Caribbean Region Profile. Retrieved 11 Feb 2016, from UNEP (2005): http://
www.unep.org/regionalseas/programmes/unpro/caribbean/instruments/r_profile_car.pdf
21. Alonso, D., Barbosa, H., Duque, M., Gil, I., Morales, M., Navarrete, S., … Vasquez, J.: Con-
ceptualización del Subsistema de Áreas Marinas Protegidas en Colombia. Documento de Tra-
bajo (Versión 1.0). Proyecto COL75241 Diseño e implementación de un Subsistema Nacional
de Áreas Marinas Protegidas (SAMP) en Colombia. Retrieved 06 June 2016, from INVE-
MAR (2015): http://www.invemar.org.co/documents/10182/14479/cartilla-conceptualizacion-
final-web.pdf
22. Jones, P.J.: Governing Marine Protected Areas: Resilience through Diversity. Taylor and Fran-
cis, New York (2014)
23. Speed, J.R., Levine, A.: What makes a “successful” marine protected area? The unique context
of Hawaii’s fish replenishment areas (Elsevier, Ed.) Mar. Policy 44, 196–203 (2014, September
12)
24. Coralina: Acuerdos 021 y 025 de 2005 Por medio del cual se delimita internamente el Área
Marina Protegida de la Reserva de la Biosfera Seaflower y se dictan otras disposiciones.
Retrieved 18 Feb 2016, from Registro Unico Nacional de Areas Protegidas (2005): http://
runap.parquesnacionales.gov.co/reportes/generardocadminzip/aid/838
25. Fanning, L., Mahon, R., McConney, P., Angulo, J., Burrows, F., Chakalall, B., … Toro, C.: A
large marine ecosystem governance framework. Marine Policy, 31(34), 434–443. Retrieved 19
Mar 2016, from Marine Policy (2007, July): http://www.sciencedirect.com/science/article/pii/
S0308597X07000097
26. Vallega, A.: The regional approach to the ocean, the ocean regions, and ocean regionalisation—a
post-modern dilemma. Ocean Coast. Manag. 45, 721–760 (2002)
27. Lopez, V., Krauss, U.: National and Regional Capacities and Experiences on Marine Inva-
sive Species, Including Ballast Waters, Management Programmes in the Wider Caribbean
Region—A Compilation of Current Information. United Nations Environment Programme.
CAB International Caribbean and Latin America Regional Centre (CLARC), Trinidad &
Tobago (2006)
28. Biggs, C.: Indo-Pacific Red Lionfish: Pterois Volitans, Invasion of the Western
Atlantic and Caribbean Oceans. Retrieved 23 Aug 2015, from http://www.washington.
edu (2009): http://depts.washington.edu/oldenlab/wordpress/wp-content/uploads/2013/03/
Pterois-Volitans_Biggs.pdf
29. Morris, J.J.: Invasive Lionfish: A Guide to Control and Management, vol. 1. Gulf and Caribbean
Fisheries Institute Special Publication. Marathon, Florida, USA (2012)
30. IMO: International shipping facts and figures—information resources on trade, safety, secu-
rity, environment. Retrieved 10 July 2016, from IMO-Knowledge Centre section: Ships and
Shipping—Facts and Figures (2012 March 5): http://www.imo.org/en/KnowledgeCentre/
ShipsAndShippingFactsAndFigures/TheRoleandImportanceofInternationalShipping/
Documents/International%20Shipping%20-%20Facts%20and%20Figures.pdf
31. AGCS: Panama Canal 100 Shipping safety and future risk. Retrieved 17 Jan 2016, from
Allianz Global Corporate & Specialty (2014): http://www.agcs.allianz.com/assets/PDFs/risk%
20bulletins/Panama-Canal-100_low-res.pdf
32. UNCTAD.: Review of Maritime Transport 2015. Retrieved 10 July 2016, from UNCTAD/RMT
(2015): http://unctad.org/en/PublicationsLibrary/rmt2015_en.pdf
33. The World Bank.: Panama Canal expansion: A smart route for boosting infras-
tructure in Latin America. Retrieved 30 July 2016, from The World Bank(2016,
May 07): http://blogs.worldbank.org/voices/panama-canal-expansion-smart-route-boosting-
infrastructure-latin-america
34. Rodrigue, J.-P., Ashar, A.: Transshipment hubs in the New Panamax Era: The role of the
Caribbean. J. Transp. Geogr. 10 (2015 October 2)
35. US Department of Transportation.: Panama Canal Expansion Study_Phase I Report.
Retrieved 17 Jan 2015, from U.S. DEPARTMENT OF TRANSPORTATION MARITIME
ADMINISTRATION (2013, November 20): http://www.marad.dot.gov/wp-content/uploads/
pdf/Panama_Canal_Phase_I_Report_-_20Nov2013.pdf
36. Webster, M.: Redrawing Global Shipping Routes: The Panama Canal Gets an Upgrade.
Retrieved 17 Jan 2016, from Brown Brothers Harriman (2015, December): https://www.
bbh.com/blob/10958/f7d1d1c2fd63c146c469fb0d384b1c75/redrawing-global-shipping-
routes–the-panama-canal-gets-an-upgrade-pdf-data.pdf
37. Harrould-Kolieb, E.R., Herr, D.: Ocean acidification and climate change: synergies and chal-
lenges of addressing both under the UNFCCC. Clim. Policy J. Retrieved 22 Oct 2015 (2012,
January): http://doi.org/10.1080/14693062.2012.620788
38. Hassellöv, I.-M., Turner, D.R., Lauer, A., Corbett, J.J.: Shipping contributes to ocean acidifi-
cation. Geophys. Res. Lett. Retrieved 22 Nov 2015, (2013, June 6): http://onlinelibrary.wiley.
com/doi/10.1002/grl.50521/epdf
39. Bekker, P.: The World Court Awards Sovereignty Over Several Islands in the Caribbean
Sea to Colombia and Fixes a Single Maritime Boundary between Colombia and Nicaragua.
Retrieved from The American Society of International Law (2013, Jan 15): http://www.
peacepalacelibrary.nl/ebooks/files/354062581.pdf
40. Gorricho, J.: Global Pacebuilding II: High Time for Environmental Diplomacy in the Caribbean.
Retrieved 7 July 2016, from Governance and Development (2012, November 28): http://www.
governanceanddevelopment.com/2012/11/global-peacebuilding-ii-high-time-for.html
41. Gual, S.M., Wang, T.C., Robinson, C.R.: Science Diplomacy. Retrieved June 2018, from science
diplomacy.org (2017, March): http://www.sciencediplomacy.org/article/2017/international-
collaboration-in-connecting-scientists-policy
42. Gluckman, P.D., Turekian, V., Grimes, R.W., Kishi, T.: Science Diplomacy: A Pragmatic Per-
spective from the Inside. Retrieved 20 June 2018, from Science & Diplomacy (2017, Decem-
ber): www.sciencediplomacy.org/article/2018/pragmatic-perspective
Part X
Wearable Technology and Assistance
Devices
Quantum Well Infrared Photodetector
for the SWIR Range
Pedro Pereira, Lesslie Guerra, G. M. Penello, M. P. Pires, L. D. Pinto,

R. Jakomin, R. T. Mourão, M. H. Degani, M. Z. Maialle and P. L. Souza
Abstract An InGaAs/InAlAs superlattice infrared photodetector is developed to

reach the forbidden gap in the SWIR range for arsenides, between 1.7 and 2.5 µm,
appropriate for surveillance imaging. The figures of merit of the device are deter-
mined to be 120 K for the BLIP temperature and 2.1 mA/W and 3 × 106 Jones for the
best responsivity and detectivity, respectively, obtained at 120 K under +4 V bias.
Possible approaches to improve the device performance are addressed.
Keywords QWIPs · SWIR · III-V semiconductors
1 Introduction
Infrared photodetection was for a long time dominated by the Hgx Cd1 − x Te detec-
tors, which can be easily tuned to reach any operation wavelength in the mid-infrared
range (2–20 µm), by simply varying the alloy composition [1]. Since these devices
are based on band-to-band transitions, varying the Hg content in the alloy leads to
different band gap energies and, consequently, different threshold detection wave-
P. Pereira · L. Guerra · L. D. Pinto · P. L. Souza (B)

LabSem/CETUC, Pontifícia Universidade Católica, Rio de Janeiro 22451-900, Brazil
e-mail: plustoza@cetuc.puc-rio.br
P. Pereira · L. Guerra · G. M. Penello · M. P. Pires · L. D. Pinto · R. Jakomin · R. T. Mourão ·
M. H. Degani · M. Z. Maialle · P. L. Souza
DISSE-Instituto Nacional de Ciência e Tecnologia de Nanodispositivos Semicondutores,
PUC-Rio, Rio de Janeiro 22451-900, Brazil
G. M. Penello · M. P. Pires · R. T. Mourão
Instituto de Física, Universidade Federal do Rio de Janeiro, Rio de Janeiro 21945-970, Brazil
R. Jakomin
Campus de Xerém, Universidade Federal do Rio de Janeiro, Duque de Caxias 25245-390, Brazil
M. H. Degani · M. Z. Maialle
Faculdade de Ciências Aplicadas, Universidade Estadual de Campinas, Limeira, São Paulo
13484-350, Brazil

https://doi.org/10.1007/978-981-13-9155-2_29
364 P. Pereira et al.
lengths. However, due to difficulties in producing a homogeneous material over a

large area, these devices are not adequate for focal plane arrays, in which application
each pixel in a matrix of devices should have exactly the same performance [2–4]. In
the past decade, the most investigated alternative has been the quantum well infrared
photodetector (QWIP) fabricated with III-V semiconductors, which are based on
intraband transitions [5, 6]. Intraband transitions are, in principle, intrinsically lim-
ited to energies below the heterostructure conduction band offset. In the case of
III-V semiconductors, this means that energies above 500 meV cannot be detected,
which corresponds to wavelengths shorter than 2.5 µm. As a consequence of lattice
mismatch to the substrate, band-to-band photodetectors using III-V semiconductors
can only be fabricated for wavelengths shorter than 1.7 µm. So, there is a sort of
forbidden gap for these materials between 1.7 and 2.5 µm. This spectral interval is,
though, of major importance for satellite imaging and surveillance [7].
In 2016, an InGaAs/InAlAs superlattice structure with a central quantum well
with a different thickness from the otherwise periodic structure has been proposed to
reach this forbidden gap [8]. Absorption and photocurrent have been demonstrated
at 2.1 µm at 77 K, as well as, the possibility of tuning this operation wavelength
within the SWIR range by slightly changing the thicknesses of the quantum wells
and barriers.
In this work, the behavior of this device, which from now on will be called SLIP, as
a function of temperature is addressed, the different figures of merit are determined
and operation up to 160 K is demonstrated.
2 Experimental Details
The InGaAs/InAlAs SLIP structures have been grown by metalorganic vapor-phase

epitaxy lattice matched to the InP substrate at 100 mbar and 650 °C. Trimethylindium,
trimethylgallium, trimethylaluminum, arsine and phosphine were used as sources for
In, Ga, Al, As and P. After the deposition of an InP buffer layer, the QWIP structure,
which consists of a twenty-period SL with a wider central quantum well (CQW), is
grown. This modified SL is itself repeated twenty times. An InAlAs spacer of 30 nm
is deposited between each SL. A scheme of the structure is depicted in Fig. 1. The
CQW is n-doped with Si using a SiH4 flux to reach 2 × 1018 electrons per cm3 . The
whole active layer is sandwiched between two n-doped contact layers with the same
doping level as the CQW. The thicknesses of the CQW, barrier (B) and superlattice
quantum well (SLW) for the main investigated SLIP are nominally 7, 3 and 1.5 nm,
respectively.
Absorption and photocurrent measurements were carried out with Fourier trans-
form infrared spectroscopy (FTIR) in a waveguide mode, since perpendicular inci-
dent radiation does not couple to intraband transitions in QWIPs. For this purpose,
the grown samples were polished to produce a 45° coupling facet on which light is
focused at normal incidence. Additionally, this waveguide geometry allows multi-
ple reflections of the beam inside the sample increasing the interaction of radiation
Quantum Well Infrared Photodetector for the SWIR Range 365
Fig. 1 Scheme of the

photodetector structure.
Adapted from [8]
with the quantum wells (QWs) and therefore the absorption itself. The responsivity
measurements were carried out using a calibrated blackbody source.
Calculations using the split-operator method were employed to design appropriate
SLIP structures. In the simulations, the Schrödinger equation is solved for the desired
potential obtaining the eigenstates and eigenvalues. As the energies involved are
comparable to the bandgap of the material, a nonparabolic approximation is used.
The correction of these effects is considered in the effective mass of the electron,
which becomes dependent on the electron energy, rendering the calculation self-
consistent.
3 Results and Discussion
The calculated band structure of the previously designed SLIP structure is depicted
on the left panel of Fig. 2 [8]. In blue are the probability densities of the even parity
states, while in green are the odd parity ones. In gray are the probability densities for
the miniband states. On the right are the absorption spectra for transitions starting at
the ground (blue) and first excited (green) states, showing the possibility of operation
within the forbidden gap with such structures, since absorption from the ground state
to the third excited state gives rise to a peak at 578 meV (2.14 µm). Photocurrent
spectra for different bias voltages have been detected at 77 K for such structure at
this energy, as reproduced in Fig. 3 [8], demonstrating that, in fact, a SLIP can detect
radiation in the band 7 of the SWIR (2.105–2.155 µm). A second photocurrent peak
around 300 meV (4 µm), attributed to the optical transition between the ground state
and the miniband states, in which case the current is produced by tunneling from one
period to the subsequent one, is also observed.
Even though it has already been demonstrated that SLIPs based on III-V semicon-
ductors can reach this important wavelength range [8], it is crucial to determine up to
Fig. 2 Left panel: probability densities for the first four localized states (blue for even and green for
odd parity states) and for the miniband states (gray) superimposed on the conduction band profile
on the growth direction. Right panel: absorption spectra for transitions with the ground state (blue)
and the first excited state (green) as the initial states for the best-fit structure thicknesses Adapted
from [8]
Wavelength (m)
108 6 4 2
100
80
Photocurrent (arb. u.)
60
40
20
0.2 0.4 0.6 0.8 1.0 1.2

Energy (eV)
Fig. 3 Photocurrent as a function of bias for the investigated SLIP structure Adapted from [8]
Fig. 4 Photocurrent as a
function of temperature
under a +2 V bias
which temperature the device can work and its figures of merit such as responsivity,
detectivity and BLIP temperature, temperature at which the current generated due to
the background radiation equals the dark current.
Figure 4 shows the photocurrent spectra for temperatures between 80 and 160 K
at +2 V bias. It is clear that the peak at 587 meV is detected up to 160 K with an
evident raise in noise as the temperature increases. The lower energy peak around
300 meV is only observed up to 120 K most likely because carriers excited to the
miniband have the additional possibility of being thermally excited to states out of the
miniband, which do not allow current propagation, reducing the tunneling current.
Figure 5 shows the dark current versus voltage curves (I × V ) for temperatures
between 80 and 300 K in black using a heat shield with steps of 20 K. At null bias, the
dark current is of the order of pA and for +2 V bias is below nA at 80 K, increasing
to mA at room temperature. In green is the current versus voltage curve generated by
incident background radiation. From this plot, we determine the BLIP temperature
to be 120 K. The activation energy of the dark current generation mechanism can
be determined from the I × V data, building an Arrhenius plot for each different
bias. The determined activation energies are shown in Fig. 6 for biases between +
5 and −5 V. The activation energy for null bias is 295 meV, which agrees with the
energy difference between the ground state and the first miniband, indicating that
the tunneling current through the minibands is the main mechanism for dark current
generation. In order to reduce the undesirable dark current, thicker barriers between
the different superlattice periods should be implemented.
Fig. 5 Dark current versus

voltage for temperatures
between 80 and 300 K in
steps of 20 K. In green, the
current versus voltage curve
generated by incident
background radiation,
showing a BLIP temperature
of 120 K
Fig. 6 Activation energy as

a function of bias obtained
from the data of Fig. 5
The responsivity of the SLIP, current generated per incident radiation power, as
a function of bias is depicted in Fig. 7 for temperatures between 80 and 160 K. The
highest responsivity is obtained with a +4 V bias at 120 K and is equal to 2.1 mA/W.
At the highest operating temperature, 160 K, the device provides a 2.0 mA/W respon-
sivity when the applied bias is +2 V. The relatively low responsivity is somewhat
expected, since the oscillator strength for the optical transition between ground state
and third excited state is intrinsically low, as shown in the right panel of Fig. 2.
There one can clearly see that the absorption is more than two orders of magnitude
lower than the one from the ground state to the first excited state. Figure 8 shows the
detectivity as a function of bias of the SLIP, essentially responsivity divided by the
noise, for the same temperatures. The highest detectivity of 3 × 106 Jones is achieved
under three different conditions, namely temperatures 120, 140 and 160 K with +4,
+3 and +2 V biases, respectively. This occurs because as the bias voltage increases
not only the photocurrent increases, but also the dark current and, consequently, the
noise.
Fig. 7 Responsivity as a
function of bias for
temperatures between 80 and
160 K
Fig. 8 Detectivity as a
function of bias for
temperatures between 80 and
160 K
4 Conclusion
An InGaAs/InAlAs superlattice infrared photodetector, here called SLIP, which can

reach band 7 of the SWIR (around 2.0 µm operation wavelength) has been investi-
gated in terms of its figures of merit. Even though this previously forbidden wave-
length has been reached, the performance of the device still falls short from being
competitive. The optical transition involved to absorb 2.1 µm is from ground state
to third excited state, which has a relative small probability of taking place. An alter-
native would be to seek different designs, using genetic algorithms, for instance, to
be able to have the final state of the optical transition of lower order, which would
strengthen the absorption’s oscillator strength. Another important issue is reducing
the dark current and, therefore, the noise. It has been shown that the main mechanism
for generation of the dark current is the tunneling current through the superlattice
miniband states. Using a wider InAlAs barrier between each of the 20 periods should
significantly reduce the dark current, since it decays exponentially with the layer
thickness.
Acknowledgements This work was partially supported by FAPERJ, CNPq, FINEP, CAPES and
FAPESP.
References
1. Rogalski, A.: HgCdTe infrared detector material: history, status and outlook. Rep. Prog. Phys.
68(10), 2267 (2005)
2. Shen, S.: Comparison and competition between MCT and QW structure material for use in IR
detectors. Microelectron. J. 25, 713–739 (1994)
3. Goldberg, A.C., Kennerly, S.W., Little, J.W., Shafer, T.A., Mears, C.L., Schaake, H.F., Winn,
M.L., Taylor, M., Uppal, P.N.: Comparison of HgCdTe and quantum-well infrared photodetector
dual-band focal plane arrays. Opt. Eng. 42, 30–47 (2003)
4. Downs, C., Vandervelde, T.: Progress in infrared photodetectors since 2000. Sensors 13,
5054–5098 (2013); Author, F., Author, S., Author, T.: Book title, 2nd edn. Publisher, Loca-
tion (1999)
5. Liu, H.C.: Quantum dot infrared photodetector. Optoelectron. Rev. 1, 1–6 (2003)
6. Rogalski, A.: Infrared detectors: an overview. Infrared Phys. Technol. 43, 187–210 (2002)
7. See www.landsat.usgs.gov or www.landsat.gsfc.nasa.gov
8. Guerra, L., Penello, G., Pires, M., Pinto, L., Jakomin, R., Mourão, R., Degani, M., Maialle,
M., Souza, P.: Detecting infrared radiation beyond the bandoffset with intersubband transitions.
IEEE Photonics Technol. Lett. 28, 1641–1644 (2016)
Independent Feeding of People Affected
with Osteoarthritis Through a Didactic
Robot and Visual Control
Arturo Jiménez, Katherine Aroca, Vicente Hallo, Nancy Velasco

and Darío Mendoza
Abstract This chapter presents a work that will help in the consumption of the
food of people with osteoarthritis. In this way, the patient will avoid discomfort in
the joints of their hands. The system consists of an artificial vision algorithm and
independent feeding part. The vision algorithm allows the detection of the face,
localization, and tracking the person’s mouth. The independent feeding part consists
of a didactic robotic arm. The robotic arm takes the food from the dish. The vision
algorithm detects and tracks the face, then locates the position of the mouth. Finally,
the robotic arm delivers the food to the user. The program was developed in Python
using OpenCV and Dlib libraries. The face alignment method has an average of 94%
of effectiveness.
Keywords Visual control · Recognition and monitoring · Robot feeding people ·

Osteoarthritis
1 Introduction
Osteoarthritis is characterized by progressive degenerative changes primarily involv-

ing the hips, knees, spine, hands, and feet. Osteoarthritis causes stiffness, inflam-
mation, pain, deformity, and weakness. The person with osteoarthritis cannot hold
objects such as a spoon. Military troops are more likely to develop osteoarthritis than
civilians.
Robots are one strategy to help with osteoarthritis. Robots are directly relationship
with computer control. Computer vision offers the robot information about its work
A. Jiménez · K. Aroca · V. Hallo (B) · D. Mendoza

Departamento de Energía y Mecánica, Universidad de las Fuerzas Armadas ESPE, Sangolquí,
Ecuador
e-mail: vdhallo@espe.edu.ec
N. Velasco
Departamento de Ciencias Exactas, Universidad de las Fuerzas Armadas ESPE, Sangolquí,
Ecuador

https://doi.org/10.1007/978-981-13-9155-2_30
372 A. Jiménez et al.
environment so that this information can be used for its own control. An eyeinhand
system is a robot equipped with an autofocus camera.
There are many designs and constructions of self-feeding robots like Handy1,
Winsford feeder, Neater Eater, My Spoon, Meal Buddy, and Mealtime Partner Dining
System [1]. Meal Buddy [2] has three degrees of freedom in robotic arm and three
bowls. After the system scoops the food, the robotic arm scrapes the surplus food off
the spoon with the rod on the bowls. Handy1 [3] consists of a five degree of freedom
robot, a tray unit and a gripper. Handy1 allows a user to select food from any part of
the tray.
Some robots are for specific food like the design of feeding robot for Korean food
[4]. This robotic system consists of two robotic arms: One is a food transferring
robotic arm and one is a food grasping robotic arm. Moreover, some students build
robots to help with osteoarthritis in [15]. They utilized their robot models to educate
the public on osteoarthritis.
We propose a low-cost system to facilitate feeding to patients with osteoarthritis.
For this, the hardware part is represented by a robotic arm. The software part is
implemented in Python and image processing libraries.
2 Development
2.1 Description of Tools
The Raspberry Pi 3 is a reduced-board computer. One of the great advantages is its free
use both educationally and privately. Its operating system offers an adapted version
of Debian, known as Rasbian [5]. A low-weight high-resolution camera is required
for the recognition and monitoring of the user’s mouth, so that the robotic arm has
no problems when it comes to mobilizing and being compatible with the Raspberry
Pi 3. The camera that meets the requirements of the system is the FACECAM 1000X
for its weight of 50 g, size (60 × 20 × 22 mm), resolution (1 MP, 1280 × 720, 640
× 480 pixels), and the sensor (CMOS HD 720p pixels) that make this alternative the
most convenient for this project.
2.2 System Flow
The project is composed of the following processes (Fig. 1).

Independent Feeding of People Affected with Osteoarthritis … 373
Fig. 1 System flow diagram
2.3 Face Detection Algorithm
The Haar Cascade function of the OpenCV library is used in face detection, which
uses pre-trained classifiers to identify faces. The front face classifier function is
used. The video input is loaded and it is transformed to grayscale mode for a better
processing of the image. Then the detection of the face is done by the detectMultiScale
function, the same as with a copy in scale of grays finds all the faces in each frame
[6].
2.4 Face Tracking Algorithm
It is not necessary to apply face detection in each frame in real-time videos. Instead,
tracking algorithms can be used because it is easier to track the face during the video.
When the person makes a movement, the detection algorithm has to return to zero
in order to detect the new position of the face; so a tracking algorithm is needed
to complement the detection algorithm. This not only improves the tracking speed
eliminating the need to redetect the face in each frame. The system also becomes
more robust for two reasons. First, it retains the facial characteristics after detecting
the face and second, it becomes more resistant to rotation and photometric changes
[7, 8].
This project is designed to track only one face (one user) because tracking more
than one face can generate errors in the control of the didactic robotic arm.
Dlib allows to track the position of an object as it moves frame by frame in a video
sequence [9].
Fig. 2 Face detection using

Haar Cascade
The Dlib correlation tracker is used to track the face. It extracts the characteristics
of the detected face, to subsequently track it. After that, a rectangle is drawn in the
area where the face has been detected (Fig. 2).
It is necessary to calculate the area of the rectangle that contains the face to know
if Haar Cascade has detected the face. If it is greater than zero, the detector works
well and initiates the tracker.
2.5 Mouth Detection
Dlib allows to track the position of an object as it moves frame by frame in a video
sequence. This correlation tracker has a bounding box to follow the object in the
video frame. Correlation trackers, as the name implies, work by correlating a set of
pixels from one frame to another [8].
2.6 Tracking the Mouth
The correlation tracker Dlib is used for face tracking. The same one that extracts the
characteristics of the detected face, to later track through the video sequence.
After face detection and tracking, an identifier of the geometric structure of the
face of the Dlib library is used. It is called shape_predictor_68_face_landmarks. It
is a standardized library to deliver the reference points of the face by through the
alignment in one millisecond [10]. The library delivers 68 facial points (Fig. 3),
which define the facial regions, such as the corners of the mouth, eyebrows, eyes,
nose, and mouth, using parts of a simple matrix in Python [11].
The facial alignment is performed by a set of regression trees, which consists of
taking an image in the form of a numpy matrix and returning a 68 × 2 matrix, each
row is conformed with the x, y coordinates of a common point of the facial feature in
Fig. 3 Points of facial

coordinates
the input image, in other words, it contains coordinates for a particular facial feature.
In this project, points 49 through 68 are used because they contain all the information
in the mouth [12].
2.7 Reverse Kinematics of the Robotic Arm
The inverse kinematics is necessary to perform a correct positioning of the robotic

wrist. The geometric transformation method is used in order to determine the angle
to which each of the servomotors must turn [13].
In this project, the calculations will be made first on the YZ plane of the arm
to determine the angles of each joint that allow the didactic robotic arm to take the
required position according to the coordinate because in this way the calculations
will be made in the plane for greater ease.
The arm has been divided into three vectors (Fig. 4) to facilitate the calculation.
The vectors are arm, forearm, and wrist. When the analysis is carried out, the angle
δ is calculated which is imposed by the user. The established spatial coordinate of
the armrests in this angle, and it will serve as the basis for calculating the rest of the
angles. Next, we decompose the wrist vector:
wrist Y = wrist ∗ cos(δ) (1)
wrist Z = wrist ∗ sin(δ) (2)
Calculate Y and Z in order to find the diagonal AC and the angle α:

Fig. 4 Geometrical components of inverse kinematics
Y = Y − wrist Y (3)
Z = wrist Z + Z (4)
√ 2
AC = ((Y ) + (Z )2 ) (5)
tan(α) = Z /Y (6)

α = tan−1 Z /Y (7)
Angles β and γ are calculated using the law of cosines:

cos(β) = arm2 − forearm2 + AC2 /(2∗arm∗AC). (8)

cos(γ ) = arm2 + forearm2 − AC 2 /(2∗arm∗forearm). (9)
After we found β and γ , it is possible to determine the angles that the forearm
can turn based on Fig. 4.
arm angle = 90◦ + β + α. (10)
forearm angle = γ . (11)

The triangles ABC, CDE, and ACE are located to calculate ε. After that, ε is
determinate based on the complementary angles:
ε = ε + ε + ε (12)
In triangle ABC, the following equation is obtained
ε = 180◦ − β − γ (13)
In the triangle ACE, we have the following equation
ε = 180◦ − 90◦ − α (14)
ε = 90◦ − α (15)
In the triangle CDE, we use the following equation
ε = 180◦ − 90◦ − δ (16)
ε = 90◦ − δ (17)
Then ε is:
ε = 180◦ − β − γ + 90◦ − α + 90◦ − δ (18)
The angle of rotation of the base of the robot that allows the manipulator to locate
in the X, Y, Z plane is determined, and the angle between X and Y must be known to
obtain its spatial position [14].
Angle of rotation = tan(−1) (x/y) (19)
The first image is the starting point to know the position of the mouth. A copy of
the original frame is made to guarantee a greater acceleration in the processing of
the image and further, resize with values of 320 × 240 pixels. The selected resolu-
tion makes the system more robust and tracking more efficient compared to larger
dimensions.
Basically, there is a delimiting area of the mouth; if the user moves outside the
area of the center of the image as shown in Fig. 5, then the robotic arm begins the
tracking process to be located within the center of the area. For example, if the user is
located in a sector with coordinates P(60, −100), through the serial communication
already established, the respective location characters will be sent to position the
robotic arm at the center of the image.
Fig. 5 Central area of the image
Fig. 6 Screening and monitoring of the mouth using face alignment method and Haar Cascade
3 Experimental Results
3.1 Experiments
Figure 6 shows the robotic arm tests tracking user’s mouth in face alignment method
and Haar Cascade’s.
The precision and speed of the system are considered to determine the best method
for the recognition and tracking of the mouth. If the recognition of the mouth is not
exact, the robotic arm would produce failures when feeding the user.
3.2 Results
The tests were performed on five users, inside a house, with different light settings
and in different distances. The test distance is from 5 to 50 cm. It is based on the
reach of the robotic arm. Once the user’s face is located, a green rectangle is drawn
around him and his mouth (Fig. 7).
Figure 8 shows the percentage of effectiveness of each method at different dis-
tances up to 50 cm.
The alignment method of the face of Dlib has a higher percentage of effective-
ness than the Haar Cascade method of OpenCV. The face alignment method has a
maximum percentage of 98% and a minimum of 90, 94% in average. Haar Cascade
has a maximum percentage of 70% and a minimum of 40%. These percentages were
obtained from the average effectiveness of tracking of each user.
Fig. 7 Left: face alignment with Dlib. Right: face recognition image with Haar Cascade
Fig. 8 Accuracy of tracking of the mouth according to the tracking distance

The Haar Cascade method for mouth detection is less precise and its parameters
much more difficult to adjust, while alignment of the face is more precise with simpler
adjust parameters.
Both methods are executed in a speedy manner and do not require a very powerful
processor for their execution since can run from a Raspberry Pi 3 without problems.
Face alignment method can be adapted to any environment for detection and
tracking of the mouth. Haar Cascade needs to reset of its parameters every time the
environment changes.
4 Conclusions
This project was carried out in free software to reduce costs.

The Dlib library is used to efficiently implement the mouth tracking algorithm.
Dlib allows the recording of the detected region frame by frame, the reduction of
errors and the improvement of the tracking speed.
The visual servo control provides an adequate response for the control of the
movements of the manipulator.
The face alignment method has an average of 94% of effectiveness. In the other
hand, Haar Cascade has an average of 55% of effectiveness. Therefore, face alignment
method has the best results in this application.
The calculation of the inverse kinematics allows to position the end effector of
the arm exactly. For this in space (X, Y, Z), the angles of your joints are calculated.
Assistive technology allows to considerably improve the quality of life of a war
veteran affected with osteoarthritis.
Acknowledgements We are grateful to “Universidad de las Fuerzas Armadas ESPE” to carry out
the tireless duty of teaching.
References
1. Song, W.-K., Kim, J.: Novel assistive robot for self-feeding. En Robotic Systems-Applications,
Control and Programming. InTech (2012)
2. Sammons Preston http://www.sammonspreston.com/
3. Topping, M.: Handy 1, a robotic aid to independence for severely disabled people. In: 7th
International Conference on Rehabilitation Robotics, pp. 142–147 (2001)
4. Song, W.-K, et al.: Design of novel feeding robot for Korean food. In: International Conference
on Smart Homes and Health Telematics. Springer, Berlin, Heidelberg, pp. 152–159 (2010)
5. Aranda, D.: Electrónica: plataformas Arduino y Raspberry Pi. Dalaga, Buenos Aires (2014)
6. Lee, S., Lee, C.: Illumination normalization and skin color validation for robust. de IS&T
International Symposium on Electronic Imaging 2016, Seoul (2016)
7. Castrillón, M., Déniz, O., Hernández, D.: A comparison of face and facial feature detectors
based. Mach. Vision Appl. 22, 481–494 (2011)
8. Dollár, P., Appel, R., Belongie, S., Perona, P.: Fast feature pyramids for object detection. IEEE
Trans. Pattern Anal. Mach. Intell. 36(8), 1532–1545 (2014)
9. Klare, B.F., Klein, B., Taborsky, E., Blanton, A., Cheney, J., Allen, K., Grother, P., Mah, A.
Jain, A.K.: Pushing the frontiers of unconstrained face detection and recognition: IARPA Janus
Benchmark A. In: de IEEE Conference on Computer Vision and Pattern Recognition (CVPR),
Boston (2015)
10. Liao, S., Jain, A.K., Li, A.Z.: A fast and accurate unconstrained. IEEE Trans. Pattern Anal.
Mach. Intell. 38(2), 211–223 (2015)
11. Cheney, J., Klein, B.: Unconstrained face detection: state of the art baseline and challenges.
In: International Conference on Biometrics (ICB), Phuket (2015)
12. Kazemi, V., Sullivan, J.: One millisecond face alignment with an ensemble of regression trees.
Estocolmo (2017)
13. Baki Koyuncu, A.M.G.: Software development for the kinematic. World Acad. Sci. Eng. Tech-
nol. 1(6), 1575–1580 (2007)
14. Cruz, A.B.: Cinemática inversa, de Fundamentos de robótica. McGraw-Hill, Madrid (2013)
Adjustable Structure in Height
for Assisted Feeding in Patients
with Osteoarthritis
Katherine Aroca, Arturo Jiménez, Nancy Velasco, Vicente Hallo

and Darío Mendoza
Abstract This article presents the design and building of an adjustable height struc-
ture for the assisted feeding of people with osteoarthritis. The project will improve
the quality of life of these patients. The design is governed by the INEN (Ecuado-
rian Standardization Service) norm. The fundamental criteria in the document are
oriented to create and size a mechanical structure matching with the anthropometric
measurements of the users. The parameters include the feeding area, mechanism
materials, and the correctly defining a height adjustment.
Keywords Ergonomics · Osteoarthritis · Feeding · Mechatronic design
1 Introduction
The osteoarthritis is the most common joint disease. According to international stud-
ies, more than 70% of adults older than 50 years of age present radiological signs of
osteoarthritis in any joint; women are the most appropriate. The osteoarthritis affects
joints like hands, hip, or knee [1]. Besides, osteoarthritis is one of the main causes of
disability in the retired militaries. It affects a greater proportion of retired militaries
than in comparison with civilians. The most frequent reasons are stress and injuries
caused by repetitive basic training and combat risks such as bomb explosions, which
directly affect the cartilage cells and soft tissues [2].
Currently, there are some structures to feed people with physical disabilities, such
as Obi, which is a simple minimalist structure that works with a robotic arm, two
push buttons and a plate divided into four compartments for different types of food
K. Aroca · A. Jiménez · V. Hallo (B) · D. Mendoza

Deptartamento de Energía y Mecánica, Universidad de las Fuerzas Armadas ESPE, Sangolquí,
Ecuador
e-mail: vdhallo@espe.edu.ec
N. Velasco
Deptartamento de Ciencias Exactas, Universidad de las Fuerzas Armadas ESPE, Sangolquí,
Ecuador

https://doi.org/10.1007/978-981-13-9155-2_31
384 K. Aroca et al.
Table 1 Data for sizing the Dimension Reference Measure (mm)

surface of the table height
A Popliteal height 470
E Lumbar support height 180
[3]. There are many designs and constructions of self-feeding robot like Handy1,
Winsford feeder, Neater Eater, My Spoon, Meal Buddy and Mealtime Partner Dining
System [4]. Handy1 [5] consists of a five degree of freedom robot, a tray unit, and a
gripper. Handy1 allows a user to select food from any part of the tray. Meal Buddy
[6] has three degrees of freedom in robotic arm and three bowls. After the system
scoops the food, the robotic arm scrapes the surplus food off the spoon with the rod
on the bowls.
The study presented below emphasizes the design of a height adjustable table for
patients with osteoarthritis. For this, it is necessary to determine technical criteria
to implement a structure with good space of operation according to the physical
characteristics of the user.
2 Elements of the Structure
The structure will have a space for: a robotic assistance arm with a range of at least
520 mm and lightweight, food plates (soup, entrée, dessert), a drinking glass with a
straw, a spoon with its spoon holder, electronic and drive elements, a screen, and a
height adjustment mechanism.
2.1 Feeding Area
The maximum feeding area is determined by the maximum reach of the robotic arm,
in this case 1040 mm2 . Within this space, the user can be fed, and Table 1 shows
the elements comprising it. Figure 1 shows the elements of the maximum feeding
area. Number one represents the prototype of robotic arm. Number two represents
the area of work or feeding area and number three represents the maximum robotic
arm reach.
2.2 Support Table
The INEN norm 1641:1988 [7] determines nominal values of the free space for the
accommodation of legs as shown in Fig. 2. The table width (T ) has a minimum value
of 580 mm; it should be defined such as requirement for the implementation of the
Adjustable Structure in Height for Assisted Feeding in Patients … 385
Fig. 1 Elements of the maximum feeding area
Fig. 2 Free space for legs accommodation according to NTE INEN 1641:1988 standard
elements on the table with a value of 810 mm. In addition, the depth at the feet level is
determined by a minimum value (V ) of 600 mm. After adding a structure for driving
mechanism operated by patient’s feet on the back of the mechanical structure, its
height will be increased to 670 mm for user comfort.
The dimensions of the table are 810 mm × 670 mm for width and depth,
respectively.
The Pythagorean theorem is used to calculate the exact location radius of food
utensils. The right triangle is formed by the oscillation arm and the union of the arm
with the final wrist or effector (Fig. 3).
If bo is measurement of the oscillation arm, bm is arm measurement + wrist
measurement, and r T is measurement of the radius of location of food utensils, thus:
bm2 = bo2 + r T2 (1)

√ √
rT = (bm2 − bo2 ) = (3172 − 1022 ) mm = 300 mm (2)
The optimum total radius of location of the food utensils is 300 mm with respect
to the initial point of location of the robotic arm. There are three food dishes in total,
386 K. Aroca et al.
Fig. 3 Triangle with real measurements used to calculate the radius of location of food utensils
with respect to the center of the robotic arm
Fig. 4 Location of food

utensils
and the location of each plate is required to calculate the separation length between
them in first place. The arc length formula is used assuming an angle of 36°, as shown
in Fig. 4.
If ∅ is the angle formed between radius, c is the length of separation between each
plate, thus,
c = 2r T2 sin(∅) (3)
c = 2(300) sin(36◦ /2) = 185 mm (4)
The separation between plates is delimited by a value of 185 mm and angle of

36°. The glass of drink is at 300 mm with an angle of 90° with respect to the second
plate of food.
The total mechanism is present in Fig. 5.
Fig. 5 Parts of the mechanical structure
The transmission of circular to continuous linear motion is used with the rack and
pinion elements in the mechanism of regulation of the height range.
2.3 Adjustable Height Range of the Table
One of the parameters of ergonomics is to reduce user’s risks and discomforts in

certain positions [8]; therefore, the regulation in height fulfills the need to adapt it
to different heights. For this purpose, the following measurements have been carried
out to obtain a good range of regulation and offer more ergonomics and comfort.
The scheme is presented in Fig. 6. In the first place, the INEN 1641: 1988 NORM
[9] is used. The minimum height (hcodo) to which the support table should be fixed
depends on popliteal height in the sitting position (A) and the distance of the space
from the chair to the elbow known as lumbar support height (E).
hcodo = A + E (5)
The data for sizing the surface of the table height is shown in Table 1.
The minimum height of adjustment of the work surface is given by the height
hcodo:
hcodo = (470 + 80) mm = 650 mm (6)

388 K. Aroca et al.
Fig. 6 Seated workplace
Fig. 7 Sitting position of women from 60 to 90 years
Table 2 Data for sizing the Dimension Reference Measure (mm)

surface of the table height
30 Popliteal height 360
37 Lumbar support height 190
The sitting position of adult women is taken as the minimum height of adjustment
of the work surface. In the same way, the popliteal height in the sitting position (30)
and the distance of the space from the chair to the elbow (37) are added, in order
to obtain the minimum height at which the support table should be fixed [10]. The
details are in Fig. 7.
The data for sizing the surface of the table height is shown in Table 2.
The minimum height of adjustment of the work surface is given by the height
hcodo, where:
hcodo = (360 + 190) mm = 550 mm (7)
In another study based on the INEN 1641: 2015 standard [7], a minimum height
value of 650 mm is established for fully adjustable tables, up to a height of 850 mm.
In this case, according to previous studies, a height of 550 mm is established, up
to a maximum of 670 mm, in other words, an interval of 120 mm, to satisfy the need
to regulate the height of all the elderly.
2.4 Pinion and Rack Design
Due to the low cost and improvements in the mechanical properties, the plastic
materials have increased their utility within the industry. In the market, there is a
great variety of pinions with standard zips, mostly with pressure angle of 20° and
made of nylon [11], for applications of uniform shock, safe operation, low friction,
corrosion resistance, and good properties in how much to wear.
The rack and pinion elements must comply with certain design requirements to
avoid possible failures during the movement of the mechanism, mainly for plastic
gears are supporting the impact of the tip of the tooth and the bottomland or chamfer
of the tooth of the rack, which produces fracture in the material due as the failure by
fracture between plastic gears due to the impact of the tip of the tooth of the pinion
with the bottomland or chamfer of the tooth of the rack. This defines a minimum
number of teeth of the gear, due to the established pressure angle, not less than 18
teeth are used [12].
The material used is nylon. The allowable bending stress S at is 41 MPa. The
minimum reference space for pinion location DReference is 92 mm. The length of the
rectangular zip l is 120 mm. The number of teeth selected from the pinion zp is 21.
The geometry of the rack and pinion is defined by [13]. Thus, the pinion and rack
module is:
m = DReference /z p = 92/21 mm = 4 mm (8)
The diameter of the step DP is equal to the multiplication of pinion or rack module
m and number of teeth selected from the pinion zp :
D P = mz p = (4) (21) mm = 84 mm (9)
The diameter step pd is equal to the multiplication of pi (π ) and pinion or rack

module m:
pd = π m = π (4) = 12.57 mm (10)

390 K. Aroca et al.
The number of steps carried out (n°steps) is equal to a length of the rectangular
zip divided by diametral step of the pinion or rack:
n ◦ steps = 1/ pd = 120 mm/12.57 mm = 9.56 steps (11)
The number of turns of the pinion n°turnsp is equal to the number of steps carried
out
(n°steps) divided by the number of teeth selected from the pinion zp :
n ◦ turns p = (n ◦ steps)/z p = 9.56/21 = 0.45 Turns (12)
The revolution per minute of the pinion np is equal to the number of turns of the
pinion divided by the thickness of the tooth t:
◦ turns n p = n ◦ turns p /t = (0.47Turns/1.8s) (2π rev/1Turns) (60s/1min) = 95 RPM
p
(13)
The zipper speed V ac in mm/s is equal to the multiplication of diametral step of

the pinion or rack pd , the number of teeth selected from the pinion zp , and revolutions
per minute of the pinion np :
Vac = pd Z p n p = 12.57 mm × 21 × 1.59REVS = 419 mm/s (14)
The dimensions of head a, root b, and spacing c are 4, 5, and 1 mm, respectively.
The external diameter Do is equal to a primitive diameter of the pinion Dp plus
double the head a:
Do = D p + 2a = 84 + 2 × 4 = 92 mm (15)
The diameter of the root DRP is:
DRP = D P − 2b = 84 − 2 × 5 = 74 mm (16)
The total depth ht is:
h t = a + b = 4 + 5 = 9 mm (17)
The depth of work hk is:
h k = a + a = 1 + 1 = 2 mm (18)
The tooth thickness t is:
t = pd /2 = 12, 57/2 = 6.285 mm (19)

The face width F is:
F = 12(m) = 12(4) = 48 mm ≈ 50 mm (20)
The critical load on a tooth will be generated when the load is at the highest point
of contact in a single tooth, so the force will have used the value of the mass that will
be lifted, this one, a tangential force will generate a tension momentum or maximum
bending at the base of the teeth of the rack and pinion. To solve this requirement was
used a design software, it is estimated an approximate value of the total mass to be
lifted of 36.16 kg which consists of the robotic assistance arm, the food plates, food
utensils, to screen, the table, the support profiles of the table, and the approximate
mass of the user’s arms.
The tangential force W t :

Wt = mass(gravity) = 36.16 Kg 9.81 m/s2 = 354.73 N (21)
2.5 Bending Stress on the Teeth of the Pinion and Rack
These calculations are based on AGMA Standards. If K a is an application factor for

the flexural strength, K s is the size factor for the flexural strength, K m is the load
distribution factor for the flexural strength, K B is the factor thickness of the crown,
K v is the dynamic factor for the flexural strength, K v dynamic factor for the flexural
strength, and then, the tension or bending stress on the teeth σ t is:
σt = (Wt /F m J )((K a K s K m K B )/K V ) (22)
The effort for maximum tension in the pinion is:
σtp = (354.73/(50 × 4 × 0.33))((1 × 1 × 1.126 × 1)/0.85) = 7.12 MPa (23)
Therefore, it is true that the maximum bending stress on the pinion σ tp is less than
the permissible bending stress S at :
σtp < Sat → 7.12 MPa < 41 MPa (24)
If J p is geometry factor for the pinion and J r is geometry factor for the rack, the
factors in the flexural stress equation are the same for the rack except for the J value
of geometry so the effort in the rack can be calculated as follow:
σtr = σtp (Jp /Jr ) = 7.12 (0.33/0.35) = 6.71 MPa (25)

392 K. Aroca et al.
Fig. 8 von Mises stress analysis of the pinion
Also, the maximum bending stress on the rack is less than the allowable bending
stress:
σtr < Sat → 6.71 MPa < 41 MPa (26)
2.6 Maximum Bending Stress of the Pinion and Rack
It is the maximum bending at the base of one of the teeth of the pinion giving a value
of 8.99 MPa using the von Mises deformation method, therefore:
σtp < Sat → 8.99 Mpa < 41 Mpa (27)
The von Mises effort of the pinion is less than the allowable bending stress.
Figure 8 shows the stress through design software.
In the same way, the maximum bending at the base of one of the teeth of the rack
gives a value of 6.84 MPa, thus:
σtr < Sat → 6.84 Mpa < 41 Mpa (28)
The von Mises effort of the rack is less than the allowable bending stress. Figure 9
shows the analysis of efforts through a design software.
Fig. 9 von Mises stress analysis of the rack
Table 3 Minimum height

Study type Minimum support table adjusting height (mm)
INEN 1641:1988 Norm 650
Study of the anthropometric registry in females 550
between 60- and 90-year-old in Mexico
INEN 1641:2015 Norm 650
3 Experimental Results
The feeding area has the dimensions 1040 × 1040 for width and depth, respectively.
The support table has the dimensions 810 × 670 for width and depth, respectively.
Then, the dimensions of the support table have lower values compared to the dimen-
sions of the feeding area allowing to take these dimensions as part of the design of the
mechanical structure. Table 3 shows the minimum height obtained through a study
of the INEN norms and an anthropometric registry study conducted in Mexico [10].
Based on Table 3, the minimum height required to adjust the adjustable support
table will be 550 mm, based on the anthropometric registry study of women aged
60–90 years in Mexico. Since these users are those of shorter stature, and the maxi-
mum height will be 670 mm, to allow the regulation of the height of the mechanical
structure for the other beneficiaries. The results obtained from the analysis of bending
stresses are in Table 4.
The stress measurements obtained by the AGMA standards and computer-aided
engineering are very similar; therefore, it guarantees the reliability of these calcu-
394 K. Aroca et al.
Table 4 Results obtained Study type Pinion Rack

from the analysis of bending
stresses Bending stress according to AGMA (MPa) 7.12 6.71
Bending stress according to von Mises (MPa) 8.99 6.84
Permissible nylon bending stress (MPa) 41 41
lation methods. Also, these values are lower than the permissible bending stress;
therefore, the material selected for the rack and pinion is nylon.
4 Conclusions
The variation in height in each elderly affected with osteoarthritis allows that the
design of the table adjustable in height offers greater ergonomics, comfort, and
effectiveness when feeding and also reduces risks or discomfort that the user may
have in certain positions, improving their quality of life.
Among the technical criteria implemented in the research is the area of power
the one that limits the dimensions of width and depth of the mechanical structure in
order to avoid shocks and discomfort with the elements that make up protecting the
user from being hurt.
The correct distribution of the spaces for the food utensils and the robotic arm
allow the patient to position themselves in a suitable place to taste their food.
The height adjustment mechanism, formed by means of the rack and pinion ele-
ments, allows the vertical translation of the mechanical structure. The range of height
regulation is delimited by the studies of the INEN standards and an anthropomet-
ric study carried out in Mexico. AGMA standards and computer-aided engineering
are proven to be effective and reliable tools to select nylon as a material for the
implementation of the elements and thereby comply with the economic and design
parameters necessary for the selection of materials.
References
1. Solis, U., García y A. de Armas, V.: Rasgos demográficos en la osteoartritis de manos. Revista
Cubana de Reumatología 16(3), 275 (2014)
2. Cuddigan Law, [En línea]. Available: https://www.cuddiganlaw.com/library/making-the-
service-connection-for-arthritis.cfm?fbclid=IwAR2vnmcOQHx9nETuGUp_bWKDo9Zr9se_
pnASD0rHiyfjOPaJTFH4HuVRQvs [Último acceso: 22.01.2018]
3. Decker, J.: [En línea]. Available: https://meetobi.com/ [Último acceso: 2019 Enero 6] (2019)
4. Song, W.-K., Kim, J.: Novel assistive robot for self-feeding. En Robotic Systems-Applications,
Control and Programming. InTech (2012)
5. Topping, M.: Handy 1, a robotic aid to independence for severely disabled people. In: 7th
International Conference on Rehabilitation Robotics, pp. 142–147 (2001)
6. Sammons Preston. http://www.sammonspreston.com/
7. N. I. 1641:2015, Muebles de oficina, escritorios y mesas. Requisitos, Quito (2015)

8. Herrera Saray, P.: Ergonomía y el hábitat para la tercera edad. Páginas 87, 35–89 (2010)
9. N. I. 1641:1988, Muebles de oficina, escritorios y mesas. Requisitos, Quito (1988)
10. De Lavalle Herrera, Y.: Diseño y ergonomía para la tercera edad, 1st ed., México: Universidad
Nacional de México, p. 304 (2014)
11. Roydisa, Roydisa [En línea], (2012). Available: https://www.roydisa.es/wp-content/uploads/
2012/12/engranajes.pdf [Último acceso: 24 Junio 2017]
12. Carro Suárez, J., Flores Salazar, F., Flores Nava, I.: Análisis de esfuerzo de contacto en engranes
rectos aplicando el criterio de Hertz con ingeniería asistida por computadora. Científica 16(1),
25–32 (2012)
13. Mott, R.: Diseño de Elementos de Máquinas. In: Guerrero, P. (ed.) Pearson Education, México
(2006)
Voice-Controlled Assistance Device
for Victims of Gender-Based Violence
Miguel A. Domínguez , David Palomeque , Juan M. Carrillo ,

José Mª Valverde , Juan F. Duque , Bruno Pérez and Raquel Pérez-Aloe
Abstract One of the biggest problems that society is currently facing is violence
against women. In recent years, tangible progress in protecting and saving the lives
of female victims of intimate partner/family-related homicide has not been made,
so targeted responses are clearly needed. In this work, an electronic device to help
victims of gender-based violence who live with their aggressor has been designed.
The system is built on Bluetooth Low Energy technology allowing a wireless com-
munication between device and mobile phone with a low power consumption. The
device is controlled by three different commands and is capable of sending mes-
sages through a mobile phone to a Control Center. Depending on the nature of the
received messages, the Control Center will take the appropriate measures to assist
the victim. The design has been made paying special attention to a reduced size so
that the device can easily be camouflaged in any accessory of the victim’s jewelry,
thus going unperceived to the possible aggressor.
Keywords Gender-based violence victims · Assistance device · Voice-controlled
M. A. Domínguez · J. M. Carrillo · J. M. Valverde · J. F. Duque · B. Pérez · R. Pérez-Aloe (B)

University of Extremadura, 06006 Badajoz, Spain
e-mail: raquel@unex.es
M. A. Domínguez
e-mail: madominguez@unex.es
J. M. Carrillo
e-mail: jmcarcal@unex.es
J. M. Valverde
e-mail: valverde@unex.es
J. F. Duque
e-mail: duque@unex.es
B. Pérez
e-mail: bperez@unex.es
D. Palomeque
Institute of Microelectronics (US-CSIC), 41092 Seville, Spain
e-mail: palomeque@imse-cnm.csic.es

https://doi.org/10.1007/978-981-13-9155-2_32
398 M. A. Domínguez et al.
1 Introduction
Nowadays, one of the problems of great social impact is the violence against women.
According to a new research published by the United Nations Office on Drugs and
Crime (UNODC), released for the International Day for the Elimination of Violence
against Women, a total of 87,000 women were intentionally killed in 2017. Around a
third (30,000), were killed by their current or former intimate partner, someone they
would normally expect to trust [1]. The main health consequences of this violence
are: death and injury, depression, problems with alcohol consumption, sexually trans-
mitted diseases, unwanted pregnancies and abortions, babies with low birth weight,
among others [2].
Since the 1980s, Spain has focused its attention on the problem of violence suf-
fered by women, reflecting his concern in several legal and regulatory programs
[3–12]. It is worth highlighting, the agreement signed between several Spanish Min-
istries for the implementation of the protocol of action of the monitoring system by
telematics means of the fulfillment of the restraining orders in matters of gender-
based violence. The monitoring system consists of a radiofrequency transmitter that
emits a radio frequency (RF) signal that is received by a track unit. Both are devices
for the accused. The track unit is a GPS tracking device that incorporates the basic
functionalities of a mobile phone in addition to the reception of the RF signal emit-
ted by the RF transmitter. The device for women is a track unit, practically equal to
the defendant’s and incorporates an external radio frequency antenna that allows to
detect the radio frequency signal of the RF transmitter of the accused. Figure 1 shows
the devices for the accused and the corresponding for the woman. When an alarm
occurs, the Control Center performs the communications provided in the protocol of
intervention for each type of alarm. Despite these actions, gender-based violence in
Spain still constitutes one of the main social problems.
Although there are multiple mechanisms that currently have the victims of gender
violence to protect themselves and ask for help, the reality is that the number of
fatalities does not seem to decrease and remains at a very high number. It should
Fig. 1 Devices of the monitoring system by telematics means: for the accused (a) and (b), (c) is
the device for women
Voice-Controlled Assistance Device for Victims of Gender-Based … 399
be noted that, only in the last year, 60 women were killed, of which 78% had never
processed a complaint against their aggressor [13]. Therefore, the current conception
of surveillance of the aggressor is insufficient for a real prevention of situations of
gender violence and targeted responses are clearly needed.
In order to provide a possible solution to this problem, in this paper we present
a device for asking help, designed for people suffering from gender-based violence,
including those who have not reported their aggressor and continue to live with him.
The purpose is to develop a system of personalized help, through electronic devices
easily camouflaged in the accessories of the victims, such as watches, necklaces,
bracelets, brooches, as well as a set of actions and strategies, aimed at returning to a
normal life this group at risk of social exclusion.
The present paper has been organized as follows: Sect. 2 describes the system
proposed. The voice capture system and its characteristics are detailed in Sect. 3. In
Sect. 4, the design of the Bluetooth application is mentioned. The conclusions are
set out in Sect. 5. Finally, Sect. 6 is dedicated to the acknowledgments.
2 Proposed System
The main goal has been to develop a device of small size that the victim or potential
victim of gender-based violence would carry at all times. The device is able to detect
a limited number of voice commands through a relatively simple hardware, which
respects the conception of low consumption device. The system will be built on
Bluetooth Low Energy (BLE) technology, which will allow to establish a wireless
communication between device and mobile phone with a low power consumption
thanks to the transmission of small data packets instead of a continuous flow of
information.
The smartphone will perform alert functions via calls and SMS, as well as GPS
location, thus allowing the device designed to be small in size and then easily cam-
ouflaged. It, also, will incorporate a microphone that will allow the smartphone to
perform GPS locator functions and launch calls to the authorities, depending on the
detected voice command. On the other hand, the smartphone will fulfill the functions
imposed under an Android application. The use of some capabilities that, a priori, the
majority of low-mid-range smartphones have (GPS and Bluetooth 4.0) is essential
to reduce the cost and its size to the minimum possible. The entire system described
is outlined in Fig. 2.
Regarding the products focused on the prototyping and development of systems
based on BLE technology, RFduino has been the selected for this work. Specifically,
the RFD22301 chip includes a microcontroller based on an ARM Cortex M0 archi-
tecture (nRF51822) that implements BLE technology. In addition, this platform is
fully compatible with the Arduino IDE environment and with the libraries developed
for Arduino. RFD22301 incorporates a 10 bits analog-to-digital converter, serial
port, a microcontroller operating at 16 MHz, 128 kB of Flash memory and 8 kB
of RAM [14]. To have an easy access to the chip’s ports, the module RFD22102
Fig. 2 Block diagram of the voice command recognition system
“RFduino DIP” has been used [15] along with the RFD22121 “USB Shield.” This
device provides serial conversion-USB necessary to load compiled code on the chip,
thanks to the built-in FTDI [16]. Regarding the development environment, RFduino
is programmable through Arduino IDE in a language that mixes functions of C and
C++ [17]. Arduino IDE incorporates code editor, compiler and serial monitor, being
the only software debugging element available.
With respect to the smartphone, Android, which is based on the Linux kernel,
has been the operating system selected. Today it is present in smartphones, tablets,
smart watches, televisions, and automobiles, among others, being already the most
popular operating system in the world [18].
There are several environments that enable the development of Android appli-
cations in Web languages (HTML5, PHP, CSS3); however, it has been selected
Android Studio (IDE facilitated by Google) because it offers greater possibilities for
optimization and use of the hardware, as well as the use of APIs that facilitate the
implementation of services such as Bluetooth or location in a relatively simple way.
IDE Android Studio incorporates various utilities for programming: code editing,
compilation in real time, debugging, possibility of simulation of mobile devices, and
others.
As an Android device, we have used a Motorola Moto G, with Android Lollipop
5.0. In order to implement the libraries related to BLE technology in an Android
application, it is necessary that the device has API Level 18 or higher, that is, Android
4.3 or higher.
Regarding the BLE protocol just highlight that it is built on three fundamental
layers: application, host, and controller. This work mainly deals with the highest-level
layer, that is, the application layer.
3 Voice Capture System
The voice capture system is shown in Fig. 3. As can be seen, consists of a microphone,
a sound amplifier, and an analog-to-digital converter. Since the voltage variations will
be very small, it will be necessary to pass the electrical signal through an amplifier
Fig. 3 Functional block diagram of the audio capture system
Fig. 4 Audio amplifier with low-pass type characteristic
so that, finally, it is digitized, processed and the result sent through the Bluetooth
module. The amplified acoustic signal will be a signal with a continuous frequency
spectrum between 150 Hz and 4000 Hz, which is the range of frequencies on which
the male and female voice extends [19]. For an adequate sampling of the audio by
the ADC, it is necessary to ensure the Nyquist Theorem.
The amplification structure selected is based on a non-inverting configuration with
Operational Amplifier (OpAmp). In order to comply with the Nyquist criterion and
avoid the possible spectrum overlapping, a capacitor has been included to reduce
the bandwidth of the input signal. The amplifier is shown in Fig. 4, along with the
capacitor Cf added to provide the amplifier with low-pass characteristics. Its value
will be selected following an empirical procedure. The OpAmp used has been the
LM324. It is a single-supply low power OpAmp. The resistors Rcm1 and Rcm2 are
responsible for providing the input signal with a common mode voltage, necessary
due to the 3.3 V single-supply operation of the OpAmp. On the other hand, resistors R1
and R2 set the gain of the circuit according to the expression (1) with R1 implemented
with a potentiometer for a manual adjustment of the gain.
R1
Gain = 1 + (1)
R2
In addition, the C 1 condenser decouples the input signal and the common mode
voltage of the circuit. Finally, the C 2 capacitor allows that the common mode feedback
will be complete. The values of the discrete components are the following: R1 =
218 k, R2 = 2.2 k, Rcm1 = 22 k, Rcm2 = 10 k, C 1 = 4.7 µF, C 2 = 4.7 µF so
the maximum value of the gain will be approximately 100 V/V, i.e., 40 dBs.
The total harmonic distortion of the amplifier has been obtained for a 10 mV
input signals at 1 kHz frequency obtaining a value of 0.3%, more than acceptable
considering that the maximum input signal will be around 2 mV (voltage level for
conversations of great intensity).
4 Design of the Bluetooth Application
The effective sampling frequency of RFduino was determined resulting to be about

13 kHz, which, in the first instance, allows the development of an audio applica-
tion, since the Nyquist Theorem would be respected for signals within the human
conversation frequency spectrum.
Regarding the roles in the BLE network, the smartphone would take a role of
observer whereas for the implemented device the role will be broadcaster.
4.1 Audio Processing
The digital processing of the audio signal through RFduino has been implemented
through the μspeech library [20]. Thanks to this open-source software library, it is
easy implement phoneme recognition functions in RFduino. It is written in C++ and
works by taking a number of audio samples that evaluates and associates with a
phoneme, distinguishing between English phonemes associated with the characters
‘e’, ‘h’, ‘v’, ‘f ’, ‘s’, and ‘I’. This library will be modified to adapt it to the needs of
this work.
4.2 Operation and Modification of the µspeech Library
It is worthwhile to mention that it is not intended to perform a complex digital

audio processing, since this requires advanced hardware (a DSP would be necessary)
and a higher power consumption. The valuable of this method is the possibility of
distinguishing a few voice commands from a simple hardware with a minimum power
consumption allowing the device to operate with a coin cell battery.
The μspeech library works by calculating the complexity of the sampled signal.
To do this, take a number of samples (32 by default) that are stored in an array of
integers. Afterward, discrete derivation operations in absolute value and the discrete
integral of the absolute value of the samples are performed. These operations are
reflected in Eq. (2).

|array[i] − array[i − 1]|
complexity = (2)
|array[i]|
In a first approximation, it can be deduced that signals with a waveform with

important high-frequency contents will have a higher value of complexity, thus being
able to distinguish some phonemes from others. In addition, with the aim of filtering
noise, the library performs a buffer of a 7 size in which a new value of complexity is
stored in each sample collection. Finally, an amplification factor adjusts the weight
of the derivative action against the integral action of the algorithm.
Once the complexity of the sample vector has been calculated, the obtained value
is compared with the tabulated coefficients for each of the phonemes, selecting the
corresponding one. All phonemes are recognized by this algorithm excepting the
phoneme /f/, which is obtained by comparing a new parameter with a tabulated
coefficient.
Finally, in each execution it is possible to calibrate the level of the ambient noise
present through a method that takes 10,000 samples and performs the average, set-
ting an ambient sound level that is subtracted from each of the measures that are
subsequently taken.
μspeech can distinguish 6 phonemes in the English language, which should not
necessarily correspond to phonemes of the Spanish language. Thanks to a developed
script and empirical tests, it was determined that the Spanish phonemes that this
library allows to distinguish were: /i/, /a/, /g/, /s/, and /f/ [21].
4.3 Development of a Script for Voice Command Detection
Considering the environment in which the designed device is going to be used, a

program has been developed that distinguishes three different Spanish voice com-
mands: “socorro,” “informacion,” and “GPS.” These words contain phonemes that
can be distinguished one from the other. The “socorro” command is distinguished
Table 1 Results after a voice command acquisition test

Cf capacitor (pF) Command Hits Errors Hit rate (%) Average rate (%)
15 Socorro 18 2 90 96.67
GPS 20 0 100
Información 20 0 100
100 Socorro 16 4 80 88.33
GPS 20 0 100
150 Socorro 4 16 20 61.67
GPS 20 0 100
by the /s/phoneme, whereas the “informacion” command will be discerned by the

/f/phoneme. The “GPS” command is recognized thanks to the /g/phoneme. To imple-
ment this script, the syllable class—defined in the library μspeech—has been used.
It acts as a “container” of detected phonemes, and an array of characters that allows
us to know the history of phonemes detected in chronological order.
4.4 Frequency Response of the Amplifier
As already mentioned, the choice of the value of capacitor Cf has been made following
a completely empirical procedure. This has consisted of a test in which the three
commands to be detected were repeated 20 times. The results obtained are shown in
Table 1.
It is concluded that the best response is obtained using a 15 pF capacitor in the
feedback path. These tests have been carried out in ideal conditions of loudness
(without ambient noise). The results were very similar with certain levels of ambient
noise.
4.5 Design of a Prototype
For the design of a prototype, the shield of the RFduino RFD22127 has been used [22].
It allows the circuit to be powered through a AAA 1.5 V battery since it implements
a DC/DC boost converter that sets a supply voltage of 3.3 V. Three LEDs have been
also included so they will light up when any of the voice commands were detected.
The estimated current consumption is around 15 mA. In the design of the printed
circuit board (PCB) special attention has been paid to the layout in order to size the
board so that have the dimensions and shape similar with those of the Shields used.
Fig. 5 Photograph of the PCB with components and shields
Fig. 6 Miniaturized circuit and necklace where it can be placed a circuit top view, b bottom view
and c top and bottom part of the necklace
All traces have been arranged in the lower layer. In addition, three capacitors have
been included in the path from the power supply to ground to filter noises. Figure 5
shows the photograph of the PCB with all components and Shields.
As already mentioned, the final goal of this work is that the device can be carried
by a victim of gender-based violence camouflaged in some jewelry accessory. To
achieve this, it is necessary to minimize the size of the designed prototype. Figure 6a,
b shows the top and bottom part of the new prototype implemented, respectively. The
dimensions are 28.6 × 25.5 mm2 which allow it to be placed in a necklace as shown
in Fig. 6c.
4.6 Development of the Android Application
The application has a graphical user interface (GUI) which shows how the phone
scans for advertising packets during a limited period of time, then going to sleep. If
advertising packets are detected, the information will be read to discern if it comes
from the implemented device (through the name) and what is the detected voice
Fig. 7 Screenshots of the Android application a main screen, b “socorro” (SOS) screen,
c “información” (info) screen and d “GPS” screen
command (through the user data included in the package). The functions, depending
on the detected command, are:
– Voice command “socorro”: GPS location via satellite and telephony antennas,
ability to send SMS alerts with this information and to launch a telephone call.
– Voice command “informacion”: collecting the battery level of the device and send-
ing SMS with this information.
– Voice command “GPS”: GPS location and sending SMS with information regard-
ing the coordinates (latitude and longitude).
Figure 7 shows the different screenshots of the Android application depending on
the functions carried out.
Once the SMS is sent to the Control Center, the appropriate steps will be taken in
order to protect the victim from its possible aggressor or simply store the received
information.
5 Conclusions
The prototype of an electronic device to help victims of gender-based violence has

been designed. The device is controlled by voice commands and is capable of sending
messages through a mobile phone to a Control Center. Depending on the nature of
the messages, the Control Center will take the appropriate measures to assist the
victim or simply store the information received. The design has been done paying
special attention to a minimum power consumption so a coin cell battery can be used
and a reduced size so that the device can easily be camouflaged in any accessory of
the victim’s jewelry. In this way, the device can go unnoticed to the aggressor with
which the victim could even to be living together.
Acknowledgements The authors are grateful to the Junta de Extremadura and FEDER (Fondo
Europeo de Desarrollo Regional “Una manera de hacer Europa”), for financial help by project
IB16211.
References
1. United Nations Office on Drugs and Crime, UNODC: Global Study on Homicide, Vienna
(2018)
2. World Health Organization, Department of Reproductive Health and Research, London School
of Hygiene and Tropical Medicine, South African Medical Research Council: Global and
regional estimates of violence against women. Prevalence and health effects of intimate partner
violence and non-partner sexual violence. World Health Organization, Switzerland (2013)
3. Organic Law 1/2004, of December 28, on comprehensive Protection Measures against Gender-
based Violence
4. Organic Law 3/1989, of June 21, updating the Criminal Code, which introduces the criminal
activity of habitual violence in article 425
5. Organic Law 11/2003, of September 29, on Measures on Citizen Security, Domestic Violence
and Social Integration that brings habitual violence to Article 173 of the Criminal Code and
introduces for the first time the “criminal activity of occasional mistreatment” in his article 153
6. Law 27/2003, of July 31, regulating the Order for the Protection of Victims of Domestic
Violence
7. Organic Law 5/2010, of June 22, which modifies the Organic Law 10/1995, of November 23,
of the Penal Code
8. Organic Law 2/2009, of December 11, reform of the Organic Law 4/2000, of January 11, on
rights and freedoms of foreigners in Spain and their social integration
9. Royal Decree-Law 3/2013, of February 22, which modifies the regime of fees in the area of
the Administration of Justice and the system of free legal assistance
10. B.O.E of June 6, 2014, Instrument of ratification of the Convention of the Council of Euro-pa
on prevention and fight against violence against women and domestic violence, done in Istanbul
on May 11, 2011
11. Organic Law 3/2007, of March 22, for the effective equality of women and men
12. Guide to criteria for judicial intervention against gender-based violence, 2013
13. Ministry of Health, Social Services and Equality of the Government of Spain on Gender-
based Violence Homepage, http://estadisticasviolenciagenero.msssi.gob.es/. Last accessed
2018/12/19
14. RFduino (RFD22301) Homepage, http://www.rfduino.com/product/rfd22301-rfduino-ble-
smt/index.html. Last accessed 2015/06/01
15. RFduino (RFD22102) Homepage, http://www.rfduino.com/product/rfd22102-rfduino-dip/
index.html. Last accessed 2015/06/01
16. RFduino (RFD22121) Homepage, http://www.rfduino.com/product/rfd22121-usb-shield-for-
rfduino/index.html. Last accessed 2015/06/01
17. Arduino programming language Homepage, http://www.arduino.cc/en/pmwiki.php?n=
Reference/HomePage. Last accessed 2015/06/01
18. Android Homepage, https://www.android.com/. Last accessed 2018/012/26
19. Nortel Networks, Voice Fundamentals, Reference Guide (2014)
20. GitHub directory of the μspeech library, https://github.com/arjo129/uspeech/releases. Last
accessed 2015/06/13
21. Association of phonemes and letters of Spanish, Homepage, http://www.livingspanish.com/
correspondencia-fonetica-grafia.htm. Last accessed 2015/06/13
22. RFduino (RFD22127) Homepage, http://www.rfduino.com/product/rfd22127-single-aaa-
battery-shield-for-rfduino/l. Last accessed 2015/06/13
Part XI
Analysis and Signal Processing
Analysis and Evaluation of the
Positioning of Autonomous Underwater
Vehicles Using Acoustic Signals
Enrique V. Carrera and Manolo Paredes
Abstract Autonomous underwater vehicles (AUVs) present as much potential as

unmanned aerial vehicles, or drones, for being used in military missions. Major
defense-oriented applications of AUVs will be intelligence, surveillance, and recon-
naissance tasks, besides payload delivery and communication support. Nevertheless,
AUVs have several constraints that make it difficult to rely completely on their
autonomous functions. Particularly, AUVs do not have access to accurate position-
ing mechanisms like GPS. Thus, expensive inertial navigation techniques have been
developed for these vehicles in the last decades. However, small and low-cost AUVs
require newer and simpler positioning techniques that can be adapted to specific
military missions. Based on that, this work analyzes and evaluates common trilat-
eration techniques based on the propagation of acoustic signals in the water for a
precise underwater positioning. Although the results show that the method known as
time difference of arrival (TDoA) presents good accuracy, its application in military
missions is still limited due to the large number of noise sources required; therefore,
several future works in this direction are additionally proposed.
Keywords Autonomous underwater vehicles · Underwater positioning ·

Trilateration techniques · Underwater sound propagation
E. V. Carrera (B) · M. Paredes

Departamento de Eléctrica, Electrónica y Telecomunicaciones, Universidad de las
Fuerzas Armadas ESPE, Sangolquí 171103, Ecuador
e-mail: evcarrera@espe.edu.ec
M. Paredes
e-mail: dmparedes@espe.edu.ec
M. Paredes
Centro de Investigación de Aplicaciones Militares CICTE, Universidad de las
Fuerzas Armadas ESPE, Sangolquí 171103, Ecuador

https://doi.org/10.1007/978-981-13-9155-2_33
412 E. V. Carrera and M. Paredes
1 Introduction
Unmanned aerial vehicles, or drones, have rapidly grown in popularity during the
past years. They have broken through rigid barriers in traditional industries which
otherwise seemed impenetrable by similar technological innovations. Today, drones
are key elements in activities like aerial photography, search and rescue operations,
precision agriculture, and express shipping and delivery. In fact, even though the
military use of drones has a large history [7], military applications are currently the
main service of this type of vehicles [3, 16]. It should also be noted that unmanned
vehicles have been an important catapult for the battlefield industry and security
operations, improving the fight against organized and transnational crime, and other
threats that attempt to violate or alter citizen security.
However, illegal activities are also being developed in coastal profiles and off-
shore, and drones are not able to fully cover the needs existing in underwater environ-
ments. For instance, security operations require support to combat drug trafficking
in submersible vehicles, guerrillas next to zones covered by water, criminal activities
through rivers and estuaries, etc. It is then essential to use the benefits and capabili-
ties of unmanned vehicles for these operations in order to support military activities
in those areas. Thus, the need of developing the so-called unmanned underwater
vehicles arises. These vehicles are divided into remotely operated vehicles (ROVs),
which are controlled by a remote human operator, and autonomous underwater vehi-
cles (AUVs), which operate with no direct human intervention. AUVs are the most
complex and expensive vehicles since they completely rely on autonomous func-
tions, and they have gradually evolved offering capabilities in some missions and
applications that no other platforms can offer [1]. As a matter of fact, due to the tech-
nological innovations in the past years, small and inexpensive AUVs are becoming
feasible and profitable.
Although AUV field is still an emerging sector, AUVs will be part of many aca-
demic, commercial, and military applications [5]. In fact, military usage of AUVs
is a sector of increasing demand, due to their ability of conducting missions that no
other equipment can perform [8]. However, current existing alternatives face sev-
eral challenges like position uncertainty, limited and noisy communication, bounded
time autonomy, overall reliability, among others. In addition, the usage of AUVs
into military operations deserves a deeper analysis, since it is necessary to exploit
the principles of the war and carry out military missions effectively. Hence, it has
been promoted the creation, strengthening, and improvement of R+D+i processes
to generate technological solutions that energize and optimize the use of all kind of
resources in military operations.
Following this line of reasoning, this work seeks to solve a main AUV problem:
underwater positioning. This is a challenging task since the GPS network is not
available to submersed devices, and considering that radio signals do not propagate
very far in water, AUV positioning and navigation are still areas of active develop-
ment [14]. When georeferenced navigation is necessary, current AUVs use internal
navigation systems, such as inertial guidance systems that measure and integrate
Analysis and Evaluation of the Positioning of Autonomous Underwater … 413
acceleration or a Doppler velocity log system. A controller then integrates velocity

to dead-reckoned positions, which are generally updated via GPS. GPS update rep-
resents a risk to underwater missions since the AUV must emerge jeopardizing its
covertness. Accurate navigation without GPS has been also proposed using bottom-
terrain maps, which must be developed in advance. In test conditions, AUVs might
even navigate using a set of acoustic beacons, forming an LBL network [12].
Based on the above, this paper analyzes and evaluates common trilateration tech-
niques used for Cartesian positioning, but adapted to underwater scenarios through
the usage of acoustic signals [6, 15]. Specifically, a simulator of an AUV was imple-
mented to compare received signal strength indication (RSSI), time of arrival (ToA),
and time difference of arrival (TDoA) techniques for the positioning of a vehicle
inside a large volume of water. The results show that TDoA presents the best accu-
racy, but the number of noise sources is relatively high for a 3D positioning. Thus,
some future works are also proposed for the positioning of AUVs, especially using
microphone arrays (i.e., hydrophone arrays).
The rest of this paper is organized as follows. Section 2 introduces AUVs, sound
propagation in water, and trilateration techniques. Section 3 presents the simulation
of three common trilateration techniques for AUV positioning. The main results and
their implications are analyzed in Sect. 4. Finally, Sect. 5 concludes this paper and
discusses some future works.
2 AUV Positioning
An important step for the development of powerful AUVs is to know the vehicle’s
position adequately, using any available tool or technology. With this requirement in
mind, the basics of AUVs, the propagation of acoustic signals in water, and common
trilateration techniques are discussed below.
2.1 AUV Basics
Many different AUVs have been designed in the past five decades, ranging in size
from portable AUVs to very large vehicles. Large AUVs have advantages in terms of
endurance and sensor payload capacity, but smaller AUVs benefit significantly from
lower logistics. Today, the AUV market is evolving, and new designs are following
commercial requirements rather than being purely experimental [2].
AUVs must basically control themselves while accomplishing a predefined task
[18]. For that, the following list represents the main technologies that have been
addressed over the last years [1, 5]:
– Autonomy. Topics such as mission planning, perception and situation assessment

and intelligent systems have been studied. All these are hard problems, and the
issue of autonomy still remains unsolved.
– Energy. Endurance of AUVs has increased from a few hours to tens of hours.
Some systems now contemplate missions of days or months. However, extended
endurance is at the expense of sensing capability and very limited transit speeds.
Better battery-powered systems are been still researched.
– Navigation. Early AUVs relied on dead reckoning for their navigation. Inertial
navigation systems were available for expensive AUVs, while acoustic transponder
navigation provided good accuracy but at a significant logistics cost. In recent
years, AUVs have taken advantage of GPS when the vehicle surfaces. However,
there is strong interest in being able to navigate relative to the environment within
which the system exists. Without a doubt, an effective and robust positioning
system will provide a significant increment in AUV capabilities.
– Sensors/3D imaging. The paradigm of sensor radio integration has changed the
way AUVs work today. New sensors are smarter, lower power, highly reliable,
smaller in size, etc. In fact, much work continues now on the development of
imaging systems with higher resolutions and artificial vision support.
– Communications. In the past years, there have been significant advances in acous-
tic communications such that relatively low error rate communications are possible
over ranges of kilometers at bit rates of a few kbps.
In summary, there are still technological problems to be solved. The economic

viability of the technology also has to be demonstrated, and AUVs must be proven
in an operational regime in order to be finally applied in military missions.
2.2 Sound Propagation in Water
In the water, acoustic signals propagate faster than in the air and their absorption
at low frequencies is weak, making the water a good conductor of acoustic waves
[10]. However, speed propagation of acoustic signals in the water depends on several
parameters like salinity, depth, and temperature. Del Grosso’s equation, an up-to-date
alternative to the UNESCO algorithm, determines the existing relationship among
all these parameters [19]. The base velocity considered by Del Grosso’s equation is
1402.4 m/s, and instead of depending on the depth, this equation uses pressure as
an input parameter. Obviously, the pressure depends on the depth and geographical
latitude.
Del Grosso’s equation was derived from fitting it to experimental data, and it
has an associated uncertainty in its prediction of sound speed. There are also other
proposed equations and algorithms, but the choice of a particular model depends
on the accuracy and precision acceptable for the application in which it is being
employed. In any case, the range of validity for Del Grosso’s equation is temperature
between 0 and 30 ◦ C, salinity between 30 and 40 parts per thousand, and pressure
between 0 and 1000 kg/cm2 .
The absorption of acoustic signals in the water also depends on some properties
such as temperature, salinity, and acidity of the water, as well as the frequency of the
acoustic signal [9]. This absorption generates some losses in the transmitted energy
from the source to the receiver. However, the major contribution to transmission
losses is the spreading of the acoustic wave as it propagates away from the source. It
is well known that sound spreads such that its intensity decreases exponentially with
distance.
2.3 Trilateration Techniques
Trilateration is the process of determining absolute or relative locations of points

by the measurement of distances to known locations [17]. When it is possible to
establish the intensity of some transmitted signals, the most common trilateration
techniques are RSSI, ToA, and TDoA. However, if it is also possible to establish the
angle of arrival of the transmitted signals, the trilateration technique knows as angle
of arrival (AoA) might be employed.
Any of these techniques can be implemented by the vehicle (i.e., client-based
positioning) or the surrounding infrastructure (i.e., infrastructure-based positioning)
[4]. In the case of a client-based positioning, the surrounding infrastructure includes
N acoustic or radio-frequency (RF) sources transmitting some specific signals, and
the vehicle listens to those transmissions. In the case of an infrastructure-based
positioning, the vehicle emits an acoustic or RF signal and N transducers in the
surrounding infrastructure listen to that transmission.
AUVs will prefer client-based positioning techniques, since the autonomy of these
vehicles is increased and less energy is wasted listening rather than transmitting
acoustic signals. Because of that, the main trilateration techniques are described as
a client-based implementation, although an infrastructure-based positioning is also
possible for each of the variants.
RSSI. This technique measures the strength of the signals received from the sur-
rounding acoustic sources at known locations. Since the intensity of a signal is
attenuated by the travelled distance, the distance of each source can be modeled as
a function of its intensity. However, this modeling is hard to achieve accurately due
to multipath interference, path loss, Doppler distortion, and other fading effects. An
important advantage of this technique is that no time synchronization is required
between sources and receivers.
ToA. In this technique, the distance is computed as a function of the time required
by the signal to travel from the source to the receiver. Since the acoustic signals
maintain almost constant speed, the accuracy of this trilateration technique is very
good. However, the main problem here is to keep synchronized sources and receivers
in order to determine the time of flight of each transmitted signal. Geometrically, both
Fig. 1 Geometry employed

by ToA/RSSI and TDoA
using three sources
RSSI and ToA describe circumferences centered in the sources and the position of
the vehicle is determined by the intersection of those circumferences as explained in
Fig. 1.
TDoA. This method determines the vehicle position measuring the time differences
between every pair of received signals. Geometrically, the differences in time gen-
erate hyperbolas that intersect in the position of the vehicle as shown in Fig. 1. This
technique requires that only the acoustic sources are synchronized among them,
being clearly advantageous an infrastructure-based TDoA positioning technique as
there is only one source.
3 Underwater Positioning Simulation
In order to analyze and evaluate different positioning techniques for AUVs, an under-
water vehicle simulator has been developed. The technical details of the simulator
and its default configuration are presented in the following subsections.
3.1 Simulation Platform
The simulator was implemented in MATLAB® and includes a very complete model
of the sound propagation in water. The acoustic sources can be configured in space
and number with specific directivity, while the hydrophones are always omnidirec-
tional. The speed propagation is modeled according to Del Grosso’s equation, and
the absorption of the sound is modeled by the Ainslie and McColm algorithm. After
determining the existing response to the discrete unit impulse between any source
and the receivers, the response to a discrete pulse is computed. Obviously, multipath
interference and path loss are considered in these calculations in order to obtain a
realistic pulse response.
Using the corresponding responses to the pulses emitted by the acoustic sources,
the simulator is able to compare RSSI, ToA, and TDoA algorithms using both
client-based and infrastructure-based trilateration techniques. The computed posi-
tions through these methods can be compared for any specific AUV position, or the
average error for a fraction of the whole volume can be determined. The fraction
of volume used in this last case is the centered volume limited by a scaling factor
between 0 and 1. Figure 2 allows us to understand better the volume corresponding
to a specific r factor.
In addition, the simulator implements an intuitive graphical interface shown in
Fig. 3. Through the interface, the final user can configure the size of the water con-
tainer, its latitude, the water temperature and salinity, and the reflection coefficients of
every wall in the container. Moreover, the sampling frequency for digital processing
and the duration of the pulses emitted by the sources are also configurable.
Fig. 2 Volume determined by a scaling factor r
Fig. 3 Graphical interface of the implemented simulator

3.2 Default Configuration
The default configuration of the simulator considers the parameters found in a lake
at the Andean region of Ecuador. Basically, we use a latitude of 0.216667◦ , a tem-
perature of 4 ◦ C, and a salinity of 10 ppm. The reflection coefficients are 1.0 for the
surface and 0.46 for the other walls of the water container. The default size of the
water container is 5 × 5 × 3 m, and the average errors are obtained examining points
separated 0.2 m in each direction along the volume scaled by a factor r (see Fig. 2).
Lastly, the sampling frequency used in the computation of the unit impulse response
is 160 kHz and the duration of the pulses is 500 ms.
4 Results
The main results obtained by the simulator are presented below. A parameter space
study is also included before discussing the main implications of these values.
4.1 Accuracy Results
The default configurations for comparing the three main trilateration techniques are:
five noise sources, temperature of 10 ◦ C, and a water container of 5 × 5 × 3 m. The
results shown in Table 1 confirm that TDoA is the most accurate technique. Although
ToA has similar results, ToA requires additional time synchronization among acoustic
sources and receivers. Remember that TDoA only requires that the acoustic sources
are synchronized.
Other import observation from Table 1 is the importance of keeping away the
vehicle from the acoustic sources. When the operational zone of the AUV is contained
inside an r factor of 0.8 (i.e., approximately half of the total volume), the accuracy
is lower than 10 cm. Furthermore, if the operational zone is inside an r factor of 0.6,
the accuracy could be as low as 1 cm.
Results for an infrastructure-based positioning instead of a client-based position-
ing also show the same trend.
Table 1 Average position accuracy (in cm) for different trilateration techniques
Technique Scaling factor r (volume)
1.0 (100%) 0.8 (51%) 0.6 (22%)
RSSI 326.2 29.0 3.4
ToA 163.0 9.6 1.1
TDoA 162.8 9.6 1.1
Table 2 Average position accuracy (in cm) for TDoA using different configurations
Temperature (◦ C) Scaling factor r (volume)
1.0 (100%) 0.8 (51%) 0.6 (22%)
4 130.8 5.6 1.0
10 162.8 9.6 1.1
20 169.5 12.1 1.1
Number of sources Scaling factor r (volume)
1.0 (100%) 0.8 (51%) 0.6 (22%)
4 114.0 58.3 25.7
5 162.8 9.6 1.1
6 25.7 3.2 0.9
Size of the container (m) Scaling factor r (volume)
1.0 (100%) 0.8 (51%) 0.6 (22%)
5×5×3 162.8 9.6 1.1
10 × 10 × 6 177.0 6.0 1.4
20 × 20 × 12 107.6 3.3 1.4
4.2 Parameter Space Study
In order to determine how some parameters included in the simulator affect the
accuracy of the TDoA technique, Table 2 presents accuracy values for different tem-
peratures, number of noise sources, and size of the water container.
We can see that low temperatures slightly improve accuracy, even for high val-
ues of r . On the other hand, the number of acoustic sources significantly affects
the accuracy of TDoA. Basically, five acoustic sources are the minimum number
of allowed sources for getting an acceptable accuracy. Working with six acoustic
sources, instead of five, still causes a noticeable improvement.
In addition, although the initial configuration of the water container might seem
a bit small, we can see that accuracy values are similar for larger configurations of
the container. In the case of a container of 20 × 20 × 12 m, the positioning accuracy
is lower than 5 cm for an r factor of 0.8.
4.3 Discussion
These results definitively place TDoA as the best trilateration technique without using
angles of arrival. Results are pretty good, especially for low temperatures, using at
least five noise sources, and big containers of water. In summary, an accuracy of a
few centimeters can be easily achieved by TDoA.
The main problem of TDoA is the high number of noise sources required in
the surrounding infrastructure. This problem is enlarged if military applications are
considered, since enemy forces can easily be aware of the presence of such devices
in the battlefield.
Despite that, we have already started the hardware implementation of the TDoA
technique to validate previous results in the field. In order to detect the starting point
of the acoustic pulses, a variation of Goertzel’s algorithm [11] has been implemented
in a Raspberry Pi motherboard. At this moment, time difference errors lower than
0.1 ms are achieved with this hardware platform.
5 Conclusions
This work demonstrates that trilateration techniques are good enough to provide a
positioning accuracy below 10 cm in a broad range of configurations. The imple-
mented simulator considers a complex model for the sound propagation in the water,
providing acceptable trends in the comparison of trilateration techniques based on
signal intensities. We expect that this type of technology can be used soon in military
application after minimizing the number of noise sources.
For the reduction of noise sources, as a future work, we plan to extend the simulator
for comparing the AoA technique to the other trilateration techniques. In fact, we
would like to study microphone/hydrophone arrays [13] to combine AoA and TDoA
techniques in order to reduce significantly the number of noise sources. Finally, the
hardware implementation of the TDoA technique is almost finished and we expect
to make tests in the field using a commercial ROV in a lake of Ecuador to validate
the results presented along with this document.
Acknowledgements This work was partially supported by the Universidad de las Fuerzas Armadas
ESPE under Research Grant PREDU-2016-005.
References
1. Allard, Y., Shahbazian, E.: Unmanned underwater vehicle (UUV) information study. Technical
report, DRDC-RDDC-2014-C290, OODA Technologies Inc., Montreal, Canada (2014)
2. Antonelli, G.: Underwater Robots, 4th edn. Springer, Berlin (2018)
3. Baek, H., Lim, J.: Design of future UAV-relay tactical data link for reliable UAV control and
situational awareness. IEEE Commun. Magaz. 56(10), 144–150 (2018). https://doi.org/10.
1109/MCOM.2018.1700259
4. Bensky, A.: Wireless positioning technologies and applications. Artech House (2016)
5. Button, R.W., Kamp, J., Curtin, T.B., Dryden, J.: A survey of missions for unmanned undersea
vehicles. RAND National Defense Research Institute, Santa Monica, CA (2009)
6. Carrera, E.V., Perez, M.S.: Event localization in wireless sensor networks. In: 2014 IEEE
Central America and Panama Convention (CONCAPAN XXXIV), pp. 1–6. IEEE, Panama,
Panama (11 2014)
7. Cook, K.L.B.: The silent force multiplier: The history and role of UAVs in warfare. In: 2007
IEEE Aerospace Conference. pp. 1–7 (March, 2007)
8. Damian, R.G., Jula, N., Paturca, S.V.: Autonomous underwater vehicles - achievements and
current trends. Sci. Bull. Nav. Acad. 21(1), 85–89 (2018)
9. Doonan, I.J., Coombs, R.F., McClatchie, S.: The absorption of sound in seawater in relation to
the estimation of deep-water fish biomass. ICES J. Mar. Sci. 60(5), 1047–1055 (2003)
10. Etter, P.C.: Underwater Acoustic Modeling and Simulation. CRC Press (2018)
11. Jiang, J., Brewer, R., Jakubowski, R., Tan, L.: Development of a piano frequency detect-
ing system using the goertzel algorithm. In: 2018 IEEE International Conference on Elec-
tro/Information Technology (EIT), pp. 0346–0349. IEEE (2018)
12. Khan, R., Ercan, M.F., Metarsit, L., Le, A.N.S., Lim, K.V., Tan, W.T., Ang, J.L.: Underwater
navigation using maneuverable beacons for localization. In: OCEANS 2016 MTS/IEEE, pp. 1–
5. IEEE, Monterey, CA (2016)
13. Paredes, D., Apolinario, J.: Shooter localization using microphone arrays on elevated platforms.
In: 2014 IEEE Central America and Panama Convention (CONCAPAN XXXIV), pp. 1–6.
IEEE, Panama, Panama (2014)
14. Paull, L., Saeedi, S., Seto, M., Li, H.: AUV navigation and localization: A review. IEEE J.
Ocean. Eng. 39(1), 131–149 (2014)
15. Perez, M.S., Carrera, E.V.: Acoustic event localization on an arduino-based wireless sensor
network. In: 2014 IEEE Latin-America Conference on Communications (LATINCOM), pp. 1–
6. IEEE, Cartagena, Colombia (2014)
16. Solodov, A., Williams, A., Al Hanaei, S., Goddard, B.: Analyzing the threat of unmanned aerial
vehicles (UAV) to nuclear facilities. Secur. J. 31(1), 305–324 (2018)
17. Stojanović, D., Stojanović, N.: Indoor localization and tracking: methods, technologies and
research challenges. Facta Univ. Ser. Autom. Control. Robot. 13(1), 57–72 (2014)
18. Wadoo, S., Kachroo, P.: Autonomous underwater vehicles: modeling, control design and sim-
ulation. CRC Press (2016)
19. Wong, G.S., Zhu, S.m.: Speed of sound in seawater as a function of salinity, temperature, and
pressure. J. Acoust. Soc. Am. 97(3), 1732–1736 (1995)
Part XII
Chemical, Biological and Nuclear Defense
Evaluation of 3D Printing Parameters
on the Electrochemical Performance
of Conductive Polymeric Components
for Chemical Warfare Agent Sensing
Joseane R. Barbosa, Pedro H. O. Amorim, Mariana C. de O. Gonçalves,

Rafael M. Dornellas, Robson P. Pereira and Felipe S. Semaan
Abstract Aiming at the use of 3D printing as a tool in the design of electronic com-
ponents and sensors, the present work has developed a preliminary evaluation of the
influence of printing conditions on the electrochemical behavior of PLA and com-
mercial carbon conductor-based composites. For this, characterization experiments
were carried out by means of infrared spectroscopy and thermal analysis, pointing
to a clear variability of behavior of different types of PLA, as well as the influence
of this variation on the thermal behavior of the composites, which consequently
influences the processes of 3D printing. Experiments allowed a speculation of the
conductor load present in the conductive filament (about 20.5%), although it was not
possible to define its category. The print modes, in turn, proved to be a decisive fac-
tor for the electroanalytical behavior in front of a redox probe, being responsible for
improving or worsening the resolution, or favoring the exchange in a given direction.
In all cases, the relations between currents (cathodic and anodic) were acceptable,
although the loss of resolution, translated in the form of spacing between peaks, has
greatly contaminated the calculation of electroactive areas in some cases. Clearly, an
optimization of the printing process, as well as a better control in the particle size,
will allow obtaining high electroactive areas, with adequate resolution, even with low
driver loads, which shows a promising line currently in development in our Group.
Keywords Conductive 3D printed component · Surface effects · Electrochemical

sensing
J. R. Barbosa · P. H. O. Amorim · M. C. de O. Gonçalves · R. M. Dornellas · F. S. Semaan (B)

Analytical Chemistry Department, Fluminense Federal University, Niteroi RJ, Brazil
e-mail: felipesemaan@id.uff.br
R. P. Pereira · F. S. Semaan
Military Institute of Engineering, Rio de Janeiro RJ, Brazil

https://doi.org/10.1007/978-981-13-9155-2_34
426 J. R. Barbosa et al.
1 Introduction
Among the current trends in electroanalytical chemistry, the development of cus-

tomized sensors for lab-free applications has received special attention. Aiming to
combine sensitivity, selectivity and velocity with small sample amounts, such sen-
sors are highlighted in the literature as a promising technology among other detection
systems for defense proposals, due to reduction of the sample-operator contact [1, 2].
In this regard, some reports summarize current efforts performed in order to optimize
analytical capabilities of in locu analysis. Various examples of electrodes have been
described in the literature for defense-related applications, and the next section lists
some of these works.
Ismail and Matsuda [3] have developed a procedure for organophosphate chemical
warfare agent determination using a 3-cyanopropyltriethoxysilane film by cyclic
voltammetry, reaching linearity in the range of 0.5–10 pg mL−1 , with a detection
limit of 0.25 pg mL−1 .
Mahyari [4] performed amperometric determinations of thiodiglycol using elec-
trodes based on a mixture of nano-sheets of graphene, silver nanoparticles and ionic
liquid (octylpyridinium hexafluorophosphate), detecting a pair of well-defined redox
peaks in a range of 10–3700 µmol L−1 , with detection limit of 6 µmol L−1 .
Sikarwar et al. [5] analyzed thiodiglycol stimulant using the xanthine oxidase as
electrode modifier onto a gold substrate. The linear response in cyclic voltammetry
was obtained in the range between 1.94 and 9.70 mmol L−1 .
Elliot et al. [6] developed a square wave voltammetric procedure for the quantifi-
cation of pyocyanin using a transparent carbon microelectrode (bare or modified by
gold nanoparticles).
Hondred et al. [7] prepared printed graphene electrodes via non-inkjet lithography.
These electrodes were modified by platinum nanoparticles to improve electrical con-
ductivity. The biosensor was able to rapidly measure an insecticide with 3 nmol L−1
detection limit.
Wang et al. [8] initiated a different approach while coupling a detection platform
to popular smartphones by a plug-able to perform the electrochemical measurement
for the determination of low nitrate concentrations in water, being a very reliable
proof of concept for lab-free applications.
Shim et al. [9] also used a portable potentiostat, performing square wave voltam-
metry measurements with electrochemical sensors, exchanging all data via Bluetooth
to a smartphone. Analytical signals for the observed peaks were compared to those
obtained on a lab-based commercial potentiostat, without significative difference.
Fan et al. [10] tested nanocomposites synthesized by graphene and gold nanopar-
ticles as customized agent for the determination of some specific enolases using
a microfluid analytical paper-based device controlled by smartphone. Such system
was used under differential pulse voltammetric conditions, allowing a good linear
relationship ranging from 1 to 500 ng mL−1 with a detection limit of 10 pg mL−1 .
Evaluation of 3D Printing Parameters … 427
Kanchi et al. [11] and Ainla et al. [12] described different characterization projects
for a universal wireless electrochemical system in a wide range of voltammetric
techniques, with compatible results from all assessed systems.
Colozza et al. [13] reported a new paper-based electrochemical sensor for detec-
tion of sulfur mustards. The working electrode was modified with a nanocomposite
consisting of Prussian blue and carbon black nanoparticles.
From these examples, it is easy to conclude that use of smartphones and other
controlling systems is no longer a challenging task but, on the other hand, customizing
sensors is still a decisive step for a reliable application. It becomes important to note
that many reports describe the use of carbon-based substrates, which is not only
cheap and easy to manipulate, but also versatile in terms of chemical modification.
Carbon composites have numerous advantages over conventional substrates such
as easy surface renewal (allowing the removal of adsorbates), fabrication in different
shapes and sizes, application in a wide range of pH and working potential, possi-
bility of chemical modification on the surface of the electrode or in the bulk itself,
low production cost, among others [14–20]. Among many available binders for 3D
printing, poly(lactic acid) (PLA) stands for a promising substrate, being a biodegrad-
able aliphatic polyester with melting point between 130 and 180 °C, derived from
renewable resources such as corn, rice, cassava roots and sugarcane. Among its char-
acteristics, PLA has a high mechanical resistance that promoted the study of the same
in the use of several materials [14, 21–23].
Various forms of carbon-based particles have been used as conductive phase
in composite materials suitable for electrochemical sensors [20]. Among these,
graphite [15], vitreous carbon particles [24], graphenes [25], nanotubes [26] and
fullerenes [27], among others, have shown promising properties. Graphene consists
of a monoatomic layer formed by carbon atoms with sp2 hybridization, with a sur-
face area that is almost twice as large as that of single-walled carbon nanotubes [28],
and with a long p-conjugated system, the electrons are confined in two dimensions,
which gives the graphene remarkable properties such as high electrical and thermal
conductivity, good transparency, good mechanical strength, inherent flexibility and
huge specific surface area [29–31].
3D printing is a technology that allows the fabrication of solid 3D objects, with
high precision, and with a great deal of flexibility over the adopted design, from a
digital file. The material used for printing is considered low cost, and the produc-
tion process generates minimal waste. There is a commercial drawback, related to
fabrication of conductive parts by 3D printing, using carbon black, graphene and nan-
otubes, which reflects the higher cost compared to pure non-conductive filaments.
One important point is the lack of information about the load (%) and nature of the
conductive phase employed, since a significant part of the producers and suppliers
do not declare such information, offering the product on the market as a “conductive
filament” without further technical details.
Guo et al. [32] presented in their article electrodes fabricated from PLA nanocom-
posite and multilayer nanotubes by 3D printing with a helical structure. In addition
to high electrical conductivity, this system has good rigidity, excellent sensitivity and
selectivity.
Vernadou et al. [33], in his paper on 3D-printed graphene electrodes, deal with
the use of graphene printed with pyramidal geometry using a double-extruded 3D
FDM type printer for application on lithium-ion batteries.
Palenzuela et al. [34] used graphene-PLA composite electrodes imprinted in ring
and disk form in a 3D process of melting and deposition molding and, after charac-
terization by cyclic voltammetry and scanning microscopy, applied in the detection
of picric acid and ascorbic acid.
In spite of the clear application potential and the variety of commercial filaments
available, it is interesting to consider that few previous works take into account the
printing conditions or describe them in detail, although the interfacial properties
obtained are directly affected by the relationship between nature, properties, charge
and dimensions of the dispersed conductive phase, extrusion and printing conditions.
In such context, the present work aims to exploit the influence of printing conditions
over the electrochemical performance of carbon-PLA sensors as prototypes for cus-
tomized sensor strips for lab-free use.
2 Materials and Methods
2.1 3D Printing of Prototypes
Prototypes were prepared by direct printing, using a Voolt 3D GI3 3D printer, with
Slic3r software. These test bodies/work electrodes with dimensions (2.8 × 1.5 ×
0.1 cm) (length, width and thickness) were printed in horizontal and vertical modes
and with Linear type interfacial geometries, Helbert, Archiemedean, Octagram and
concentric, with filler density of 10 and 100%.
Filaments of PLA (FPLAP), and conductive PLA (FPLAC), supplied by 3D Lab®
(Brazil), Proto-Pasta® (USA), respectively, both with a diameter of 1.75 mm, were
used as feeder filaments. Table and nozzle temperatures were adjusted following
properties described by the manufacturers and modes suggested by the printer default
(nozzle temperature between 195 and 225 °C, bed temperature circa 60 °C).
2.2 Thermal Characterization of the Material
The thermal behavior was evaluated by thermogravimetric analysis (TGA and DTG),
using a Shimadzu thermogravimetric analyzer model TGA-60, sample mass about
20 mg of pre-sampled sample in alumina samples and under heating at 10 °C min−1 ,
at room temperature at 900 °C, under a dynamic atmosphere N2 inlet at a flow
rate of 40 mL min−1 . Further additional information was obtained by means of
differential scanning calorimetry (DSC), using a Shimadzu differential scanning
calorimeter model DSC-60, between room temperature and the thermal stability
limit temperature of each evaluated material, under a heating rate of 10 °C min−1 in

heating-cooling-heating mode, under N2 atmosphere.
2.3 Electroactive Surface Response and Geometries
Voltammetric cells printed directly on ABS with diameter 2.2 and 2 cm in height
were used, with total volume of 5 mL in electrolyte solution. The working electrode
was positioned at the bottom, in a window allowing an exposure of 0.385 cm2 in the
geometric area of the sensor, and the sealing was performed with the aid of O’ring
and acetic curing silicone glue.
The electroactive areas were measured by a system of three electrodes
(Ag|AgClsat ) as reference, surgical steel needle as against electrode and the printed
working electrodes to be tested) carefully inserted into the voltammetric cell already
described, using a potentiostat, Ivium® CompactStat model (Ivium Technologies,
Netherlands).
Cyclic voltammetry (CV) in potassium hexacyanoferrate (5 mmol L−1 ) and
0.5 mol L−1 potassium chloride was used, and the Randles–Sevcik ratio, previ-
ously presented and defined, was used for calculations. Three successive scans were
performed in the potential range from −1.0 to 1.5 V, under different scan rates (10,
20, 30, 50, 75 and 100 mV s−1 ), as usual.
3 Results and Discussions
3.1 3D Printing of Prototypes
Conductive FPLAC blades were obtained in the dimensions previously described by

the FDM process employing different conditions. Such blades were suitably posi-
tioned to the bottom of BIA cell in ABS and sealed with silicone glue, having a fixed
and calculated exposed area. Due to printing parameters, the surface is more rough-
ened or regular, which must be evaluated, additionally, based on the dimensions of
the conductive particles.
3.2 Thermal Characterization of the Material
Considering the sample as an electroactive composite, it is possible to infer that the

homogeneity of charge dispersion (%) of the conductive phase, as well as particle
size and thermal conductivity, will directly influence the heat conduction process
and consequently in the relevant thermal events observable. In general, information
about the nature and loading of the conductive phase in these filaments are not pro-
vided, which points to a clear need for prior characterization by thermal techniques,
aiming to evaluate not only the thermal decomposition profile, but possible faults in
the final components obtained by the formation of conductor clumps or regions of
heterogeneity in the parts as a consequence of large differences in thermal behavior
between insulating (PLA) and conductive (undeclared) phases.
Both materials have different thermal stabilities, the FPLAC being more thermally
resistant, since the thermal decomposition starts at 312.12 °C (Tonset ), about 100 °C
higher compared to FPLAP. Another interesting aspect is the third mass loss in
FPLAC (25.4%), compared to the event in the same temperature range for FPLAP
(4.9%), indicating a percentage of conductive phase of about 20.4% and in agreement
with the observed temperatures for carbon-based phases in several reports [21, 35].
Certainly, such an approach is semi-quantitative, since the content is not stated,
as well as the nature of the driver, besides the fact that these are different PLA
structures. This behavior suggests that the conductor filler dispersed in PLA matrix
may increase the thermal stability of the material as a function of the colligative
properties, making it more resistant. It was also possible to note in the DTG two
small mass loss processes of FPLAP, between 84.94 and 109.47 °C and between
217.80 and 248.69 °C, the latter shortly before thermal decomposition, which may
suggest a minimum presence of moisture and additives used in the manufacture
of the filaments, respectively. Thermal decomposition occurs between 248.69 and
338.51 °C, where there is a large mass loss, followed by another small loss between
559.2 and 900.0 °C, with no residue in the crucible at the end of the test, as reported
in the literature [36].
Differential scanning calorimetry (DSC) for the FPLAP and FPLAC samples was
carried out by heating-cooling-heating mode [37] between room temperatures ~25
and 220 °C, based on previous results from TG curves. The percentage of crys-
tallinity was obtained through Eq. 1, where Hc is the enthalpy of crystallization in
(J/g), Hm is the enthalpy of fusion, being obtained by the area of the peak in the
DSC curve, and Hm 100 is the melting enthalpy of 100% crystalline PLA, which
according to the literature is given by 93.7 J/g [38, 39].
|Hm − Hc |
Xc (%) = × 100 (1)
H 100%
The PLA has a characteristic heating profile, with three thermal events—a
glass transition around 60 °C, crystallization around 100 °C and melting around
160 °C—and can vary between 130 and 180 °C. Comparing the first and second
heating cycles, it was observed that the FPLAC had its DSC profile more affected
by the cooling than the FPLAP; however, the glass transition temperatures (Tg ) did
not change significantly. In both cases, the glass transition is followed by a small
endothermic peak, which suggests some degree of physical aging of the materials,
since the polymers have undergone previous processing when the filament is made.
The crystallization process characterized by the exothermic peak at 101.96 °C in
the first heating for the FPLAC does not occur after the cooling process in addition
to a substantial decrease of the melt peak at 150.96 °C, suggesting that the solid upon
cooling did not form a more organized or crystalline structure due to the presence
of the conductive filler. It was observed that FPLAP presented higher crystallinity
(13.97%) than FPLAC (10.21%), which was improved with cooling since the crystal-
lization peaks (first and second) besides the melting became better defined, pointing
to a better reorganization of the polymer chains.
The melting temperatures Tm for the two samples show a significant difference,
which is attributed mainly to the structural difference between the two materials,
possibly due to different proportions between the l-lactic acid/d-lactic acid type
monomers along the polymer, resulting from different synthesis processes, in which
the chains can aggregate in points with a greater proportion of l or d, forming
clusters that make the material more crystalline, as described in Kumar et al. [40]
and Ceregatti et al. [41].
On the other hand, the presence of conductive filler dispersed in the PLA chains
promotes a better thermal stability; however, it makes it difficult to reorganize
the chains, making the material less crystalline, after successive cycles of fusion-
solidification, unlike that presented by Ceregatti et al. [41], in which the material
presented higher crystallinity in the second heating with CNT in concentrations up
to 5%.
3.3 Electroactive Surface Response and Geometries
A batch-injection cell was used to determine the electrode area and subsequent evalu-
ation of the best printing conditions (direction, fill type and surface), which was filled
by electrolyte-holder for later activation and measurements. Subsequent addition of
adequate mass of the probe allowed, under homogenization, to reach the desired
concentration of the same and then the measurements of the electroactive areas.
The determination of the electroactive area was performed using cyclic voltam-
metry between −1 and +1.5 V versus Ag|AgClsat , at different scanning rates and
three consecutive cycles for each condition. A preliminary activation step in the same
window was performed at a speed of 100 mV s−1 . Parameters such as peak shift, and
the relationship between anodic and cathodic currents, as well as the linear relations
between current and velocity values (Ip × v1/2 ) were used to evaluate the best condi-
tions of application of the material as an electrode surface. In addition, the equation
of Randles and Sevcik was used to determine the electroactive area.
Different electrode prototypes (with dimensions 2.8 × 1.5 × 0.1 cm) obtained in
the horizontal and vertical modes and with Linear, Heli, Archiemedean, Octagram
and concentric interfacial geometries, with fills of 10 and 100%, were adapted to the
described BIA cell, the seal being obtained by the use of acetic curing silicone glue,
so that the exposed geometric window remained constant throughout all the tests.
Based on the voltammetric profile, as well as the evaluation of the aspects cal-
culated, it was possible to notice that in this impression condition, the electroactive
area (0.070 cm2 , 18.18% of the exposed geometric area −0.385 cm2 ) is almost
corresponding to the percentage of conductive phase (20.45%), safeguarding the

approximations and errors. Comparing 100% with 10% of fill, following the same
printing pattern, a peak resolution loss of 10% is observed, according to the observed
peak distances, which has a direct impact on electro-area measurements, as calcu-
lated (0.158 cm2 ). However, in both fills, behaviors in cathodic and anodic regions
are compatible with each other and between the different test bodies. In this way, it
was decided by the evaluation, at this moment, of the vertical direction of impression
with 10% of filling.
Based on the results and aspects of the printing, it is noticed that using vertical
printing, the results are also contaminated by resolution loss, as expected, since other
characteristics such as resolution and filling become less important, in this way, the
subsequent evaluations in horizontal printing. Helbert geometry now shows a lower
spatial spike and compatible behaviors in the anodic and cathodic scanning regions,
and a calculated electroactive area of 0.375 cm2 , 97.40% of the exposure window
(percentage well higher than expected given the determined percentage of conductive
phase). This was followed by an exploration of Archiemedean geometry, resulting
in profiles not only better resolved (with spikes between 0.325 V peaks), but also as
better Ipa /Ipc ratios, suggesting higher reliability in the determination of electroactive
area, which in this case was 0.224 cm2 . Following the optimization, in order to
explore all allowed modes, impressions were made in the Octagram and concentric
modes. Under printing on Octagram, it was possible to observe the worst resolution
(1.05 V between peaks), culminating in a terribly contaminated electroactive area,
calculated at 0.409 cm2 ; (0.664 V), the area was the lowest, demonstrating loss of
load exchange capacity. All the results for electroactive areas were calculated through
the direct application of the Randles–Sevcik equation.
4 Final Remarks
The present work developed a preliminary evaluation of the influence of printing

conditions on the electrochemical behavior of PLA and carbon-based composites. It is
clear the behavior variability of different types of PLA, as well as the influence of this
variation on the thermal behavior of the composites, which consequently influences
the 3D printing processes. Experiments allowed a speculation of the conductor load
present in the conductive filament (about 20.5%), although it was not possible to
define its category. Such nature as well as particle size are vital aspects for a better
modeling of heat transfer phenomena by the composite, which is clearly related to
the printing mode.
The print modes, in turn, proved to be a decisive factor, being responsible for
improving or worsening the resolution, or favoring the exchange in each direction.
In all cases, the relations between currents (cathodic and anodic) were acceptable,
although the loss of resolution, translated in the form of spacing between peaks, has
greatly contaminated the calculation of electroactive areas in some cases. With this,
it can be seen that the best orientation among the evaluated geometries was Helbert.
Notably, an optimization of the printing process, as well as a better control in the

particle size, will provide higher electroactive areas, with adequate resolution, even
with low driver loads, which shows a promising line currently in development in
our Group, line is dedicated to the preparation of different filaments for future direct
printing of disposable conductive components for lab-free applications.
References
1. Cinti, S.: Electroanalysis: towards a “laboratory-free” approach. SM Anal. Bioanal. Tech. 1(1),
1004 (2016)
2. Wang, J.: Analytical Electrochemistry. 3ª edição. Weinheim: Ed. Wiley-VCH (2006)
3. Ismail, W.N.M., Matsuda, A.: Electrochemical imprinted sol-gel films detection of organophos-
phate chemical warfare agent. Def. S T Tech. Bull. 8(1), 9–16 (2015)
4. Mahyari, M.: Electrochemical sensing of thiodiglycol, the main hydrolysis product of sulfur
mustard, by using an electrode composed of an ionic liquid, graphene nanosheets and silver
nanoparticles. Microchim. Acta 183(8), 2395–2401 (2016)
5. Sikarwar, B., Shama, P.K., Singh, V.V., Boopathi, M., Singh, B., Jaiswal, Y.K.: Xanthine oxidase
immobilized electrode for electrochemical sensing of chemical warfare agent simulant. Sens.
Lett. 15(3), 227–235 (2017)
6. Elliot, J., Simoska, O., Karasik, S., Shear, J.B., Stevenson, K.J.: Transparent carbon ultrami-
croelectrode arrays for the electrochemical detection of a bacterial warfare toxin, pyocyanin.
Anal. Chem. 89(12), 6285–6289 (2017)
7. Hondred, J.A., Breger, J.C., Aalves, N.J., Trammel, S.A., Walper, S.A., Medintz, I.L., Claussen,
J.C.: Printed graphene electrochemical biosensors fabricated by inkjet maskless lithography
for rapid and sensitive detection of organophosphates. ACS Appl. Mater. Interfaces 10(13),
11125–11134 (2018)
8. Wang, X., Gartia, M.R., Jiang, J., Chang, T.-W., Qian, J., Liu, Y., Liu, X., Liu, G.L.: Audio
jack based miniaturized mobile phone electrochemical sensing platform. Sens. Actuators, B:
Chem. 209, 677–685 (2015)
9. Shim, W., Han, J.-H., Kim, S., Kwon, H.J., Pak, J.: A portable potentiostat with Bluetooth
communication for square wave voltammetry measurement. Trans. Korean Inst. Electr. Eng.
65(4), 622–627 (2016)
10. Fan, Y., Liu, J., Wang, Y., Luo, J., Xu, H., Xu, S., Cai, X.: A wireless point-of-care testing
system for the detection of neuron-specific enolase with microfluidic paper-based analytical
devices. Biosens. Bioelectron. 95, 60–66 (2017)
11. Kanchi, S., Sabela, M.I., Mdluli, P.S., Inamuddin, I., Bisetty, K.: Smartphone based bioanalyt-
ical and diagnosis applications: a review. Biosens. Bioelectron. 102, 136–149 (2018)
12. Ainla, A., Mousavi, M.P.S., Tsaloglou, M.-N., Redston, J., Bell, J.G., Fernández-Abedul, M.T.,
Whitesides, G.M.: Open-source potentiostat for wireless electrochemical detection with smart-
phones. Anal. Chem. 90(10), 6240–6246 (2018)
13. Colozza, N., Kehe, K., Popp, T., Steinritz, D., Moscone, D., Arduini, F.: Paper-based electro-
chemical sensor for on-site detection of the Sulphur mustard. Environ. Sci. Pollut. Res. 1–12
(2018)
14. Buzzetti, P.H.M., Oliveira, G.C., Azevedo, A.L.M., Michel, R.C., Ponzio, E.A., Semaan, F.S.:
Conductive composites as versatile substrates for modeling of interfaces and sensors. In:
Wythers, M.C. (ed.) Advances in Materials Science Research, vol. 21, 1st edn, pp. 1–24.
Nova Science Publisher, New York (2015)
15. Fonseca, C.A., Dos Santos, T.A.D., Resende, A.C., Semaan, F.S.: Paraffin-graphite based com-
posite electrodes as substrate for the electrodeposition of conductive films with analytical pur-
poses. Revista Virtual de Química 5(4), 538–540 (2013)
16. Amorim, R.B.S., Rabelo, A.C., Bottecchia, O.L., Muñoz, A.A., Richter, E.M.: Desenvolvi-
mento, caracterização e aplicação eletroanalítica de um compósito fluido de adesivo epóxi,
grafite e ciclo-hexanona. Quim. Nova 33(6), 1398–1402 (2010)
17. Vasiliev, V.V., Morozov, E.V.: Mechanics and Analysis of Composite Materials, 1st edn. Else-
vier Science Ltd. (2001)
18. Tallman, D.E., Petersen, S.L.: Composite electrodes for electroanalysis: principles and appli-
cations. Electroanalysis 2, 499–510 (1990)
19. McCreery, R.L.: Advanced carbon electrode materials for molecular electrochemistry. Chem.
Rev. 108, 2646–2687 (2008)
20. Corb, I., Manea, F., Radovan, C., Pop, A., Burtica, G., Malchev, P., Picken, S., Schoonman, J.:
Carbon-based composite electrodes: preparation, characterization and application in electro-
analysis. Sensors 7, 2626–2635 (2007)
21. Silva, A.L., Corrêa, M.M., Oliveira, G.C., Michel, R.C., Semaan, F.S., Ponzio, E.A.: Develop-
ment and application of a routine robust graphite/poly(lactic acid) composite electrode for the
fast simultaneous determination of Pb2+ and Cd2+ in jewelry by square wave anodic stripping
voltammetry. New J. Chem. 42, 19537–19547 (2018)
22. Pereira, R.B., Morales, A.R.: Estudo do Comportamento Térmico e Mecânico do PLA Modi-
ficado com Aditivo Nucleante e Modificador de Impacto. Polímeros 24(2), 198–202 (2014)
23. Milde, J., Hrušecký, R., Zaujec, R., Morovič, L., Görög, A.: Research of ABS and PLA materials
in the process of fused deposition modeling method. DAAAM International, Vienna (2017)
24. Kalcher, K., Svancara, I., Buzuk, M., Vytrias, K., Walcarius, A.: Electrochemical sensors and
biosensors based on heterogeneous carbon materials. Monatsh. Chem. 140, 861–889 (2009)
25. Geim, A.K., Novoselov, K.S.: The rise of graphene. Nat. Mater. 6, 183–191 (2007)
26. Rivas, G.A., Rubianes, M.D., Pedano, M.L., Ferreyra, N.F., Luque, G.L., Rodrigues, M.C.,
Miscoria, S.A.: Carbon nanotubes paste electrodes: a new alternative for the development of
electrochemical sensors. Electroanalysis 19(7–8), 823–831 (2007)
27. Miranda-Hernandez, M., González, R.: Characterization of carbon–fullerene–silicone oil com-
posite paste electrodes. Carbon 43(9), 1961–1967 (2005)
28. Scida, K., Stege, P.W., Haby, G., Messina, G.A., García, C.D.: Recent applications of carbon-
based nanomaterials in analytical chemistry: critical review. Anal. Chim. Acta 691(1–2), 6–17
(2011)
29. Vieira, J.E.D., Vilar, E.O.: Grafeno: Uma revisão sobre propriedades, mecanismos de produção
e potenciais aplicações em sistemas energéticos. Revista Eletrônica de Materiais e Processos
11(2), 54–57 (2016)
30. Zarbin, A.J.G., Oliveira, M.M.: Nanoestruturas de carbono (nanotubos, grafeno): quo vadis?
Quim. Nova 36(10), 1533–1539 (2013)
31. Castro, M.O.: Síntese de grafeno pelo método CVD. 84 f. Dissertação de Mestrado em Física.
Departamento de Física, Universidade Federal do Ceará, Ceará (2011)
32. Guo, S., Yang, X., Heuzey, M., Therriault, D.: 3D printing of multifunctional nanocomposite
helical liquid sensor. J. Name 00, 1–3 (2012)
33. Vernadou, D., Vasilopoulos, K.C., Kenanakis, G.: 3D printed graphene-based electrodes with
high electrochemical performance. Appl. Phys. A 123, 623 (2017)
34. Palenzuela, L.C.M., Novotný, F., Krupicka, P., Sofer, Z., Pumera, M.: 3D-printed
graphene/polylactic acid electrodes promise high sensitivity in electroanalysis. Anal. Chem.
(2018)
35. Azevedo, A.L.M., Oliveira, R.S., Ponzio, E.A., Semaan, F.S.: Sensor development exploiting
graphite-epoxy composite as electrode material. In: IOP Conference Series: Materials Science
and Engineering, vol. 97 (2015)
36. Auras, R., Harte, B., Selke, S.: An overview of polylactides as packaging materials. Macromol.
Biosci. 4, 835 (2004)
37. Vidovic, E., Faraguna, F., Jukic, A.: Influence of inorganic fillers on PLA crystallinity and
thermal properties. J. Therm. Anal. Calorim. (2016)
38. Du, Y., Wu, T., Yan, N., Kortschot, M.T., Farnood, R.: Fabrication and characterization of fully
biodegradable natural fiber-reinforced poly (lactic acid) composites. Compos. Part B Eng. 56,
717–723 (2014)
39. Van de Velde, K., Kiekens, P.: Biopolymers: overview of several properties and consequences
on their applications. Polym. Test. 21, 433–442 (2002)
40. Kumar, S., Bhatnagar, N., Ghosh, A.K.: Effect of enantiomeric monomeric unit ratio on thermal
and mechanical properties of poly (lactide). Polym. Bull. 73, 2087–2104 (2016)
41. Ceregatti, T., Pecharki, P., Pachekoski, W.M., Becker, D., Dalmolin, C.: Electrical and thermal
properties of PLA/CNT composite films. Revista Matéria 22(3) (2017)
Part XIII
Information and Communication
Technology in Education
Interactive System to Improve the Skills
of Children with Dyslexia: A Preliminary
Study
Jorge Buele , Victoria M. López , L. Franklin Salazar ,

Jordan-H. Edisson, Cristina Reinoso, Sandra Carrillo, Angel Soria ,
Raúl Andrango and Pilar Urrutia-Urrutia
Abstract This paper describes a virtual system for the strengthening of linguistic
abilities of children with dyslexia. To achieve this objective, an intuitive interface
has been developed. The interface consists of three games (each with three levels
of difficulty), all of them being part of a rehabilitation program. These applications
combine visual and auditory messages that complement each other, in order to pro-
vide an immersive experience and to train more than one of the five senses at a time.
The virtual environment has been developed in Unity 3D software, and the diffusion
of sounds is performed through binaural hearing aids. To evaluate the performance
of the presented proposal, a sample of eight infants (three girls and five boys) with
J. Buele (B) · L. Franklin Salazar · J.-H. Edisson · C. Reinoso · S. Carrillo · P. Urrutia-Urrutia

L. Franklin Salazar
J.-H. Edisson
e-mail: edissonpjordan@uta.edu.ec
C. Reinoso
e-mail: ci.reinoso@uta.edu.ec
S. Carrillo
e-mail: sandralcarrillor@uta.edu.ec
P. Urrutia-Urrutia
J. Buele · V. M. López
Universidad de Las Fuerzas Armadas ESPE, Latacunga 050104, Ecuador
e-mail: vmlopez2@espe.edu.ec
A. Soria
Purdue University, Lafayette 47907, USA
e-mail: asoriach@purdue.edu
R. Andrango
Universidad Técnica de Cotopaxi, 050102 Latacunga, Ecuador
e-mail: raul.andrango@utc.edu.ec

https://doi.org/10.1007/978-981-13-9155-2_35
440 J. Buele et al.
ages between 8 and 12 years was chosen. As an inclusion criterion, it is determined

that the user must have an age greater than 7 and less than 13 years, and as a criterion
of exclusion the presence of some visual impairment and/or an auditory disorder or
other disorder that affect the educational environment. It must be taken into account
that this is a preliminary study and that the experimental results in the patients will
be seen in the long term. Therefore, to evaluate the acceptance of the system, and
to make future corrections a SUS usability test has been applied, with the following
result (82.5 ± 0.52). At the end of the execution of the applications, a report is issued
in which the specialist can control how the patient has progressed during sessions.
Keywords Human–computer interaction · Dyslexia · Open educational resources
1 Introduction
Children school performance depends on their physical, emotional sensory, and psy-
chological well-being [1, 2]. When the child suffers from a disorder or syndrome,
their performance drops markedly, as well as its ability to relate to their environment
[3–5]. Among the main learning disorders include dyscalculia, non-verbal learning
disorder (NVT), disortography, dysgraphia, dysphasia, and dyslexia [6, 7]. The latter
affects the development of the infant, preventing him from acquiring efficient reading
and spelling skills despite adequate teaching and the absence of manifest sensorial
and/or neural deficiencies [8]. This language learning disability affects one in five
people in the world, and it can produce anxiety due to the subject imagining scenarios
of mockery or social rejection [9, 10]. They also have difficulty remembering words,
and problems distinguishing certain sounds, so they pronounce words with similar
sounds, and this increases confusion [11]. Determining the origin of this syndrome
coupled with the visual and/or auditory disjunction that it produces has led to the
development of several related investigations. As shown in [8], several prominent
sensory theories of dyslexia are evaluated. Particularly, the effects of experiencing
scarce reading in the brain of affected children are presented as a priority cause. In
[12], a study is presented, which aims to discover the nature of this disorder, and
to achieve this complete brain regions are examined using the non-invasive auditory
P300 paradigm in children with dyslexia and neurotypical controls.
Cases of dyslexia occur in all known languages, although it differs according
to the spelling, which triggers labor problems as well. That is why identifying it
at an early age and carrying out an adequate therapeutic process is a high priority.
In [13], we describe a cognitive behavioral therapy based on mindfulness called
Mindfulness Based Rehabilitation of Reading, Attention & Memory (MBR-RAM
©). It uses techniques of visual meditation to improve attention focusing on formats
assisted by therapists, and practice at home to rehabilitate reading deficits, lack of
visual attention, lack of visual motor coordination, and visual memory. The proposal
was tested in three children between 8 and 10 years of age, obtaining significant
clinical and statistical improvements after a trial period of 6 months. In [14], several
Interactive System to Improve the Skills of Children … 441
technological proposals that seek to remedy the reading difficulties of children who
have this disability are described. “Friendly” features are incorporated for the user,
such as altering the size and format of the text and converting text to speech. As can be
seen in [15], the use of technological tools contributes to the rehabilitation of users.
A book that incorporates a set of mental games, mazes, and coloring pages, ideal for
creative children with dyslexia, ADHD, Asperger syndrome, and autism (volume 3),
is presented. The latter being an interesting proposal that incorporates 180 pages and
that is distributed at an affordable cost to parents. The exponential development of
technology has enabled greater and better benefits to be implemented, in this case
focused on medical applications. The main tools currently available are virtual reality
(VR) and interactive systems [16, 17]. Therefore, in [18] a virtual psychometric tool
for the rehabilitation of patients with dyslexia is presented. Such tool is based on the
Nintendo Wii video game system. The results show that it is a tool that improves
attention, but it has no immediate effect on reading performance, suggesting a longer
protocol as future work.
The standard rehabilitation of this learning disorder involves a classic format of
paper and pencil training, meaning that it presents boring and routine exercises to the
subject. In this context and based on the previously cited studies, this paper proposes
the implementation of a virtual system for the rehabilitation of children suffering
from dyslexia. For this, an application has been developed in Unity software that
incorporates three games with three levels of difficulty each (easy, medium, and
hard), which must be executed in order for a better recovery process. In the first, word
games are presented, allowing the user to develop spelling and reading skills. In the
remaining games, the sense of hearing is added, and through this the identification
words, and if their syllables are in the proper order. In order to corroborate the
performance of this proposal, the respective experimental tests were carried out.
This work is composed of four sections: Sect. 1 describes the introduction and
the works related to the topic. Section 2 shows the description and design of the
implemented system. Section 3 presents the tests and results, and finally in Sect. 4
the conclusions and future work are presented.
2 Methodology
2.1 Materials
Hardware The implemented system has a commercial brand laptop, where the
keyboard and mouse are used as input devices and the screen is the visual output to
the patient. Binaural hearing aids are used to provide complementary therapy with
sound outputs, when it is required. The general diagram of the project is presented
in Fig. 1.
Software The software used for the development of the interactive interface is Unity
3D and its complementary packages. This software has tools that allow the imple-
442 J. Buele et al.
Fig. 1 General diagram of the implemented system
mentation of virtual environments simulating real scenarios. You can develop envi-
ronments in 2D and 3D; in this case, the application is set to work in two dimensions,
since the child’s eye development is not yet complete, and taking into account the
ergonomics. Video games must be executed in order, since the training program has
been carried out with a pre-established sequence.
2.2 Methods
Sample It is made up of eight children (three female and five male), and their demo-
graphic characteristics are described in Table 1. The criteria for the selection of the
sample are: age (between 8 and 12 years) and the diagnosis of dyslexia (with/without
dysgraphia). Exclusion criteria include: presence of other disorders in the educational
setting and other psychological or neurological diseases. Also, there are some visual
deficiency and/or an auditory disorder.
Development of scripts In this section, the development of each one of the imple-
mented scripts are explained, for example, how these flat text files manage the infor-
mation entered by the user and allow him or her to interact with the environment.
Based on the flow diagram presented in Fig. 2, scripts are programmed to perform
the necessary tasks in the proposed rehabilitation program.
Table 1 Demographic characteristics of the sample

Item Gender Age (year) Education (years)
1 F 8 4
2 F 11 6
3 F 11 6
4 M 9 4
5 M 9 4
6 M 10 6
7 M 12 7
8 M 12 6
Fig. 2 Flowchart of the application developed

444 J. Buele et al.
Video games designed to encourage the enrichment and compensation of language

skills and linguistic functions of children with dyslexia. The main interface requires
the child to enter their name and age, using the computer keyboard and thus start.
In the selection menu, there are three games (with two levels of difficulty each) that
allow the user to train their language skills, while entertaining. In addition, an option
is presented to obtain and save the data of each user, since this can be used to evaluate
their evolution. Figure 3 shows the main screen of the application, and the screen for
entering a user’s data is in Fig. 4.
Procedure In this subsection, the developed games and a brief description are
shown.
Fig. 3 Main screen of the application
Fig. 4 Screen for user registration

Fig. 5 Interface of the first game presented to the user
(i) In the first game, an incomplete word is presented on the screen and a series of
syllables are placed, and the user has to choose the complement to said word
and thus make sense of it. Figure 5 presents the developed interface.
(ii) In the second game, the child is presented with a word composed of two sylla-
bles, and in parallel, an auditory message is reproduced. It must be determined
if the sound corresponds to the word that is shown on the screen, meaning that
if the order of the syllables is the right one or is the other way around; according
to that, “yes” or “no” is chosen. Figure 6 shows the interface.
(iii) In the third application, a word is played through the headphones and on the
screen two words of two or three syllables (depending on the level of difficulty)
each one are presented. The user has to choose between these two options,
which is the word that has the correct order of their syllables. Figure 7 shows
the developed interface.
Fig. 6 Interface of the second game running

446 J. Buele et al.
Fig. 7 Interface of the third game that the user uses
3 Tests and Results
3.1 Tests
The participants were subjected to sessions of 30 min, two times per week, for a
period of 5 weeks. Figure 8 shows how the tests were developed in two of the eight
users. For the execution of tasks, it is recommended that a psychologist or educational
therapist conducts the session and supervises each video game. Although it has been
designed intuitively, an adult that is not specialized can direct it. The exercises were
held in private sessions, and outside school hours with the intention of avoiding some
kind of social pressure, and the indications were repeated until the child understands
them.
3.2 Results
At the end of the execution of the games, the interface has an option that exports
the data to a file with a.txt extension. This report contains information detailing the
name of the patient and his score obtained, so that the specialist can verify if there
has been an improvement. In addition, the SUS evaluation test has been applied to
users, also known as the system usability scale created by John Brooke in 1986. The
system usability scale (SUS) provides a reliable and fast tool for the measurement
of usability, thus allowing the evaluation of a wide variety of products and services,
including hardware, software, mobile devices, Web sites, and applications. The SUS
test can be used in small samples with very reliable results and is valid since it can
effectively differentiate between usable and non-usable systems. For this reason,
in the developed application the scores have been normalized in order to obtain a
percentile classification, which is (82.5 ± 0.52), and as the obtained result is superior
to 68, it means that the developed system has a good degree of acceptance.
Through this research, patients with dyslexia have been given a virtual tool to
strengthen their linguistic abilities. It is a valid technological proposal because it
has visual and audible feedback to improve concentration and the learning process.
In this way, conventional exercises have been set aside, with the intention of provid-
ing greater understanding, confidence, and better experiences to the student. It should
be clarified that the present work is a preliminary study that shows the development
and application of video games with the use of Unity 3D software in a mixed set of
children. Being an approved system, there are different types of experimental results
in each patient, and it depends on the reactions each of them presents. As a metric
to determine the validity of this prototype, the SUS usability test was implemented,
which has demonstrated that the interface presented to the user is easy to use, intu-
itive, and friendly since it is focused in evaluating infants. As an important remark,
this research has been evaluated from the point of view of the acceptance generated
by this system, but not by evaluating the approach of educational therapy, which is
part of a broader investigation scope.
Despite obtaining positive results, the authors of this research are making mod-
ifications to the system based on the feedback obtained from the comments of the
users. In addition, what is proposed, as future work is to present a more personal-
ized interface that can be applied to a larger sample, also enabling the assessment
of what is the impact it has on adolescents, and even on adults who have this syn-
Fig. 8 Patients runs the games as part of their rehabilitation sessions

448 J. Buele et al.
Table 2 SUS results

Question Result (N =
8)
Mean SD
1. I think I would like to use this virtual reality system frequently 4 0.63
2. I found the virtual reality system to be unnecessarily complex 2.2 0.4
3. I thought that the virtual reality system was easy to use 4.4 0.8
4. I think I would need the support of a technical person to be able to use this 2.8 0.4
virtual reality
5. I found that the various functions in this virtual reality system were well 4.2 0.75
integrated
6. I thought there was too much inconsistency in this virtual reality system 1.8 0.4
7. I would imagine that most people would learn to use this virtual reality 4.8 0.4
system very quickly
8. I found the virtual reality system to be very cumbersome to use 1.4 0.48
9. I felt very confident using the virtual reality system 4.8 0.4
10. I needed to learn a lot of things before I could get going with this virtual 2.6 0.48
reality system
Global score (total) 82.5 0.52
drome. It should be emphasized that for the test users, 3D games could already be
implemented, which will provide another type of experimental results (Table 2).
Acknowledgements This work was financed in part by Universidad Técnica de Ambato (UTA)
and Dirección de Investigación y Desarrollo (DIDE) under project PFISEI 26.
References
1. Cameron, C.E., Cottone, E.A., Murrah, W.M., Grissmer, D.W.: How are motor skills linked to
children’s school performance and academic achievement? Child. Dev. Perspect. 10(2), 93–98
(2016)
2. Baroutsis, A., Lingard, B.: Counting and comparing school performance: an analysis of media
coverage of PISA in Australia, 2000–2014. J. Educ. Policy 32(4), 432–449 (2017)
3. Björkenstam, E., Dalman, C., Vinnerljung, B., Weitoft, G.R., Walder, D.J., Burström, B.: Child-
hood household dysfunction, school performance and psychiatric care utilisation in young
adults: a register study of 96 399 individuals in Stockholm County. J. Epidemiol. Community
Health 70(5), 473–480 (2016)
4. Skerfving, S., Löfmark, L., Lundh, T., Mikoczy, Z., Strömberg, U.: Late effects of low blood lead
concentrations in children on school performance and cognitive functions. Neurotoxicology
49, 114–120 (2015)
5. Ranning, A., Laursen, T., Agerbo, E., Thorup, A., Hjorthøj, C., Jepsen, J.R.M., Nordentoft,
M.: School performance from primary education in the adolescent offspring of parents with
schizophrenia and bipolar disorder—a national, register-based study. Psychol. Med. 48(12),
1993–2000 (2018)
6. Mayes, S.D., Frye, S.S., Breaux, R.P., Calhoun, S.L.: Diagnostic, demographic, and neurocog-
nitive correlates of dysgraphia in students with ADHD, autism, learning disabilities, and neu-
rotypical development. J. Dev. Phys. Disabil. 30, 1–19 (2018)
7. Berninger, V.W., Richards, T.L., Abbott, R.D.: Differential diagnosis of dysgraphia, dyslexia,
and OWL LD: behavioral and neuroimaging evidence. Read. Writ. 28(8), 1119–1153 (2015)
8. Goswami, U.: Sensory theories of developmental dyslexia: three challenges for research. Nat.
Rev. Neurosci. 16(1), 43–54 (2015)
9. Eden, G.F., Olulade, O.A., Evans, T.M., Krafnick, A.J., Alkire, D.R.: Developmental dyslexia.
Neurobiol. Lang. 815–826 (2016)
10. Knoop-van Campen, C.A., Segers, E., Verhoeven, L.: The modality and redundancy effects in
multimedia learning in children with dyslexia. Dyslexia 24(2), 140–155 (2018)
11. Diamanti, V., Goulandris, N., Stuart, M., Campbell, R., Protopapas, A.: Tracking the effects of
dyslexia in reading and spelling development: a longitudinal study of Greek readers. Dyslexia
24(2), 170–189 (2018)
12. Papagiannopoulou, E.A., Lagopoulos, J.: P300 event-related potentials in children with
dyslexia. Ann. Dyslexia 67(1), 99–108 (2017)
13. Pradhan, B., Parikh, T., Sahoo, M., Selznick, R., Goodman, M.: Current understanding of
dyslexia and pilot data on efficacy of a mindfulness based psychotherapy (MBR-RAM) model.
Adolesc. Psychiatry 7(1), 44–55 (2017)
14. Caute, A., Cruice, M., Marshall, J., Monnelly, K., Wilson, S., Woolf, C.: Assistive technology
approaches to reading therapy for people with acquired dyslexia. Aphasiology 32(sup1), 40–42
(2018)
15. Brown, S.J., Games, D.: Brain Games, Mazes & Coloring Pages-Homeschooling with
Minecraft: Dyslexia Games Presents an Activity Book-Great for Creative Kids with Dyslexia,
ADHD, Asperger’s Syndrome and Autism, vol. 3. ACM, USA (2018)
16. Ali, M., et al.: iCBLS: an interactive case-based learning system for medical education. Int. J.
Med. Informatics 109, 55–69 (2018)
17. Renganathan, S.M., Stewart, C., Perez, A., Rao, R., Braaten, B.: Preliminary results on an
interactive learning tool for early algebra education. In: 2017 IEEE Frontiers in Education
Conference (FIE), pp. 1–5. IEEE (2017)
18. Pedroli, E., Padula, P., Guala, A., Meardi, M.T., Riva, G., Albani, G.: A psychometric tool for
a virtual reality rehabilitation approach for dyslexia. Comput. Math. Methods Med. 2017, 1–6
(2017)
Use of Information and Communication
Technologies in the Classroom:
An Exploratory Study in Professional
Military Education
Nuno Alberto Rodrigues Santos Loureiro and

Lúcio Agostinho Barreiros dos Santos
Abstract The study focuses on the use of information and communication technolo-
gies in professional military education, delimited to the present moment and to the
current professors and students of the Portuguese Armed Forces Military University
Institute. This study aims to analyse the relationship between ICT and teaching–learn-
ing methodologies in the context of professional military education, with emphasis
on the concept of innovation in the classroom, on teaching–learning styles and on the
role of ICTs in learning. The study is based on a mixed research strategy combining
deductive and inductive approaches, materialized in a case study, with data collected
through questionnaires, interviews and documentary analysis. This article focuses on
the conceptual framework and its interconnection with the methodological strategy
for the empirical study. The results presented refer to the pilot validation test of the
data collection instruments and indicate that, in general, students and teachers share
the same teaching–learning styles, both students and teachers show a good propen-
sity to use ICT in the classroom and an almost complete substitution of traditional
lessons, lectures and conferences by distance learning with a face-to-face approach
in the discussion of case studies is possible.
Keywords Professional military education · Teaching–learning methodologies ·

Information and communication technologies
1 Introduction
Military Higher Education in Portugal consubstantiates itself in the accomplish-

ment of courses and cycles of studies, within the military sciences, preparing highly
qualified cadres, developing command, direction and leadership skills. Within the
N. A. R. S. Loureiro (B) · L. A. B. dos Santos

Military University Institute Research and Development Centre (CIDIUM), Military University
Institute, Lisbon, Portugal
e-mail: nuno.a.loureiro@gmail.com
L. A. B. dos Santos
e-mail: labs0892@gmail.com

https://doi.org/10.1007/978-981-13-9155-2_36
452 N. A. R. S. Loureiro and L. A. B. dos Santos
Portuguese Armed Forces, this responsibility is assigned to the Military University

Institute (MUI), formally created in October 2015. In today’s changing environment,
new generations (e.g. Millennials and Zeds) learn differently. The increasing use of
information and communication technologies (ICT) in the context of military educa-
tion makes it easier to achieve training objectives by concentrating on the selection
and integration of technological tools with the highest educational returns, adequate
for the teaching of the main military competences for the twenty-first century.
The study object of this research is the role of ICT in the teaching–learning
process in the context of military professional education (PME), and is delimited at
three levels:
• Temporally, to the academic year of 2018/2019;
• Spatially, at the MUI covering the students who attend and the teachers teaching
in the Senior Officer Course (SOC), Joint Staff Course (JSC) and Flag Officer
Course (FOC);
• Conceptually, in the analysis of the relationships between students ‘learning styles
and teachers’ teaching styles, the attitude of students and teachers towards the
introduction of ICT in the classroom and the adaptation of new teaching–learning
methodologies to curricular areas.
The general objective of the research, starting from the concept of innovation in
the “Classroom”1 (in its physical and virtual dimensions), is to analyse the teach-
ing–learning styles of teachers and students of the Post-Graduated Studies Depart-
ment of the MUI, evaluating the role of ICT and their suitability for PME.
It is important to analyse the relationships established between the key variables
of the study, which are expressed from some specific objectives (SO).
SO1 Identify and relate the students’ learning styles and teachers’ teaching styles
in order to understand their impact on the teaching–learning process;
SO2 Identify and analyse the attitudes of students and teachers, towards the
increased use of ICT in the “Classroom”, in order to understand their impact
on the teaching–learning process;
SO3 Evaluate the appropriateness of the new teaching–learning methodology to the
curricular areas, resulting from the use of ICT in the “Classroom”, in order to
identify opportunities for their introduction into the teaching–learning process.
The study seeks to answer the following research questions:
1. What are the learning styles preferred by the students and what are the teaching
styles adopted by teachers, what is the relationship between them, and how do
they contribute to the teaching–learning process?
1 The traditional classroom concept presents it as a space, usually a room of variable dimensions,
with environmental, dispositional, work and study conditions, where the formal teaching–learning
process takes place, whose development dates back to the century with the democratization of access
to education [1]. The “Classroom” concept of the future was created in 2011 under the iTEC Project
under the coordination of the European Schoolnet, with the purpose of designing a new proposal of
organization of classroom space and practices, by harmonious conjugation of three key elements:
space, technology and pedagogy [2].
Use of Information and Communication Technologies … 453
2. What are the attitudes of students and teachers towards the use of ICT in the
“classroom”, what is the relationship between them, and how do they contribute
to the teaching–learning process?
3. Which, among the new ICT teaching–learning methodologies, are the most
appropriate to the curricular contents of the PME?
In the present article we intend, essentially, to develop the conceptual framework
and its interconnection with the methodological strategy, that is, the definition of the
instruments of data collection for the complete empirical study to be developed. In this
way, the results presented now refer to the validation pilot test. To this end, the article
is structured in three chapters, besides the introduction and conclusion. Initially, a
review of the literature is done, highlighting the PME new challenges facing the
VUCA2 context of today, the importance of technology in the development of new
teaching–learning methodologies, the emergence of new learning environments, the
teaching–learning styles and the acceptance of the technology. Secondly, a detailed
description of how the study is being conducted is given, including an explanation
of the concepts and the operationalization of the variables under study. Next, the
preliminary data collected are summarized, characterizing the sample and analysing
the relevance of the data collected.
2 Literature Review
The military profession, analysed from a vocational or professional perspective [3],

requires a broad set of very comprehensive competencies to fulfil its duties and
obligations, within its demanding functions of command, direction, leadership, staff
and execution, among others. The education and training given should be in line
with what the military must do to fulfil the mission. Thus, systemic approach should
include four steps: analysis, curriculum design, development and implementation.
In the construction of course references, functional analysis is essential and takes
into account the work context, organizational systems, functional relations and their
future application [4].
Another challenge facing the armed forces is related to the comprehensiveness
and demands of knowledge, since it is increasingly necessary to develop more com-
plex knowledge, skills and attitudes, such as critical knowledge and understanding
of military operations; different joint military capabilities; deep understanding of the
operational planning process, rules of engagement and risk management; understand
and deal with the complexity and diversity of tasks; be very adaptable to different
positions and functions, in contexts of change and new challenges; demonstrate
authority, autonomy, professional integrity, sustained commitment to the develop-
ment of processes at the forefront of work and innovation.
2 Concept introduced by US Army War College to describe the multilateral world resulting from
the end of the Cold War: Volatile, Uncertain, Complex and Ambiguous.
In the classroom, technology can span all types of tools. The latest technologies
allow us to experiment classroom methodologies, both physical and virtual, that
were not previously possible, depending fundamentally on what we want to accom-
plish. Teaching with technology can deepen student learning by supporting training
objectives. However, it can be a huge challenge to select the “best” technology tools
without losing sight of the learning objectives. Once identified, the integration of
these tools can also be another challenge, although it can become an experience with
high educational returns.
2.1 Technology and New Teaching–Learning Methodologies
It is common to find some reluctance of higher education institutions regarding

the adoption of new procedures, processes and models of teaching–learning, being
sometimes more common to exist a conservative position on the part of the military
institution [5].
But the recovery of the backwardness that may exist in higher education institu-
tions, whether civilian or military, may be based on a strategy, not to go through all
unfulfilled development stages, but to neglect the intermediate and try as much as
possible to keep abreast of the current stage of scientific and technological devel-
opment by adopting new technologies to support the teaching–learning process. It
may, however, be necessary to evaluate the implications of “burning” some steps.
Another perspective on the role of technology in the teaching–learning process
is brought to us by Peter J. Denning and Susan L. Higgins that report a marked
increase in the discussion of technological advances in learning environments, in
particular the massive open online courses (MOOC) Internet-based platforms, thus
being an automated learning environment (ALE). In a military context, senior leaders
are asked to go beyond basic skills and are required to be at least at the “proficient”
level, which cannot be achieved by recourse to a mere ALE, as those leaders work
in environments where rules are constantly changing, so the challenge in PME is to
go beyond technologies when trying to develop higher levels of leadership [6].
According to Neal [7], although there has been a growing use of simulators in
training programs, however, care should be taken to evaluate in advance the contexts
of training in which this use allows effectively gains in effectiveness and those in
which this does not happen. As an example, this author points out that topics related to
military doctrine, politics or history are typically treated through lectures, discussion
and debate, and should not be addressed by technological products that would be
downright inappropriate.
As Schatz et al. [8] argue, the military context requires a permanent development
of skills in its staff, and this process is also based on the teaching and training given
to them: “[…] military personnel require an expanded set of competencies, higher
levels of nuanced skills such as critical thinking and emotional intelligence, and more
efficient and agile pathways to expertise, and that achieved at least in part depends
on revising the military learning enterprise”.
2.2 Emerging Learning Environments
Referring to WEB 2.0 technology, Foon Hew and Sum Cheung [9] affirm that this
is an interactive tool that facilitates bidirectional interaction, requiring the students
to contribute to the placement of information on the site available, interacting and
sharing their ideas and knowledge in a collaborative way. According to the authors,
the technological tools of WEB 2.0 allow teachers to apply a constructivist approach
to teaching–learning. There are three possible ways to use WEB 2.0 tools within the
scope of teaching: Blogger; Active Worlds; and Teachertube [9, 10 cited by 11].
Pinel [12] states that students are required to acquire twenty-first century ICT skills
because they are related to the real world, develop as students construct knowledge,
to contribute to society.
An Australian publication on educational technology edited by Education Tech-
nology Solutions (ETS), recently published an online article entitled Technology
That Will Shape Education In 2017, in which it present a guide to the main trends
in educational technology that will have a relevant impact on way of learning and
teaching. The way in which education professionals learn, teach and collaborate is
changing significantly with the emergence of new technologies, such as virtual real-
ity, augmented reality, learn from anywhere, teaching from anywhere mobile devices,
collaboration technology, gamification, coding, evolving learning spaces and styles
and the maker movement [13].
2.3 Learning Styles
The assimilation and processing of information and knowledge are different from
individual to individual. They can do this by seeing or hearing, reflecting or acting,
reasoning in a logical or intuitive way and analysing or visualizing. Students have dif-
ferent strengths and preferences in how they receive and process information, which
means that they have different learning styles [11, 14]. Teaching methodologies also
vary, with some teachers preferring lectures and others favouring demonstration or
leading students to self-discovery; some focus on principles and others on applica-
tions; some emphasize memory and others emphasize understanding.
Difficulties arise when there are incompatibilities between the student’s learning
style and the teacher’s teaching style; students may become bored and inattentive in
class, poor test performance, and lose motivation regarding courses or curriculum
[15–17]. Empirical studies have confirmed the hypothesis that teachers who are able
to adapt their teaching style to include the two poles of each dimension, ideal for
most (if not all) students in a class, are more effective [14].
The Felder and Silverman model, in the 2002 update [18], ranks a student’s learn-
ing style from one category or another in each of the four dimensions through the
answers given to four questions:
1. What kind of information the student preferentially perceives: sensory (visions,
sounds, physical sensations, concrete, practical, fact-oriented and intuitive think-
ing) or intuitive (possibilities, perceptions, hunches, abstract thinking, innovative,
oriented to theories and meanings underlying)?
2. Through which sensitive channel is the most effectively perceived external infor-
mation: visual (prefer visual representations of displayed material such as figures,
diagrams, flowcharts, graphs and demonstrations) or verbal (prefer written and
spoken explanations)?
3. How does the student prefer to process information: actively (through involve-
ment in physical activity or discussion, experiencing things and working in
groups) or reflexively (through introspection, prefer to work alone or with a
single family partner)?
4. How does the student progress towards understanding: sequentially (linear think-
ing process and learning in small incremental steps in continuous steps) or glob-
ally (holistic thinking process and learning in big leaps)?
The same model states that teaching style can also be defined, in one or another
category of each of the four dimensions, in terms of answers to four questions:
1. What kind of information is emphasized by the teacher: concrete (factual) or
abstract (conceptual, theoretical)?
2. Which form of presentation is preferred by the teacher: visual (visual images,
diagrams, films and demonstrations) or verbal (verbal talks, readings and discus-
sions)?
3. What mode of student participation is facilitated by the teacher’s presentation:
active (do students speak, move and reflect) or passive (do students watch and
listen)?
4. What kind of perspective is provided on the information presented: sequential
(step-by-step, showing “the trees”) or global (relevance (presenting “the forest”)?
Table 1 (see Appendix) summarizes learning and teaching styles in their different
dimensions and categories, as proposed by Felder and Silverman [18].
The Index of Learning Styles is an online research tool used to evaluate pref-
erences in four dimensions (active/reflexive, sensory/intuitive, visual/verbal and
sequential/global), built from the model of learning styles formulated by Felder and
Silverman [11, 18]. The scale was developed and validated by Felder and Soloman
[15–17]. In their original version, respondents answer online to 44 questions, with
two alternative answers, “a” and “b”, and submit responses online, and receive a
report on their preferences regarding the four dimensions under review.
2.4 Acceptance of Technology in the Classroom
Acceptance of technology results from users’ beliefs and attitudes about their use,
which was measured with the perceived utility and perceived ease of use in the
technology acceptance model (TAM) [19, 20]. The Technology Acceptance Barriers
questionnaire is an instrument that intends to measure the extent and nature of the
barriers to the integration of ICT in the perspective of technology acceptance, com-
piled by Gu, Zhu, and Guo [21] into four constructs often referred to as predictors
of ICT acceptance: expectation of results, adjustment of technology to task, social
factor and personal factor.
Outcome expectancy (OE) results from internal beliefs and attitudes about the
use of ICT, and it is measured as perceived utility and perceived ease of use, being
considered the most important predictor of the use of technology [20, 22, 23]. Task-
technology fit (TTF) is the degree to which technology helps an individual to accom-
plish his or her tasks [24]. It is based on the assumption that users accept the technol-
ogy because of its potential benefits, such as performance improvement, regardless
of their attitude. Social factor (SF) is a type of social norm defined as “perceived
social pressure to carry out behaviour or not” [25, 26]. Studies suggest that perceived
social influence has a significant positive influence on individual beliefs about the
usefulness of technology and positively and significantly affects the use of ICTs
[22, 27, 28]. Personal factors (PF) construct includes computer self-efficacy and per-
sonal innovation through technology, widely recognized as explanatory factors that
influence the use of IT by end-users [27, 28].
3 Methodology
The study is based on a mixed research strategy [29] combining hypothetical-

deductive [30] and inductive [31] approaches, materialized in a cross-sectional study
[32], with data collected at a given moment, through questionnaires, interviews and
documentary analysis. Due to the temporal, spatial and conceptual boundaries defined
for this research, the target population of the present study is composed of students
and professors of the SOC, JSC and FOC of the Military University Institute, in the
academic year 2018/2019.
The preliminary results presented are based only on the pilot test, submitted to a
convenience sample composed of the students and professors of the SOC, JSC and
FOC of the Military University Institute, in the academic year 2017/2018.
3.1 Questionnaire
For the empirical study, a data collection instrument was prepared based on the
following scales: (1) Index of Learning Styles (ILS), proposed by Soloman and Felder
[17]; and (2) Technology Acceptance Barriers (TAB), proposed by Gu, Zhu, and Guo
[21]. It was also complemented by ad hoc items designed to collect biographical and
professional data on the one hand and, on the other hand, to verify the traditional
teaching methodologies currently used in curricular areas and to identify the most
appropriate ICT-based teaching methodologies to each of the specific curricular areas.
In the present case, the scales were adapted to the Portuguese language and to the
military context.
In order to develop a comparative analysis between preferences of learning styles
and teaching styles between students and teachers, the results of the ILS question-
naire were transformed into a seven-point Likert scale, according to Table 6 (see
Appendix—Table 6).
All items of TAB questionnaire were also measured using a seven-point Likert
scale ranging from “strongly agree” to “strongly disagree”.
Regarding the ad hoc items introduced in the questionnaire, some asked the
respondent to choose the three most traditional teaching methodologies3 currently
used in each curricular area; others asked the respondent to choose the three ICT-
based teaching methodologies4 best suited to each of the specific curricular areas.
The questionnaire, necessarily different between students and teachers, after a
brief introduction, explaining the objectives of the study and guaranteeing the treat-
ment of the data in an anonymous and confidential way, and its exclusive use for the
mentioned effects, was thus composed of the four sections shown in Table 2 (see
Appendix).
In order to adapt the questionnaire formulation to the Portuguese language and the
military context, three sequential stages were followed: (1) translation and retrover-
sion, (2) pretest and (3) pilot test. As the items were initially expressed in the English
language, the retroversion was applied to guarantee a precise translation into Por-
tuguese. Subsequently, a pretest was applied, using a small number of respondents
(focus group) to test the adequacy of the questions and their comprehension. This
focal group, consisting of 10 subjects from each course (30 participants) suggested
some modifications in the items of the questionnaire, in order to be better understood
by the target population. After the pretest, a pilot test was performed and submitted
to all individuals of the SOC, JSC and FOC of the 2017/2018 school year, whose
objective was to test the reliability of the items (consistency).
3 Lesson/Lesson-Debate; directed discussion/learning discussion/topic discussion; individual appli-
cation work/application work group; CPX/CAX computer assisted; field exercises; field trip;
seminar/panel/symposium; workshop; lecture/ conference/cycle of conferences; presentation and
defence of work.
4 Digital Book; ongoing training online; gamification; project-based learning (PBL); case study;
team-based learning (TBL); flipped classroom; coding; virtual reality; computer-based training
(CBT); augmented reality; learn from anywhere/teach from anywhere mobile devices.
3.2 Pilot Test Sample and Preliminary Results
The pilot test universe consisted of 76 teachers and 225 students. The questionnaire
was answered by 46 teachers and 150 students. After validation of the answers given,
there were some errors and incongruities, which led to the elimination of some of
them. The final number of validated respondents was 24 teachers and 124 students,
corresponding, respectively, to 31.6 and 56.4% of the target population. Their demo-
graphic and professional characteristics can be seen in Table 3 (see Appendix).
As preliminary results from a pilot test, they should be interpreted as exploratory
data for the preparation of the questionnaire that will be submitted to the target
population of the study, SOC, JSC and FOC students of the 2018/2019 school year.
Reliability was estimated by the Cronbach’s alpha coefficient, and the items of
the study achieved a reliable result (ILS > 0.60, use of ICT > 0.97).
The answers to the questionnaire were, therefore, submitted to a first analysis,
using the statistical package for the social sciences (SPSS V.23) with regard to
descriptive statistics and comparative analysis of means, centred on two aspects:
(1) analysis of learning style profiles and (2) analysis of the barriers to the use of
ICT in the “Classroom”.
3.3 Learning Style Profiles Analysis
Table 4 (see Appendix—Table 4) summarizes the learning style profiles reported by

students and teachers of each of the courses under analysis (SOC, JSC and FOC),
and reads as follows, for example, of the 101 SOC students who completed ILS,
91% were classified as having a “sensory” profile (implying that 9% were classified
as having an “intuitive” profile) 88% prefer “visual” input of content (so 12% prefer
to “verbal” content input) and so on.
It is verified that an expressive majority of the students of the SOC and JSC show
preference for the “Sensory”, “Visual”, “Active” and “Sequential” style. Regarding
the FOC students, an expressive majority shows a preference for the “Visual” style,
but only a relative majority shows a preference for the “Sensory” and “Active” style,
even reversing the preference in the field of “Understanding” with a minority to opt
the “Sequential” style.
Regarding how teachers perceive the learning styles of the students, the majority
follow the preferences expressed by students, however, with regard to the dimen-
sion “Perspective”, a significant majority of teachers presents a “Sequential” style,
contrary to the majority of FOC students.
The analysis of ILS responses also allows us to identify (see Appendix—Table 5)
that a large percentage of students and teachers have a mild preference for a particular
learning style. This suggests that both students and teachers may be able to switch
between styles, adapting more easily to the “classroom” environment. However, in
the case of the students of all the courses, there is a moderate or strong preference for
the “Visual” style in the “Input” dimension, and also, in the case of the SOC and JSC
students, a moderate or strong style preference of “Sensory” perception, not being
followed by the teachers in this preference. Particular attention should be given to
these situations.
Therefore, with regard to global learning and teaching styles (see
Appendix—Table 7), and by gender (see Appendix—Table 8), the results indicate
no significant differences in the preferences and attitudes of students and teachers.
According to the branch of the armed forces to which the respondents belong
(see Appendix—Table 9), the results indicate significant differences in the air force
branch in the dimensions of “Perception/Content”, “Input/Presentation” and “Un-
derstanding/Perspective”.
According to the age group to which the respondents belong (see
Appendix—Table 10), the results indicate significant differences in the 51–55 group
in the four dimensions.
According to the course the respondents attend/teach (see
Appendix—Table 11), the results indicate no significant differences in the
preferences and attitudes of students and teachers, except in the dimension
“Understanding/Perspective” within the FOC students and teachers.
According to the effective service time to which the respondents belong (see
Appendix—Table 12), the results indicate significant differences in the 31–35 group
in the four dimensions.
3.4 Analysis of Barriers to the Use of ICT in the “Classroom”
Given the results of the data collected, both pupils and teachers indicate a good
outcome expectation with the use of ICT in the “Classroom”, consider ICT rea-
sonably adjusted for use in the context of professional military training, consider
themselves capable of to use and consequently accept the introduction of technology
into teaching–learning processes. The deepening of this study will allow us to see
how concretely it can be done.
3.5 Analysis of Adequacy of Teaching–Learning

Methodologies to Curricular Areas
Taking into account the preliminary results achieved (see Appendix—Tables 13, 14
and 15), whereas in the SOC, the most used traditional teaching–learning method-
ologies are the “Lesson/Lesson-debate”, followed by “Directed Discussion/Learning
Discussion/Topic Discussion” and “Lecture/Conference/Lecture Series”, in JSC
preference is given to “Directed Discussion/Learning Discussion/Topic Discussion”,
followed by “Lesson/Lesson-debate” and “Lecture/Conference/Lecture Series”. This
preference is emphasized in FOC, with the primacy given to “Directed Discussion/

Learning Discussion/Topic Discussion” followed by “Lecture/Conference/Lecture
Series” and “Lesson/Lesson-debate”. As regards the teaching–learning methodolo-
gies based on ICT, considered more appropriate, the results point to a preference
for the “Digital Book”, the “Continuing education online” and the “Case Study”,
although the “Team Based Learning” and “Gamification” were also referred.
These results seem to indicate an almost complete substitution of traditional
lessons, lectures and conferences by distance learning with a face-to-face approach
in the discussion of case studies.
4 Conclusions
Being part of a broader research on teaching–learning methodologies in PME,

through an approach to the introduction of ICT in the classroom, this paper has
essentially sought to develop the conceptual framework and their interconnection
with the methodological strategy, that is, the definition of the instruments of data
collection for the complete empirical study to be developed. The results presented
refer to the pilot validation test of these same data collection instruments.
The first stage of preparation of data collection instruments is in the process of
being finalized, a preliminary analysis indicates that:
• The adapted ILS scale has an acceptable internal consistency (alpha > 0.60) and
the adapted TAB scale has a very good internal consistency (alpha > 0.97);
• In general, students and teachers share the same teaching–learning styles; however,
when analysed according to control variables, there are significant differences;
• Both students and teachers show a good propensity to use ICT in the classroom;
• An almost complete substitution of traditional lessons, lectures and conferences
by distance learning with a face-to-face approach in the discussion of case studies
is possible.
In the light of these preliminary results, the data already collected will be further
analysed in the continuation of the study, and it will be sought to gather more infor-
mation and complete the objectives of the study, which will be a major contribution
to improving the quality of PME in the Portuguese Armed Forces.
Appendix
See Tables 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14 and 15.

Table 1 Learning and teaching styles dimensions and categories

Preferred learning style (student) Corresponding teaching style (teacher)
Categories Dimensions Categories Dimensions
Sensory → Perception Concrete → Content
Intuitive Abstract
Visual → Input Visual → Presentation
Verbal Verbal
Active → Processing Active → Student participation
Reflexive Passive
Sequential → Understanding Sequential → Perspective
Global Global
Table 2 Questionnaire characterization and number of items

Students
Section A—Biographical and professional data (control variables: Total—5 items
age, sex, effective service time, branch and course attended)
Section B—Learning styles index: identify students’ learning 11 items per dimension
preferences Total—44 items
Section C—Barriers to technology acceptance: identify students’ OE—8 items
propensity to use technology in learning PF—7 items
TTF—7 items
SF—4 items
Total—26 items
Section D1—Identify the current teaching/learning methodologies SOC—9 items
most appropriate to the curricular context of the SOC, JSC and JSC—16 items
FOC, for each of its curricular areas FOC—14 items
Section D2—Identify the new teaching/learning methodologies
best suited to the curricular context of the SOC, JSC and FOC, for
each of its curricular areas
Teachers
Section A—Biographical and professional data (control variables: Total—7 items
age, sex, effective service time, branch, time in teaching functions,
course teaching and specific teaching training)
Section B—Teaching styles index: collect teachers ‘perceptions 11 items per dimension
about students’ learning preferences Total—44 items
Section C—Barriers to technology acceptance: identify teachers’ OE—8 items
propensity to use technology in learning PF—7 items
TTF—7 items
SF—4 items
Total—26 items
Section D1—Identify the current teaching/learning methodologies SOC—9 items
most appropriate to the curricular context of the SOC, JSC and JSC—16 items
FOC, for each of its curricular areas FOC—14 items
Section D2—Identify the new teaching/learning methodologies
best suited to the curricular context of the SOC, JSC and FOC, for
each of its curricular areas
Table 3 Characterization of the pilot test sample

Type Age Gender Branch Effective Years of Teaching Course Number of
service time teaching training valid
samples
Teacher 36 < age < Female Navy = 16 < YoS < < 1 year Yes = CPOS = 24
40 = 20% = 4% 16% 20 = 16% = 24% 32% 100%
41 < age < Male = Army = 21 < YoS < 1 < years No = CEMC =
45 = 60% 96% 52% 25 = 44% <3= 68% 68%
52%
age > 51 = Air Force 26 < YoS < > 3 years CPOG =
20% = 32% 30 = 20% = 24% 48%
YoS > 51 =
20%
Mean = Mean =
44.28 25.68
Min = 38 Min = 19
Max = 55 Max = 37
Student age < 36 = Female Navy = YoS < 16 = Yes = CPOS = 124
31.5% = 17.7% 38.7% 26.6% 32% 81.5%
36 < age < Male = Army = 16 < YoS < No = CEMC =
40 = 35.5% 82.3% 30.6% 20 = 43.5% 68% 9.7%
41 < age < Air Force 21 < YoS < CPOG =
45 = 10.5% = 30.6% 25 = 6.5% 8.9%
46 < age < 26 < YoS <
50 = 8.9% 30 = 11.3%
age > 51 = YoS > 51 =
13.7% 12.1%
Mean =
19.81
Mean = Min = 9
39.94
Min = 33 Max = 35
Max = 54
Table 4 Learning style profiles reported

Sample Perception: Input: visual Processing: Understanding: N
sensory (%) (%) active (%) sequential (%)
Students SOC 91 88 70 77 101
Students JSC 92 100 75 67 12
Students FOC 55 82 64 27 11
Sample Content: Presentation: Student Perspective: N
concrete (%) visual (%) participation: sequential (%)
active (%)
Teachers SOC 62 60 61 60 24
Teachers JSC 50 58 58 59 17
Teachers FOC 67 64 60 78 11
464
Table 5 Learning style profiles reported

Sample Perception: sensorial—intuitive Input: visual—verbal Processing: active—reflexive Understanding:
sequential—global
Mod/Str Mild (%) Mod/Str Mod/Str Mild (%) Mod/Str Mod/Str Mild (%) Mod/Str Mod/Str Mild (%) Mod/Str
Sen (%) Int (%) Vis (%) Ver (%) Ati (%) Ref (%) Seq (%) Glo (%)
Students 65 31 4 75 20 5 34 53 13 42 50 8
SOC
Students 75 17 8 83 17 0 33 67 0 17 75 8
JSC
Students 18 55 27 73 9 18 9 82 9 9 55 36
FOC
Sample Content: concrete—abstract Presentation: visual—verbal Student participation: Perspective: sequential—global
active—passive
Mod/Str Mild (%) Mod/Str Mod/Str Mild (%) Mod/Str Mod/Str Mild (%) Mod/Str Mod/Str Mild (%) Mod/Str
Con (%) Abs (%) Vis (%) Ver (%) Ati (%) Pas (%) Seq (%) Glo (%)
Teachers 33 54 13 50 17 33 50 17 33 13 83 4
SOC
Teachers 12 82 6 24 41 35 12 82 6 12 76 12
JSC
Teachers 18 82 0 45 45 9 18 82 0 9 91 0
FOC
N. A. R. S. Loureiro and L. A. B. dos Santos
Table 6 ILS results conversion scale
Table 7 Index of learning/teaching styles—global

Dimensions Perception Input Processing Understanding
Categories Sensory Intuitive Visual Verbal Active Reflexive Sequential Global
Students moderate moderate mild mild
Dimensions Content Presentation Student Participation Perspective
Categories Concrete Abstract Visual Verbal Active Passive Sequential Global
Teachers mild mild mild mild
Table 8 Index of learning/teaching styles—by gender index

male moderate strong mild mild
Students
female moderate moderate mild mild
male mild mild mild mild
Teachers
female moderate moderate mild moderate
Table 9 Index of learning/teaching styles—by branch of the Armed Forces

Navy moderate moderate mild mild
Army moderate moderate mild mild
Students
Air
moderate strong mild mild
Force
Navy mild moderate mild mild
Army mild moderate mild mild
Teachers
Air
mild mild mild mild
Force
Table 10 Index of learning/teaching styles—by age group

51-55 moderate moderate mild mild
Students 41-45 moderate moderate mild mild
51-55 mild mild mild mild mild
46-50 mild mild mild mild
Teachers 41-45 mild mild mild mild
Table 11 Learning/teaching styles index—by course

SOC moderate moderate mild mild
Students FOC mild moderate mild mild
JSC moderate strong mild mild
SOC mild mild mild mild
Teachers FOC mild mild mild mild
JSC mild mild mild mild
Table 12 Learning/teaching styles index—by effective time of service

Students 21-25 moderate moderate mild mild
Teachers 21-25 mild mild mild mild
Table 13 SOC results on teaching methodologies
Teaching Methodologies
Traditional ICT-based
– Lesson/ – Lecture/ – Directed – Individual – Exercises of – Digital – Continuing – Case – Team – Gamification
Lesson- Conference/ Discussion/ Application Computer Book Education Study Based
Debate Lecture Learning Work/Group Assisted Online Learning
Series Discussion/ Application Command (TBL)
Topic Work Posts
Discussion (CPX/CAX)
– Resource 1 2 3 1 2 3
Management
– Communication 3 1 2 2 3 1
and Leadership
– Logistics and 1 2 3 1 2 3
Operational
Research
– Military 1 3 2 1 2 3
Use of Information and Communication Technologies …
Organization
– Military 1 3 2 3 2 1
Strategy
– Public 1 2 3 1 2 3
International
Law
– Military History 1 2 3 1 2 3
of Portugal
– Military Units 2 1 3 1 3 2
Framework
– Military 1 2 3 3 2 1
Operations
Planning
467
Table 14 JSC results on teaching methodologies
468
Teaching methodologies
– Lesson/ – Lecture/ – Directed – Exercises of – Digital – Continuing – Case – Team Based
Lesson- Conference/ Discussion/ Computer Book Education Study Learning
Debate Lecture Learning Assisted Online (TBL)
Series Discussion/ Command
Topic Posts
– Organizations 2 3 1 1 2 3
Management
– Resource 1 3 2 1 2 3
Management
– Command and 2 3 1 3 2 1
Leadership
– Communication 1 3 2 3 1 2
and Public
Relations
– Public 1 3 2 1 2 3
International Law
– Geopolitics 2 3 1 2 1 3
– International 3 2 1 1 2 3
Context
– War and Peace 3 1 2 1 2 3
(continued)
Table 14 (continued)
Teaching methodologies
– Lesson/ – Lecture/ – Directed – Exercises of – Digital – Continuing – Case – Team Based
Lesson- Conference/ Discussion/ Computer Book Education Study Learning
Debate Lecture Learning Assisted Online (TBL)
Series Discussion/ Command
Topic Posts
– Military History 1 2 3 1 2 3
– Strategic Planning 1 3 2 2 1 3
– Strategic Studies 2 3 1 1 2 3
– Doctrinal 2 3 1 2 1 3
Operations
Framework
– National 1 2 3 2 3 1
Operations
Planning
– Operational 2 3 1 2 1 3
Management
– Exercise Planning 1 2 3 1 2 1
and Conduct
469
Table 15 FOC results on teaching methodologies
470
– Lesson/ – Lecture/ – Directed – Exercises of – Academic – Digital – Continuing – Case Study – Team Based
Lesson- Conference/ Discussion/ Computer Seminar/ Book Education Learning
Debate Lecture Learning Assisted Panel/ Online (TBL)
Series Discussion/ Command Symposium
Topic Posts
– The Strategic 3 2 1 1 2 3
Programming Model
– Threats and Conflicts 2 1 3 2 3 1
– Strategic Objectives 3 2 1 2 3
and Interests
– Means and Strategic 3 2 1 3 2
Instruments
– Strands of the 3 2 1 2 3
National Strategy
– Organizational 3 2 1 3 2
Management and
Resource
Management
– Superior 3 2 1 2 3
Organization of the
Armed Forces
– Administrative Law 2 3 1 2 3
(continued)
Table 15 (continued)
– Lesson/ – Lecture/ – Directed – Exercises of – Academic – Digital – Continuing – Case Study – Team Based
Lesson- Conference/ Discussion/ Computer Seminar/ Book Education Learning
Debate Lecture Learning Assisted Panel/ Online (TBL)
Series Discussion/ Command Symposium
Topic Posts
– Communication and 1 3 2 2 3
Media Relations
– Command and 3 2 1 3 1 2
Leadership
– Economy and 2 1 3 1 2 3
Finance
– Joint and Combined 2 3 1 2 3 1
Operations
– Operational Planning 2 1 3 3 1 2
Exercise
– Operational Activity 2 3 1 3 1 2
of the Armed Forces
471
References
1. Garrido, A.: Errar é humano! A vivência de erros e seus efeitos na produção oral sob a per-
spectiva do aluno de Inglês como Língua Estrangeira.pdf (2006)
2. Lewin, C., McNicol, S.: Criar a Sala de Aula do Futuro: conclusões do projeto iTEC (2014)
3. Moskos, C.C., Williams, J.A., Segal, D.R.: The Postmodern Military: Armed Forces After the
Cold War. Oxford University Press on Demand (2000)
4. Exército Português: Manual Didático (MD) 240-01 Qualidade da Formação (2014)
5. Downes, C.: Rapidly evolving, digitally-enabled learning environments: implications for insti-
tutional leaders, educators and students. In: Doughty, R., Wells, L., and Hailes, T. (eds.) Inno-
vative Learning: A Key to National Security. pp. 101–132. Army Press (2015)
6. Denning, P.J., Higgins, S.L.: Being in uncertainty: cultivating a new sensibility in military
education. In: Doughty, R., Wells, L.I., Hailes, T. (eds.) Innovative Learning: A Key to National
Security. pp. 133–154 (2015)
7. Neal, D.J.: Technology and its impact on defense/security thinking and learning intervention
issues. In: Doughty, R., Wells, L.I., Hailes, T. (eds.) Innovative Learning: A Key to National
Security. pp. 155–170 (2015)
8. Schatz, S., Fautua, D.T., Stodd, J., Reitz, E.: The Changing Face of Military Learning. The
Army Press
9. Foon Hew, K., Sum Cheung, W.: Use of Web 2.0 Technologies in K-12 and Higher Education:
The Search for Evidence-Based Practice (2013)
10. Bower, M., Hedberg, J., Kuswara, A.: A framework for Web 2.0 learning design. Educ. Media
Int. 47, 177–198 (2010)
11. Felder, R., Silverman, L.: Learning and teaching styles in engineering education. Eng. Educ.
78, 674–681 (1988)
12. Pinel, S.: Bringing It All Together: Literacy, ICT and the 21st Century Skills
13. Low, M.L.: Technology That Will Shape Education In 2017
14. Felder, R., Spurlin, J.: Application, reliability and validity of the index of learning styles. Int.
J. Eng. Educ. 21, 103–112 (2005)
15. Soloman, B.A., Felder, R.: Index of Learning Styles Questionnaire, pp. 1–5. NC State Univer-
sity. North Carolina State University (1996)
16. Soloman, B.A., Felder, R.: Index of Learning Styles Questionnaire (1999)
17. Felder, R.M., Soloman, B.A.: Learning Styles and Strategies (2012)
18. Felder, R., Silverman, L.: Learning and Teaching Styles In Engineering Education (2002)
19. Davis, F.D.: A Technology Acceptance Model for Empirically Testing New End-User Infor-
mation Systems: Theory and Results (1985)
20. Venkatesh, V., Smith, R.H., Morris, M.G., Davis, G.B., Davis, F.D., Walton, S.M.: User Accep-
tance of Information Technology: Toward a Unified View (2003)
21. Gu, X., Zhu, Y., Guo, X.: Meeting the “digital natives”: understanding the acceptance of
technology in classrooms. Educ. Technol. Soc. 16, 392–402 (2013)
22. Kim, C., Jahng, J., Lee, J.: An empirical investigation into the utilization-based information
technology success model: Integrating task-performance and social influence perspective. J.
Inf. Technol. 22, 152–160 (2007)
23. Lee, M.C.: Explaining and predicting users’ continuance intention toward e-learning: an exten-
sion of the expectation-confirmation model. Comput. Educ. 54, 506–516 (2010)
24. Goodhue, D.L.;, Thompson, R.L.: Task-technology fit and individual performance (1995)
25. Ajzen, I.: The theory of planned behavior. Organ. Behav. Hum. Decis. Process. 50, 179–211
(1991)
26. Ajzen, I.: The theory of planned behavior. Orgn. Behav. Hum. Decis. Process. 50, 179–211
(1991)
27. Lewis, W., Agarwal, R., Sambamurthy, V.: Sources of influence on beliefs about information
technology use: an empirical study of knowledge workers. MIS Q. 657–678 (2003)
28. Thompson, R., Compeau, D., Higgins, C., Lupton, N.: Intentions to use information technolo-
gies: an integrative model. In: End User Computing Challenges and Technologies: Emerging
Tools and Applications: Emerging Tools and Applications vol. 18, pp. 79–101 (2007)
29. Creswell, J.: Research Design: Qualitative, Quantitative, and Mixed Methods Approaches
(2013)
30. Quivy, R., Campenhoudt, L.: Manual de Investigação em Ciências Sociais. Gradiva, Lisboa
(2008)
31. Creswell, J.: Educational research: Planning, Conducting, and Evaluating Quantitative and
Qualitative Research. University of Nebraska–Lincoln (2012)
32. Bryman, A.: Quantity and Quality in Social Research. Taylor & Francis e-Library, 2004, New
York (1988)
Using Augmented Reality for Learning
Naval Operations
Anacleto Correia , António Gonçalves and Marielba Zacarias
Abstract Augmented reality (AR) is a technology that extends the perception of

human reality by allowing, for example, inserting virtual objects in real environments.
With augmented reality, learning can be more effective with graphics and interactivity
in the manipulation of content. With this work, we intend to apply the AR technology
in a process of developing a solution in learning context. For this purpose, different
types of devices were analyzed in order to select the type of equipment with better
characteristics for use in teaching/learning solutions. An agile development process
was tailored to be used in the development of AR applications. The process was then
instantiated in the development of an AR solution for learning purposes of ships’
maneuvers in formations in the context of naval operations. The effectiveness of
the AR solution was then validated through statistical tests of data collected from a
survey to a sample of users.
Keywords Augmented reality · Mobile applications · Learning application ·

Handheld display · Naval operations
1 Introduction
The augmented reality (AR) aims for a more interactive communication, closer to a
real environment, such as the simulation of the actual presence of people and objects,
or the presentation and manipulation of virtual objects close to the interlocutor with
whom the communication is done. The characteristics that define AR are the com-
bination of real and virtual, real-time interactivity, and the 3D experience [1]. With
AR, it is possible to improve the human senses by enriching the real world with
virtual information (dynamic images, spatial sounds, and haptic sensations), gener-
A. Correia (B) · A. Gonçalves

CINAV—Alfeite, 2810-001 Almada, Portugal
e-mail: cortez.correia@marinha.pt
M. Zacarias
Universidade do Algarve, FCT/DEEI· Gambelas, 8000-139 Faro, Portugal

https://doi.org/10.1007/978-981-13-9155-2_37
ated by computer in real time and positioned in space through special technological
devices [1]. So, AR aims at greater integration between the user and the display
device. This symbiosis intends more easiness in information sharing of presented in
the devices, namely with greater authenticity in the projection of game scenarios and
teaching/training contents [2].
The benefits obtained with AR are translated on an improved performance of
everyday tasks when people wearing AR device access, for example, for an on-screen
projection of contextualized information that help them to perform complex tasks.
In real or simulated war scenarios, the potentialities of AR are evident, namely in the
situational awareness of a conflict scenario and in support of the decision-making
process of the forces in the operations theater.
In the navy, the garrison of a ship needs to train in different scenarios, as closer as
possible to real situations in order to obtain a high degree of readiness. The training
on board of a ship enables the consolidation of knowledge, previously acquired,
through practice. One of the important skills to exercise is the maneuvers of the ship
in the context of a naval force.
Training is thus a basic requirement for the readiness of the ship’s crew. When the
training is performed in real environment, it is usually carried out at the stations on
board, whether the ship is part of a fleet or is standalone. Being this the ideal training
situation, however, there are obstacles to its full realization, namely the cost that is
involved navigation of several ships in open sea. An alternative to overcome this
drawback is the training in simulators or in simulated environments. As examples of
the creation of total or partially virtualized scenarios, we can have a ship (isolated or as
part of a fleet of virtual or real ships), training procedures of the ship or fleet regarding,
among others, damage limitation, battle stations for one or multiple threats (aerial,
submarine, and surface), maneuvers and evolutions, tactical operations, refueling
operations, or navigation in restricted waters.
This paper intends to report a work carried out aimed at the construction of an
AR application, to support the teaching and learning of vessel’s maneuvers in naval
formations. The article was developed in six sections. In this section, we presented the
motivation for the theme. In the second section, the technical characteristics of AR
are presented, as well as some of the existing applications in the context of the armed
forces. The methodology used for the development of the AR application is also
presented. The third section describes the process of collecting functionalities from
the target users and details the steps developed in the construction of the proposed AR
solution. The fourth section presents the process of validation of the built solution,
based on a survey designed to obtain feedback from users. Finally, in the last section,
we present the conclusions and the future work.
Using Augmented Reality for Learning Naval Operations 477
2 Context of Augmented Reality
2.1 AR Systems and Devices
Nowadays, it is possible to visualize virtual information over real objects or scenarios,

in a context of AR, through two main techniques: optical combination or video
mixing. Through optical combination, the user has a direct view of the virtual object;
that is, the object is generated on the environment that the user observes. In video
mixing, as the name implies, there are two moments in the process of composing the
final image: (1) video recording, which captures the real environment, and (2) the
subsequent addition of objects on the video that, in the end of the mixing process, will
be presented to the user, thus merging the real environment and the virtual objects
[3].
AR display devices are classified into three categories: handheld displays, spatial
displays, and head-attached displays [3]. This devices’ classification is based on how
the user interface is performed, or, in other words, how the user uses the devices.
Although the devices have the same objectives, they use different technologies and
methods of visualization—varying their optical, mechanical and electronic compo-
nents—with each of the devices being more appropriate for specific situations. In
addition, the distance of the viewer to the real object is variable on the different
devices.
In the context of the present work, we chose to use handheld displays as preferred
devices, since they are highly versatile equipment for learning, capable in most
of them to support AR technology with acceptable quality, thus conferring a good
experience of use. In addition, the fact that most students have a mobile phone makes
it easier to spread out the application. Although the head-attached displays are able
to ensure a more immersive environment, they still have the drawback of their higher
cost.
2.2 AR Marine Applications
The AR technology offers applications for purposes both recreational and profes-
sional, related, the latter, with military areas. AR began by demonstrating its potential,
especially in the area of aviation, both military and civil. The U.S. Air Force took
the first steps in using this technology with the creation of teams to work on the
creation of systems to assist the pilot during the flight and in the definition of firing
angles. With the heads-up display (HUD) system, it was possible for the pilots to
make a more precise firing line on the targets, performing the necessary calculations
for the effect [4]. This system displays the information in a transparent fixed area,
with the observer having only one viewing angle for information reading. In military
aviation, head-mounted displays (HMDs) are replacing HUDs with other advantages
and features, such as an allowed higher viewing angle [4].
In the case of the navy, there are numerous developments in this field. A helmet for
divers allows the clarification of the information in an underwater environment, as
well as the help and explanation in the assembly of equipment [5]. Another project,
GunnAR [6], allows a remotely transmitted firing order to appear on the visor of the
helmet intended for machine gun operators aboard ships. Also, for land forces there
are helmet projects for aiding and simplifying tactical scenarios, making training
more realistic [7].
In the specific area of navigation, AR applications were developed for navigation
aids (see Fig. 1). With the common objective of making the navigation more secure
and avoiding accidents at sea, these programs mainly aim to identify dangers and
ships in transit, allowing a better clarification of the situation. This type of aids
facilitates the work of the ship, mainly the larger ones, especially in situations of
greater traffic such as the entry and exit of ports [8]. Another project is the one of
building a ship bridge equipped with AR technology, capable of notifying warnings
to navigation, dangers in the area and meteorological warnings, in order to guarantee
the safety and facilitate the crossing of the ships, especially in situations of increased
danger as in arctic navigation [9].
In areas important for navigation and subsistence of a ship at sea, assistance can be
provided with the support of AR goggles allowing technicians an efficient installation
or assembly of intended equipment (e.g., antennas) [11].
Fig. 1 AR application for sailing (Source [10])

2.3 Development Methodology
In an AR project, it is necessary to adopt a methodology that allows the definition of

the users’ requirements and the construction of the solution. In multimedia software
development projects, the use of agile methodologies by application development
teams is often due to the positive results that are generally associated with it [12].
Because the development of AR solutions presents very similar contours to the
creation of multimedia software, the Scrum (see Fig. 2) was used in the realization
of the AR solution of the present work.
Thus, the construction project of the AR solution has started with a list of needs,
that is, the requirements that had been considered in the development of the solution,
being this list the execution’s guide of the project. The evaluation and estimation of
the effort and time spent in each task, presented in the list, were also carried out, in
order to verify whether the activities were feasible in the forecasted execution time.
When the list was completed, the cycle of the sprints began, ending when the proof
of concept was completed. Each sprint began with its planning, in which the tasks to
be performed by each individual during the sprint were outlined. The preparation of
a Scrum board was the solution for accounting of time. In this board, three columns
were represented: (1) the work developed; (2) in development; and (3) what was
lacking to develop. The goal at the end of each sprint was that all tasks to be located
in the third column.
Fig. 2 Scrum methodology (Source [13])

Frequent meetings were held with the project coordinator to monitor the work.
At the end of each sprint, a meeting was held with all the elements that contributed
to the sprint, team and users, where the result obtained in the sprint was presented.
Finally, the completed solution was been demonstrated to users for experimentation
and feedback. This feedback will contribute to the subsequent development of the
work in order to improve, in the following sprints and in a new cycle, aspects less
achieved in the solution to date.
3 Solution Development
The AR application being build was intended to be a proof of concept for learning
ship maneuvers in orderly formations, according to ATP I Vol. II [14]. The aim of
the application was also, besides teaching the trainees about maneuvers of ships, a
demonstration of the potential of AR on naval training.
In order to ensure that the requirements of the solution were aligned with the
needs of the learning process, a requirements’ collecting was made among seasoned
students of a navy academy. The information was collected through brainstorming
sessions with answers obtained, validated later through a questionnaire.
The collected requirements, for the application of AR, to be developed were as
follows: (a) the minimum number of vessels, in order to make the movement of
vessels and the final formation disposition noticeable, should be 3 vessels; (b) the
perspective (position and viewing angle) to observe the evolution of the formation
should be flexible, allowing different positions and angles (bow, stern, and cross);
(c) the virtual image should emerge from a sheet, allowing to be observable from all
angles; (d) the user should be able to interact with the virtual objects; (e) it should be
possible to select different weather conditions without affecting the visibility of the
objects in the scenario or the response time of the application; (f) the representation
of the guide vessel should be different from the other ships of the formation, since the
latter takes different behaviors from the other vessels in the different formations; (g)
before the ships move, they should take a position at an initial in-line formation and
the virtual ships maneuver, adopting trajectories to occupy the selected formation,
with each vessel acting independently; (h) the type of mark used to trigger the display
of the virtual formation should be a mark inserted imperceptibly in the scenario; (i)
the characteristics of the devices that users prefer to use are head-mounted displays
and handheld displays (smart phones), because they are portable.
Based on the user requirements raised, sketches, on the form of use cases, were
drawn up regarding the maneuvers of the virtual vessels in formations. The movement
of the ships was to be triggered after the reading, by the smart phone, of artificial
mark specific of each use case.
An example of a use case, the outline of which is shown in Fig. 3 is the FORM
E formation, where the vessels maneuver in order to arrive at the depicted final
formation. Starting from a straight formation, ships will slightly open the bow and
hold it, passing to a final position on the front ship’s wing. The first ship of the
formation, which will be the guide ship, keeps its movement and bow. Paired vessels
will be placed on the port side of the guide, with a 4° opening in the wing. The
odd ships will form with a 2° aperture in the starboard wing of the guide ship. In
the sketch presented in Fig. 3, the ships appear viewed from above; however, in the
developed application, they can be seen from all positions, including the vertical.
Based on the Scrum methodology, the solution was built through several sprints,
in which it was possible to improve the functionality and graphic design of the
application, with the contribution and feedback received from the users.
The requirements collected were implemented in an AR application developed
with the support of the Unity platform with Vuforia [15]. This integrated development
environment (IDE) supports multimedia development and possesses a wide range of
functionalities that facilitate the creation of AR/VR products.
The implementation of the solution was based on the following technical options:
(a) creation of the ship formation by mixing the real and virtual scenarios (see Fig. 4);
(b) recognition of the marks of the formations residing in a database (see Fig. 5);
(c) movement of vessels in the formations; (d) precipitation and rough sea in pre-
defined scenarios (see Fig. 6); (e) observation of the formation within the mark
detection distance (see Fig. 4).
Fig. 3 Formation of ships in FORM E

Fig. 4 AR solution: FORM E visible in the application running on Android phone
Fig. 5 AR solution: vessels aligned at the formation corresponding to their mark
4 Results’ Validation
In the end of the development process, it was intended to verify if the built solution
satisfies the requirements defined by the users. For this purpose, data obtained from
a survey of 40 respondents were analyzed, most of them, seasoned students from the
naval academy. Most of the respondents had already accompanied naval operations on
the bridge of a ship, where they can have a better visual perception of the maneuvers
of the ships.
In an initial phase of the survey, the knowledge of the respondents was scored
regarding how well they knew the maneuvers and evolutions, without consulting any
supporting material. Through a set of questions, where one could visualize images of
Fig. 6 AR solution: scenario with rain and rough sea
different naval formations, it was intended that the respondents were able to match the
images and the name of the formation. Most of the answers were incorrect, revealing
lack of consolidated knowledge, of the sample of respondents, regarding the depicted
maneuvers and evolutions of ships.
The following questions of the survey were answered after a previous visualization
of the AR solution and observation of the evolution of the formations evolving. It
was intended to verify whether the respondents understood the operation with the AR
solution, had seized the presented information, and improved their knowledge on the
subject. After viewing the solution, the correct response rate increased considerably.
The percentage of correct answers was above 80%. The results seemed to indicate
effectiveness in the transmission of the concepts through the AR solution and greater
knowledge capture by the respondents.
In order to statistically validate the respondents’ answers, we proceeded to the
development of hypothesis tests. It was intended to test, if it is right to generalize to
the overall population, the average opinion of the answers obtained in the sample.
As random variable of the statistical distribution, it was considered the answer to
questions posed questions, with values ranging from 1 (weak) to 6 (very good). For
the questions posed, it was intended to validate statistically if the mean of responses
was above the value 3 (satisfies).
In the presence of a sample and ignoring the mean and the variance of the popu-
lation, the t-test was used in hypotheses tests. The tests were unidirectional, with an
alpha significance level of 0.05. With these assumptions, we tested the hypothesis
regarding the question: Is the AR built solution effective for learning naval opera-
tions?
In order to evaluate the statistical significance of the results, hypothesis tests were
performed (see Table 1), to the answers obtained to the following questions in the
Table 1 Statistics of T -test

hypothesis tests, for AR
solution validation 1P 2P 3P
Average 4.85 5.15 4.65
Variance 1.207692 0.64359 1.053846
Responses 40 40 40
Hypothesis of average 3 3 3
difference
gl 39 39 39
Stat t 10.64691 16.94977 10.16542
P(T <= t) one-tailed 2.1E−13 6.91E−20 8.02E−13
Critical t one-tailed 1.684875 1.684875 1.684875
P(T <= t) two-tailed 4.21E−13 1.38E−19 1.6E−12
Critical t two-tailed 2.022691 2.022691 2.022691
survey: (1P) the visualization of the images in the AR application improves the
knowledge about the movement of ships in formations? (2P) the realism of the AR
application improves the learning of formations? (3P) the AR solution demonstrates
with quality the evolution of the formations?
To reject H 0 (α ≤ 3), it would be necessary to have t > critical value, that
is, t > 1.685, which happens for all the questions, so H 0 was rejected. Like-
wise, P one-tailed must be less than alpha (0.05) to reject H 0 , which is also
true for the three questions. It can therefore be concluded that for the col-
lected sample, H 0 should be rejected for the three questions, and therefore,
H 0 should also be rejected for Q1. The results allowed to conclude that one
cannot reject the hypothesis that the constructed AR solution is effective for the
teaching of naval operations.
5 Conclusion
This paper describes the work carried out in order to build an AR application to
support the teaching and learning of vessel maneuvers in naval formations. In the
context of the present work, a handheld display (smart phone) was used as an AR
visualization device, due to the versatility of this type of equipment for its purpose,
namely to be able to withstand AR technology with acceptable quality and good user
experience. In order to collect user requirements and develop the AR solution, the
agile methodology Scrum was used, given its versatility, namely for the construction
of multimedia applications. As a platform for the development of the AR solution,
was used Unity with Vuforia, due to the functionalities that it makes available for the
construction of games and multimedia applications. The built solution was validated
through a survey of a sample of 40 users. The answers obtained were the object of
statistical analysis to evaluate the susceptibility of generalization of the results, which

proved to be promising. Future work intends to iterate the development of the proof
of concept with new sprints and extend the AR development to new applications in
the naval context.
Acknowledgements The work was funded by the Portuguese Ministry of Defense and by the
Portuguese Navy/CINAV.
References
1. Azuma, R.T.: A survey of augmented reality. Presence: Teleoper. Virtual Environ. 6(4),
355–385. ACM Digital Library (1997)
2. Peddie, J.: Augmented Reality, Where We all Live. Springer International Publishing, Tiburon
(2017)
3. Bimber, O., Raskar, R.: Spatial Augmented Reality: Merging Real and Virtual Worlds. AK
Peters/CRC Press (2005)
4. Aukstakalnis, S.: Pratical Augmented Reality. Addison-Wesley, Crawsfordville (2016)
5. Wheeler, D.: U.S. Navy developing new augmented-reality diver’s helmet. https://bit.ly/
2SG0L4J. Accessed Apr 2018 (2016)
6. Owens, K.: The Navy tests new augmented reality helmet, automation. https://bit.ly/2GA4oCQ.
Accessed Apr 2018 (2017)
7. Seck, H.: Augmented reality training systems destined for marine schoolhouses. https://bit.ly/
2E6EncD. Accessed Apr 2018 (2016)
8. Baldauf, M., Procee S.: Augmented reality in ships bridge operation. In: ISIS 2014—Interna-
tional Symposium Information on Ships, Hamburg (2014)
9. Wingrove, M.: Augmented reality is coming to ship bridges. https://bit.ly/2DsXD2E. Accessed
Apr 2018 (2018)
10. Azadux: The UX development story. http://azadux.com/sailing/. Accessed Apr 2018 (2018)
11. Russel, K.: Cobham explores augmented reality for maritime satcom. https://bit.ly/2I4ZFLC.
Accessed Apr 2018 (2017)
12. Schwaber, K., Sutherland, J.: The Scrum Guide. Scrum Alliance (2011)
13. Medlock, J.: A short introduction to the scrum methodology. https://bit.ly/2THpnqV. Accessed
Apr 2018 (2018)
14. NATO, ATP I Vol. II, Allied Maritime Tactical Signal and Maneuvering Book (2016)
15. Unity. https://unity3d.com. Accessed Apr 2018 (2018)
16. Elvas, F.: Realidade Aumentada Aplicada a Panoramas Táticos. M.Sc. Thesis, Portuguese Naval
Academy. https://bit.ly/2Ducn10. Accessed Dec 2018 (2018)
Communication of the Quality of Higher
Education Institutions of Ecuador
Guillermo Santa María, Carina Rey, Lilian Molina and Arturo Clery
Abstract The objective of this work was to determine the fulfillment of the attribu-
tions and functions of the Ecuadorian agency, which regulates the management of
quality assurance of higher education, declared in its Communication Plan, through
an exploratory investigation of the web pages of all regulated universities. For this
purpose, standards were designed from the Plan, which collected the pertinent infor-
mation through a scale. The results are classified based on six elements that the
Plan itself highlighted, namely communication management, its dissemination and
positioning, the image of the authorities, links with other universities, management
with international institutions and the empowerment of the mission and institutional
vision. Each of these elements presents its own characteristics and is analyzed in the
present work.
Keywords Communication · Quality · University
1 Introduction
Quality, as a concept, has been developed in response to the demands of contexts.

There is agreement, among specialists, that it is a complex concept, relative and
consequently impossible to conceptualize univocally [13].
G. S. María
Universidad de Guayaquil, Guayaquil, Ecuador
e-mail: guillermo.santamarias@ug.edu.ec
C. Rey
Universidad de Barcelona, Melcior de Palau, Barcelona, Spain
e-mail: carina.rey@ub.edu
L. Molina (B) · A. Clery
Universidad Estatal Península de Santa Elena, Santa Elena, Ecuador
e-mail: lmolina@upse.edu.ec
A. Clery
e-mail: clery@upse.edu.ec

https://doi.org/10.1007/978-981-13-9155-2_38
488 G. S. María et al.
The application of the theories of quality to higher education in Latin America,

after initial resistance, has been assumed as necessary and essential.
From a vision of public policies, the existence of a new pact between the State
and the universities is now admitted: the accountability of the universities through
the fulfillment of certain quality standards [11].
During the second half of the last century and the first two decades of the present,
HEIs in Latin America, adjusted their reality to the evolution experienced by this
concept. In the 1990s, the issue of quality as an ingredient of reforms to higher
education appears on the agenda of Latin-American countries [1].
In Ecuador, the Constitution that came into effect in 1998, determined, in order
to ensure the quality objectives, that Higher Education Institutions—IES, would be
obliged to socially account, and for this purpose established an autonomous system
of evaluation and accreditation. As a result, the Higher Education Law issued in
April 2000, in accordance with the constitutional mandate, contemplates a National
System of Evaluation and Accreditation, which would be directed by a National
Council for Evaluation and Accreditation, CONEA.
In order to achieve quality in higher education, the CONEA considered a dynamic
network of interactions: the quality of the input (input), the quality of the process
(output), and the quality of the result (outcome), arguing that in that network it was
necessary to take into account some elements, such as the quality of the educational
impact in the context, the quality of the offer, the curriculum, the demand, and the
quality of the design and the implementation.
In 2008, with the coming into force of the twentieth Constitution of the Repub-
lic, it was determined that the higher education system would be governed by the
principles of responsible autonomy, co-government, equality of opportunities, qual-
ity, relevance, integrality, and self-determination for the production of thought and
knowledge, within the framework of the dialogue of knowledge, universal thought
and global scientific and technological production.
With this Constitution, the governance of the system is in charge of a public entity
of planning, regulation and internal coordination of the system, and the relationship
between its various actors with the Executive Power, this is called Higher Education
Council—CES; and, a second public technical entity for accreditation and qual-
ity assurance of institutions, careers and programs, which is called the Council for
Evaluation, Accreditation and Quality Assurance of Higher Education—CEAACES,
which cannot be made up of representatives of the institutions that are the objects of
regulation.
In 2010, the new and current Organic Law on Higher Education—LOES is
approved, and the purpose of the regulation are to define its principles and guar-
antee the right to quality higher education that strives for excellence [8]. It is not the
right to any higher education; the text of restrictive form places the word “quality”.
Admission to public institutions of higher education began to be regulated through
a system of leveling and admission, to which all applicants submit themselves; is
in charge of the Secretariat of Higher Education, Science, Technology and Inno-
vation—SENESCYT, in coordination with the Ministry of Education, regarding its
Communication of the Quality of Higher Education Institutions … 489
articulation between the bachelor’s degree or its equivalent and higher education,
seeking only the entrance of students that meet minimum standards.
With the validity of the LOES, the State regains control of the university system,
while initiating efforts for the public policies to constitute instruments of communi-
cation between the State and universities [3], and a way through which rulers seek to
change the direction, in a positive way, of the conditions of poverty and marginality
of the population [6].
The LOES enforceable, from 2010 to August 2018, defines the principle of quality
in its Article 93, as “the constant and systematic search for excellence, relevance,
optimal production, transmission of knowledge and development of thought through
self-criticism, external criticism and permanent improvement.”
The CEAACES until 2018, in addition to organizing the processes of improve-
ment, had in its task a clear orientation toward the purification. The purge ended
with the approval by the National Assembly of the Organic Law of Extinction of
Universities and Polytechnic Schools, suspended by CEAACES.
The CES, motivated by the resolution that approved the regulation for the des-
ignation and powers of the board members of the trust called “Mandate 14—more
quality,” defined that period as the final closure of a stage that represented the exis-
tence of institutions that did not meet the quality parameters [9], was also considered
a way to guarantee the right to quality higher education and establish mechanisms
to ensure social accountability, distribution and efficient use of public resources in
favor of the institutions of the higher education system.
In a second reform to the LOES (the first one was carried out in December 2016),
promoted by a new Government in August 2018, quality is redefined as “the contin-
uous, self-reflective search for improvement, assurance and collective construction
of the culture of higher educational quality with the participation of all levels of
higher education institutions and the higher education system, based on the balance
of teaching, research and innovation and the link with society, guided by relevance,
inclusion, the democratization of access and equity, diversity, responsible autonomy,
integrality, democracy, the production of knowledge, the dialogue of knowledge, and
citizen values.”
This new definition establishes that quality should be guided by the principles,
which are already determined by the norm as natural to the system: relevance, respon-
sible autonomy, integrality and dialogue of knowledge, and also incorporates new
guiding principles of quality: inclusion, equity, diversity, democracy, and citizen
values.
This definition is close to that of the United Nations Educational, Scientific and
Cultural Organization—UNESCO, which at its Regional Conference, held in the city
of Havana in 1996, agreed to define quality as “the adequacy of the being and the to
do of higher education, to its duty to be.” Where the being must be understood as
contained in the vision, mission, principles, and values of the university and what to do
corresponds to the operative and procedural aspects. These last approach universities
to the proposed ideals, through systems of learning, research, management, self-
evaluation, and external evaluation, linked to the context [5].
Extrapolating the concept of UNESCO to the Ecuadorian University System of

2018, it can be said that the being is contained in the principles that appear in Article
351 of the current Constitution and those declared in Article 93 of the LOES of 2018,
the to do is subject to the self-evaluations in HEIs and the advance or retreat toward
these ends, should be monitored by CEAACES.
2 The Evaluation System of Higher Education in Ecuador
An autonomous system of evaluation and accreditation was established in Ecuador

with the constitutional norm of 1998, of coordination and cooperation with the
National Council of Higher Education—CONESUP. It is the first occasion in which
a system is declared to address, officially, in its integrality, the higher education [10].
Up until that moment, the Ecuadorian State had maintained a traditional relationship
with its universities. University autonomy was the main concern and, in some cases,
it had become an alibi to avoid social accountability. It was inevitable that the State
initiate efforts to articulate a response to the expansion of enrollment, the prolif-
eration of new centers of higher education and the political, economic, and social
changes that the very dynamics of Ecuadorian society promoted.
The Higher Education Law of the year 2000 indicated that this system would be
directed by the National Council for Evaluation and Accreditation—CONEA and
that it would be governed by its regulations. In 2002, the General Regulations were
issued and the actions of the system were made viable. Four years had elapsed since
the mandate of the constitutional norm.
The National System of Evaluation and Accreditation of Higher Education, was
in charge of institutional self-evaluation, external evaluation, and accreditation. This
process began with the institutional request to CONEA by the university or poly-
technic school that decided to start the process. If solitude was not found, the process
did not start.
By 2008, the year in which Mandate 14 of the Constituent Assembly mandates
CONEA to prepare the report on the level of institutional performance of Ecuadorian
HEIs, from a total of 62 universities: 5 were accredited, 13 were in the process of
evaluation external, 18 had carried out a self-assessment exercise and delivered a
report, 11 were in the design stage of the self-evaluation project, 9 had not started
the self-assessment process, and 7 did not comply with the legal time [14].
The slowness with which the system had operated, the absence of relevant changes,
more allegations of corruption in the delivery of academic titles, were elements of
concern and would give rise to the decision of the National Assembly expressed in
Constituent Mandate No. 14.
The CONEA was composed of the following: 2 academics appointed by the Pres-
ident of the Republic, from three candidates drawn up by the University Assembly
and qualified by the CONESUP, 2 academics appointed by the National Congress,
from lists prepared by the University Assembly and qualified by the CONESUP; 1
member appointed by the Minister of Education and Culture; 1 member appointed by
the National Federations of the Professional Associations; 1 member designated by

the state science and technology agency; and 1 member appointed by the Federation
of the Chambers of Production.
The integration reflects a predominance of academics from the organizations
subject to evaluation and accreditation, as well as a marked interference by the
regulatory body, CONESUP.
After 10 years, from 1998 to 2008, the new constitutional norm puts the accent
on the organisms that will govern the higher education system. One of them will
be a technical body for accreditation and quality assurance of institutions, careers
and programs, and cannot be made up of representatives of the institutions subject
to regulation. The denomination of the constitutional text system, on this occasion,
is used to substantiate higher education; we must remember that in the Constitution
of 1998 the denomination of system was used to substantiate the evaluation and
accreditation processes.
This new Constitution of 2008, imperatively states that, within five years, all HEIs,
their careers and programs and postgraduate programs are evaluated and accredited,
and warns that if they do not pass the evaluation and accreditation, they will be left
out of the system of higher education.
In replacement of the CONEA, the LOES issued in 2010, will give rise to the
Council for the Evaluation, Accreditation and Quality Assurance of Higher Educa-
tion—CEAACES. For the first time, the assurance of the quality in the denomination
of the organism appears. The qualified evaluation and accreditation system is attached
into the faculties of the new Council.
CEAACES implements the Twentieth Disposition of the Constitution and carries
out the evaluations and accreditations established for the universities, Polytechnic
Schools and their extensions. It also complies with the CONEA report and evaluates,
within 18 months after the enactment of the Law, the universities and Polytechnic
Schools that were placed in category E.
The constitutional norms establish that the Law defines the coordination mech-
anisms for the higher education with the Executive Power. The LOES to comply
with this provision creates the National Secretariat for Higher Education, Science
and Technology and Innovation, SENESCYT, to coordinate actions between the
Executive Function and the institutions of the Higher Education System.
This Secretariat is headed by an official appointed by the Executive. In this way, the
presence of the State, through the officials it designates for the regulation, planning,
coordination, evaluation, accreditation, and assurance of the quality of the higher
education system, dominates almost the entire spectrum of the system.
Before 2008, the State had avoided having a predominant presence, with the
enforcement of the new Constitution and the new LOES, the State’s role became
central in the integration of the entities that govern the so-called higher education
system
The reform of the LOES of August 2018 incorporates SENESCYT as part of
the higher education system. And the task to ensure the effective compliance of
the principle of quality is delivered to the Inter-institutional Quality Assurance
System—SIAC.
This new system is integrated by the CES, the CACES, and the IES. The self-
evaluation is given a permanent role and turned into the cornerstone, which must be
carried out in reference to the purposes of the institutions. Thus, it must be understood
that the missions and visions of the universities would become the referents of the
aforementioned self-process. The process will be accredited by CACES, who will
plan and coordinate the operations.
In this way, the State remains at the forefront of the system even when the
self-evaluation processes, which are initiatives of the university institutions, acquire
centrality.
Harmonizing the dynamics of state agencies with universities will be a process
to be built. Internal processes, such as self-evaluations, and external ones, such as
accreditations, must interact, mesh, and function effectively. This interaction is on
the horizon of the Ecuadorian system.
3 Quality Assurance
The CONUEP in 1989 carried out the first diagnostic study of the Ecuadorian uni-
versity, and one of the basic problems detected was the lack of a higher education
system; instead there existed a set of independent institutions, with weak cooperation
mechanisms.
The island scenario of university institutions to which the previous study alludes
had in the defense of university autonomy a unifying link: Defending autonomy was
indispensable and there ones that when, invoking it, attributed to it thaumaturgical
properties and associated it with quality were few.
Nine years passed so that, in the Constitution issued in 1998, the constitutional
norm warns that to ensure quality objectives, universities should perform an account-
ability; for its fulfillment, an evaluation and accreditation system is formed.
This is the first normative evidence of concern for quality and its assurance.
Two years later, in 2000, the Higher Education Law ratifies that the objective
of the evaluation and accreditation system is to ensure quality, encourage ongoing
processes to improve academic quality and management, and to do so, appeals to
the processes of institutional self-evaluation, external evaluation, and accreditation
[12].
The following objective, established in the Law, was for Ecuadorian society to
be informed of the institutional performance of higher education establishments in
order to guarantee quality, promoting their depuration and improvement.
The wording of the norm’s text suffers from lack of clarity; however, its concern
for the depuration and improvement is noticed. The 14th Mandate, in 2008, includes
this intention and expresses it in its resolutions.
After the 21 universities, which in 1989 had been subject to evaluation by the
CONESUP, the creation of private and public universities had been approved in a
number close to 70. An increase of more than 200%.
In June 2006, only the Universidad del Azuay had fulfilled all the requirements
and was accredited.
A context of application of neoliberal policies in the university system, of com-
mercialization of higher education and indifference of the State in relation to defini-
tions and regulations, had introduced chaos and discredit [7]. Reconciling with the
concepts of quality was urgent as well as the need for depuration and improvement.
Reality, the political and economic context imposed their logic, while the
norm remained a dead letter and organisms limited themselves to contemplate the
decadence.
The 14th Mandate reflects the disappointment of the process entrusted to CONE-
SUP and CONEA and obliges CONESUP to prepare a report on the academic and
legal situation [2], and CONEA is required to submit a technical report to comply
with the depuration and the improvement as established in Article 91 of the Law of
Higher Education of that time.
In November 2009, CONEA delivers its report, which consists of an initial classi-
fication of Ecuadorian universities and a radical recommendation: Purge the Ecuado-
rian university system of the group of universities that did not meet the basic con-
ditions of a university institution [4]. These universities had been created during the
1990s and the early years of the twenty-first century.
The LOES of 2010 establishes for these universities identified in the 14th Man-
date, type E, the prohibition of offering new academic programs and/or postgraduate
courses; at the same time, they have to be evaluated within a period of 18 months.
This new evaluation was in charge of the new body that the LOES created in substitu-
tion of the CONEA: the Council of Evaluation, Accreditation and Quality Assurance
of Higher Education—CEAACES.
The assurance of quality defines it as actions that guarantee efficient and effective
management. And this definition extends to the Councils and accrediting bodies.
4 Final Considerations
Quality assurance appears as a component of the name of the new body; it is assumed
as one more process of the system. The concern in the previous stage had been:
evaluation and accreditation. Now, the assurance is constituted in a process to be
implemented, it stops being an objective without concretion. It is essential to shield
what has been achieved in order to avoid backing down, and in parallel, it is forced
to reach new levels of improvements.
In the period from 2010 to 2018, CEAACES carries out evaluations of careers,
exams for qualification of professional practice, depurates extensions, participates in
interventions in universities, re-categorizes, and accompanies processes of improve-
ment plans, institutional strengthening, and quality.
And finally, in 2018, we find ourselves with a quality assurance which results
show an asymmetric reality hidden in the accreditation of 55 universities out of a
total of 59, extensions 39, and 56 evaluated careers. The qualification exams for
professional practice have been applied in careers that could compromise the public
interest, putting at risk essentially the life, health, and safety of the citizens; between
2015 and 2018, the qualification exams for professional practice—EHEP have been
applied on nine occasions for students of medical careers, seven times for the career
of dentistry, and twice for students of the nursing career.
The reform introduced, in August 2018 to the LOES, has the HEIs whose quality
has been accredited from the asymmetry, as a base line. Thus, on the official website
of CEAACES, there are 18 universities in category C, 28 in category B, and 8 in
category A.
The universities categorized C, which correspond to 30% of the HEIs, should
accelerate their improvement processes in order to narrow the existing gap; in a
similar situation are categorized B of the A.
On the other hand, the norm did not contemplate or contemplates a reassurance
process for the system. Neither CONEA nor CEAACES from 2010 to 2018 have
been evaluated technically by independent bodies to ensure that their work effectively
meets acceptable standards.
Acknowledgements Thanks are due to the Universidad de Guayaquil and the Universidad Estatal
Península de Santa Elena for their administrative and economic contributions to the development
of this research, through its Public and Social Policy Observatories.
References
1. Cielo, C., Ospina, P., Vega, C.: Reforma y renacimiento: conversacones docentes sobre la
reforma universitaria en Ecuador. Universidad Andina Simón Bolívar, Quito (2016)
2. Ecuador: Legislación comparada de nueve países de América Latina y españa. Consejo
Nacional de Educación Superior, Quito (2009)
3. Ecuador: Ley Orgánica de Educación Superior (2010)
4. Ecuador: La evaluación de la calidad de la Universidad ecuatoriana: la experiencia del Mandato
14. Consejo de Evaluación, Acreditación y Aseguramiento de la Calidad de la Educación
Superior—CEAACES, Quito (2014)
5. Gutiérrez-García, R., Sainz-Venegas, P., Contreras-Cornejo, A., Lugo-Carvajal, M., Preciado-
Alatorre, T., Ornelas-Jiménez, J.: Universidad Sustentable. Instituto Tecnológico y de Estudios
Superiores de Occidente, Jalisco (2017)
6. López, M., Cifuentes, L., Palma, A.: La gestión de las políticas públicas en las universidades:
una aproximación a su caracterización en el Ecuador. Dominio de las Ciencias, 447–464 (2018)
7. Lucas, K.: De los bancos de papel a las universidades de garaje: impactos de las políticas
neoliberales en la educación superior en Ecuador. Consejo de Educación Superior CES, Quito
(2015)
8. Martínez, A., Vásquez, P.: La importancia de la evaluación en las instituciones educativas con-
forme a la nueva Ley Orgánica de Educación Superior en el Ecuador. Revista Iberoamericana
de Evaluación Educativa, 174–180 (2012)
9. Molina, L.: Evaluación de las biblotecas universitarias ecuatorianas: análisis del entorno y
propuesta de un catálogo de indicadores. Universidad Estatal Península de Santa Elena, La
Libertad (2015)
10. Rama, C.: La educación a distancia y virtual en Ecuador. Universidad Técnica Particular de
Loja UTPL-Virtual Educa, Loja (2015)
11. Toscanini, M., Aguilar, A., García, R.: Diagnóstico de las políticas públicas de la educación
superior en el Ecuador. Revista Cubana de Educacaión Superior, 161–178 (2016)
12. Valdés, M., Villegas, T.: Responsabilidad social universitaria: fundamento para la gestión socio-
cultural. Revista Cubana Educación Superior, 55–62 (2017)
13. Villarroel, C.: Gerencia, planificación y evaluación universitarias. La Revista Venezolana de
Educación Educere, 513–522 (2005)
14. Villavicencio, A.: De la universidad funcional a la universidad de la razón. Universidad Andina
Simón Bolívar, Quito (2013)
The Four Dimensions of the GDPR
Framework: An Institutional Theory
Perspective
Isabel Maria Lopes, Teresa Guarda and Pedro Oliveira
Abstract The EU general data protection regulation (GDPR) is the most important
change in data privacy regulation in 20 years. The regulation will fundamentally
reshape the way in which data are handled across every sector. The organizations
had two years to implement it. Despite this, it has been observed that, in several
sectors of activity, the number of organizations having adopted that control is low.
This study aimed to identify the factors which condition the adoption of the GDPR
by organizations. Methodologically, the study involved interviewing the officials in
charge of information systems in 18 health clinics in Portugal. The factors facilitating
and inhibiting the implementation of GDPR are presented and discussed. Based
on these factors, a set of recommendations are made to enhance the adoption of
the measures proposed by the regulation. The study used Institutional Theory as a
theoretical framework. The results are discussed in light of the data collected in the
survey, and possible future works are identified.
Keywords Regulation (EU) 2016/679 · General data protection regulation ·

Institutional Theory · Health clinics
I. M. Lopes (B) · T. Guarda

Centro ALGORITMI, Universidade do Minho, Braga, Portugal
e-mail: isalopes@ipb.pt
T. Guarda
e-mail: tguarda@gmail.com
I. M. Lopes
UNIAG (Applied Management Research Unit), Polytechnic Institute of Bragança, Bragança,
Portugal
I. M. Lopes · P. Oliveira
School of Technology and Management, Polytechnic Institute of Bragança, Bragança, Portugal
e-mail: pedrooli@ipb.pt
T. Guarda
Universidad Estatal Peninsula de Santa Elena—UPSE, La Libertad, Ecuador

https://doi.org/10.1007/978-981-13-9155-2_39
498 I. M. Lopes et al.
1 Introduction
In recent years, data protection has become a forefront issue in cybersecurity. The
issues introduced by recurring organizational data breaches, social media and the
Internet of Things (IoT) have raised the stakes even further [1, 2]. The EU GDPR,
enforced from May 25, 2018, is an attempt to address such data protection. The
GDPR makes for stronger, unified data protection throughout the EU.
Not all data is protected by the GDPR, it is only applicable to personal data. This
is defined in Article 4 as follows [3]:
“personal data” means any information relating to an identified or identifiable natural per-
son (‘data subject’); an identifiable natural person is one who can be identified, directly
or indirectly, in particular by reference to an identifier such as a name, an identification
number, location data, an online identifier or to one or more factors specific to the physical,
physiological, genetic, mental, economic, cultural or social identity of that natural person.
In a study [4] conducted in 57 health clinics (the survey was sent to 190 clinics,
but only 57 gave an effective answer) about whether they had started or completed
the process of implementation of the measures laid down in the GDPR, 39 (69%)
gave a negative answer and 18 (31%) said to have started or completed the adoption
of such measures. Among the 18 clinics which are in the implementation stage, 7
(39%) consider that the process is completed, while the remaining 11 (61%) still
have some measures to implement. These figures seem insufficient when bearing in
mind that the GDPR had already come into force six months before the conduction
of this study.
Thus, the following two research questions were formulated in order to guide the
research work:
1. Which dimensions condition the adoption of the GDPR?
2. Which recommendations might be put forward so as to enhance the adoption of
the GDPR by health clinics?
The answer to the first question aims to know the positive and negative con-
ditioning dimensions influencing the adoption of the GDPR by health clinics. In
possession of these elements, it will be relevant to produce a set of recommendations
which enable the adoption of the GDPR measures by clinics.
The structure of the present work consists of an introduction, followed by a desk
review on the general data protection regulation and the desk review of Institutional
Theory. Section 4 focuses on the research methodology, identifying the target popu-
lation and the structure of the survey. The results of the study are presented in Sect. 5,
and Sect. 6 consists of the conclusions drawn from the study. Finally, the limitations
of this research work are identified and possible future studies are proposed.
The Four Dimensions of the GDPR Framework: An Institutional … 499
2 General Data Protection Regulation
Personal data are owned by a natural person and are as such inalienable. However,
with the development of technology and the Internet, personal data are becoming
increasingly exposed and accessible to those who could use them for their own,
unlawful purposes [5]. The GDPR is a cornerstone regarding personal data treatment
regulation, which aims to respond to the new challenges in the field of data protection
generated by the evolution of technology and the globalization of markets.
It originated from the European Union package on the reform of data protection
and came into force directly and mandatorily on May 28, 2018, impacting signifi-
cantly the life of organizations.
Both the European Parliament and the Council of the European Union saw the
need to implement a more coherent and solid data protection framework, anchored
in a rigorous application of the rules, since it is crucial to generate the trust needed
for the development of digital economy in the whole internal market.
The obligatory adaptation of the organizations to the GDPR will imply a set of
legal, technological, and functional changes with a direct impact on the daily life of
the organizations, as a result of their increased responsibility with the data protection
subjects, which has been reinforced by the new legislation. On the other hand, the
transfer of responsibilities from national authorities to the organizations obliges them
to prove, at all times, full compliance with the legislation. Organizations are subject
to high fines when non-compliance is detected [6].
The GDPR aims to protect the personal data of EU residents and not just secure it.
There is no single solution for the GDPR due to its complex remit and enterprise-wide
impact. The GDPR compliance requires a 360° perspective [7]:
1. There is no silver bullet to the GDPR:
• The GDPR is the most complex piece of regulation in recent times, and it
impacts everybody.
• The depth and breadth of activity required to demonstrate and maintain com-
pliance are significant.
• There is no single solution to the GDPR.
2. The GDPR impacts the enterprises:
• The GDPR is an enterprise-wide business problem, and it requires enterprises
solutions.
• Approached in the right way, the GDPR will allow organizations to build trust
with customers/stakeholders and positively differentiate within the digital age.
3. Fragmented Approach:
• The GDPR requires a 360° perspective, and organizations that approach this
regulation from a single viewpoint are liable to an increased risk exposure.
• There are over 20 defined areas to the GDPR, all of which require a distinct
solution.
Fig. 1 Four dimensions of the GDPR framework [7]
• Cognizant has a comprehensive framework that covers all the required GDPR
areas and which can be traced back to each individual component of the leg-
islation.
The GDPR implementation program covers four distinct dimensions which are
critical to the GDPR compliance: technology, data, process, and people (see Fig. 1).
Organizations must work on these four dimensions, each of which having its own
specificities, starting the process by “preparing,” followed by “protecting,” and finally
by “maintaining,” so that the GDPR adoption can be a reality and organizations can
be in compliance with the regulation.
3 Institutional Theory
Changes in legislation, technology, and the economy generate modifications in the

organizational environment. In the face of this, the search for innovation represents
one way for the survival of organizations. The success of the organization is then
measured by the capacity to survive, change, and anticipate the market needs [8].
Therefore, organizations gradually institutionalize organizational practices in order
to face new realities, which cannot be faced using the previously existing organiza-
tional practices.
The Institutional Theory considers the processes through which structures (e.g.,
frameworks, rules, norms, and routines) are established as trustworthy guidelines
for social behavior. Also, it accounts for the way these elements are created, spread,
adopted, and adapted throughout time and space, as well as the way they fall into
decline and disuse [9]. Institutions may be conceived as high resilient social structures
that enable and constrain the behavior of social actors and that provide stability and
meaning to social life [10–12].
The authors [13] outlined the processes inherent to institutionalization as consist-
ing of four stages, namely innovation, habitualization, objectification, and sedimen-
tation. The institutionalization process starts at “Innovation,” which occurs due to
external forces such as technological change, legislation, or market forces. In this
sense, the word innovation means structural rearrangements or new organizational
practices aimed at solving organizations’ problems. Following this comes a sequen-
tial process of three stages which enables the evaluation of the institutionalization
degree of a certain social reality.
In an organizational context, the process of “Habitualization” involves the creation
of new structural arrangements in response to specific organizational problems or
sets of problems, shaped through policies and procedures of a specific organization or
set of organizations with similar problems. Hence, this is the pre-institutionalization
stage.
After the solution for the problem has been generated, it is possible to move
on to the “Objectification” process, which accompanies the spreading of the new
structure, expanding its use. Objectification implies the development of a certain
degree of social consensus regarding the structure and its growing adoption, based
on that consensus, by the organization. This process configures the semi-institutional
stage.
The stage in which institutionalization is complete is called “Sedimentation,” and
it is characterized by the adoption of the structure or organizational practice by the
whole organization for a long period of time.
Figure 2 presents a diagram with a summarized overview of the causal critical
forces in the institutionalization process.
To better understand this diagram, we applied another concept also proposed by
[14] to the stages of institutionalization to observe the institutionalization process
across different stages or levels. For those authors, the institutionalization process
Fig. 2 Component process of institutionalization [14]

occurs in three distinct stages (or phases): the pre-institutional, semi-institutional,

and total institutionalization stages.
As shown in Fig. 2, in the pre-institutional stage, causal forces external to the
organization arise from changes in technology, legislation, and market forces, which
may lead to the introduction of innovations in organizations. These forces may also
lead to the creation of new habits.
The author [12] discusses the distinction between studies focusing on the creation
of institutions and studies focusing on the change of institutions. The first ones focus
on the process and the conditions which give place to new rules, understandings, and
practices.
The second ones examine the way a set of beliefs, norms, and practices is attacked,
becomes “non-legitimate,” or falls into disuse, being then replaced by new rules,
ways, and scripts. Deep down, these two processes are related, as the institutional
creation implies the change of the existing institutions and the institutional change
implies the creation of new institutions.
4 Research Methodology
The survey (first study) was sent to 190 clinics, but only 57 gave an effective reply,
which corresponds to a response rate of 30%. The sample subjects were selected
randomly based on the kind of clinic and their location distributed throughout the 18
inland Portuguese districts as well as Madeira and the Azores [4].
Considering the research questions previously formulated, and in articulation with
the selected research method (Institutional Theory), a primary data collection tool
emerged: the interview.
For a better understanding of the results, we can group the clinics into three
clusters (Table 1):
Table 1 Clusters of GDPR implementation

Cluster Stage Number of clinics n (%) Number of clinics
interviewed
1 In compliance with the 7 (12.2) 6
regulation
2 Which are implementing 11 (19.3) 6
the measures imposed by
the regulation
3 Which are not in 39 (68.5) 6
compliance with the
regulation
Table 2 Clinics surveyed Type of clinic Clinics surveyed Clinics interviewed

and interviewed
Nursing 12 4
Dental 16 6
Ophthalmology 4 1
Medical and 15 4
diagnostics
Orthopedics 3 1
Physiotherapy 7 2
• Cluster 1—Clinics in compliance with the regulation;

• Cluster 2—Clinics which are implementing the measures imposed by the regula-
tion;
• Cluster 3—Clinics which are not in compliance with the regulation.
Altogether, 18 health clinics were interviewed, distributed equitably among the
three clusters (each cluster contributed to six interviews).
Table 2 shows the characterization of the clinics involved in the study according
to their type.
The structure of the survey resulted from a desk review on personal data protection
and the study of the legal framework Regulation (EU) 2016/679 of the European
Parliament and of the Council of April 27—General Regulation on Data Protection
[3].
5 Results
Compliance with the GDPR must be a priority and must be seen by organizations
as an opportunity. Since we are talking about health clinics, therefore, dealing with
“sensitive data,” such priority becomes even more pressing.
From the work conducted within the 18 clinics under study, the following measures
can be recommended in order to improve the adoption of the propositions laid down
in the regulation, as far as the components of the institutionalization process are
concerned (Fig. 3).
The major challenge is faced in the habitualization stage, since it is when a bigger
quantity of factors considered to be critical occurs within the institutionalization
process as far as effort, cost, and complexity are concerned.
With regard to the four dimensions presented in Fig. 1, their taking into account
increases the patients’ trust toward the health clinic.
The dimension—technology, responsible for data architecture, is crucial in order
to monitor activities in an automatized way, managing and protecting contents and
data. Thus, compliance with the GDPR implies the implementation of data protection
technology.
Fig. 3 Component process of institutionalization/adoption of GDPR
In the dimension—data, where management and security are conducted, and bear-
ing in mind that most data treatment activities are based on the use of information
systems and automatized means, it is necessary to assess the systems used, the levels
of access and the security measures implemented.
In the dimension—process, responsible for the consent of rights, the processes
are honed so as to respond to the data holders’ rights and manage data processing.
In the dimension—people, responsible for governance and oversight, it is manda-
tory to involve all the stakeholders a cultural change toward data protection by
all—with well-defined rules and responsibilities.
The conjunction of these actions to enhance the implementation of the GDPR in
the health clinics can be summarized in six essential points:
1. Review periodically—data protection is not a destination, it’s a journey! So,
review on a regular basic and correct your course accordingly.
2. Raise awareness—conduct staff training and awareness sessions. Most breaches
occur due to staff ignorance so make them aware and mitigate the risk.
3. Review policies and related documents—review existing policies and create new
ones where needed. These policies allow you to formulate processes and proce-
dures for staff to follow.
4. Make a plan—based on the audit and risk assessment, determine a roadmap
to achieve compliance. While it is important to address high-risk areas, do not
ignore the low hanging fruit. Small, easy wins can get the project off to a positive
start!
5. Identify and assess privacy-related risks—assess the risks associated with how
data are being processed. Compile a risk register.
6. Conduct an audit—know your data—how you got it; who can access it; where it
is stored; how long you keep it, etc. Create a list of recommendations to address
areas of concern.
6 Conclusions
This paper brought forward guidelines which are believed to enhance the institution-
alization of measures imposed by the GDPR in organizations. We also argue that the
use of Institutional Theory as a support to the interpretation of the adoption stage
of implementation by organizations and as a support to the projection of guidelines
which enhance the institutionalization of these GDPR measures in organizations
represents a promising means for research.
In Portugal, there is still some vagueness concerning some of the issues regard-
ing data protection, due to the fact that the Portuguese legislation has not yet been
approved (note that although the GDPR is a EU regulation and therefore is directly
enforced in Portugal and in all member states, the truth is that certain issues demand
a legal definition by the Portuguese legislator).
A considerable effort is being made by companies to be in compliance with the
GDPR. However, it is a long path and the instructions regarding the four dimensions
(technology, data, process, and people) must be properly implemented and, above all,
maintained bearing in mind that periodic improvements are necessary. This process
is never completely closed, and reviews must be conducted whenever necessary.
As a possible future work, we suggest assessing how far the certification of the
information security management system by ISO 27001 grants companies’ compli-
ance with the GDPR, since the implementation of an information security manage-
ment system by a company must ensure that all the relevant controls of risk contain-
ment associated with confidentiality, integrity, and availability are implemented and
kept functional.
Acknowledgements UNIAG, R&D unit funded by the FCT—Portuguese Foundation for the
Development of Science and Technology, Ministry of Science, Technology and Higher Education.
Project n.º UID/GES/4752/2019.
This work has been supported by FCT—Fundação para a Ciência e Tecnologia within the Project
Scope: UID/CEC/00319/2019.
References
1. Mäkinen, J.: Data quality, sensitive data and joint controller ship as examples of grey areas in
the existing data protection framework for the Internet of Things. Inf. Commun. Technol. Law
24(3), 262–277 (2015)
2. Nurse, J.R.C., Creese, S., De Roure, D.: Security risk assessment in Internet of Things systems.
IEEE IT Prof. 19(5), 20–26 (2017)
3. European Parliament and Council, Regulation (EU) 2016/679 of the European Parliament and
of the Council of 27 April 2016, Official Journal of the European Union (2016)
4. XXXX
5. Skendzic, A., Kovacic, B., Tijan, E.: General data protection regulation—protection of personal
data in an organization. In: 41st International Convention on Information and Communication
Technology, Electronics and Microelectronics, pp 1370–1375 (2018)
6. Da Conceição Freitas, M., Mira da Silva, M.: GDPR in SMEs, vol. 2018, pp. 1–6. In: 13th
Iberian Conference on Information Systems and Technologies (2018)
7. West, I.: The big scan thing!—How the EU General Data Protection Regulation (GDPR) will
affect your business! https://www.slideshare.net/CraigShipley1/digital-enterprise-festival-
birmingham-130417-ian-west-cognizant-vp-data-management-the-implications-of-the-eu-
global-data-protection-regulation-on-every-business-and-their-digital-service-providers. Last
accessed 1 Dec 2018
8. Brown, S.L., Eisenhardt, K.M.: Competing on the edge: strategy as structured chaos. Harvard
Business School Press, Boston (1998)
9. Scott, W.: Institutional Theory. Encyclopedia of Social Theory, pp. 408–414. Thousand Oaks,
Sage (2004)
10. DiMaggio, P. Powell, W.: Introduction. In: Powell, W.W., DiMaggio, P.J. (eds.) The New
Institutionalism in Organizational Analysis, pp. 1–38. University of Chicago Press, Chicago
(1991)
11. North, D.: Institutions, Institutional Change and Performance. Cambridge University Press,
Cambridge (1990)
12. Scott, W.R.: Institutions and Organizations: Ideas and Interests, 3rd edn. Sage, Thousand Oaks
(2008)
13. Tolbert, P.S., Zucker, L.G.: The institutionalization of institutional theory. In: Handbook of
Organization Studies. Sage, London (1996)
14. Tolbert, P.S., Zucker, L.G.: A institucionalização da teoria institucional. In: Clegg, S., Hardy, C.,
Nordy, W (eds.) Handbook de estudos organizacionais (pp. 196–219). Tradução de Humberto
F. Martins e Regina Luna S. Cardoso, v.1. Atlas, São Paulo (1999)
Metrics and Indicators of Information
Security Incident Management:
A Systematic Mapping Study
Alyssa Cadena, Franklin Gualoto, Walter Fuertes, Luis Tello-Oquendo,

Roberto Andrade, Freddy Tapia and Jenny Torres
Abstract The number of threats and vulnerabilities has increased rapidly in recent
years. For this reason, organizations are in need of providing improvements in their
computer security incident management (CSIM), in order to safeguard their intel-
lectual capital. Therefore, the identification and use of both metrics and indicators
are a crucial factor to manage security incidents. In this context, organizations try to
improve their level of CSIM based on standards or only according to their criteria
based on their experience. This article aims at carrying out a systematic mapping
study of academic articles conducted in this research area, in order to present a doc-
ument that describes metrics and indicators of security incidents in organizations.
The results of this work show and describe several key indicators and metrics related
to the cost, quality, and service (time) involved in dealing with such incidents. Also,
it is expected that this study serves as a strategic reference for organizations.
A. Cadena · F. Gualoto · W. Fuertes · F. Tapia

Universidad de las Fuerzas Armadas ESPE, Sangolquí, Ecuador
e-mail: akcadena@espe.edu.ec
F. Gualoto
e-mail: fgualoto@espe.edu.ec
W. Fuertes
e-mail: wmfuertes@espe.edu.ec
F. Tapia
e-mail: fmtapia@espe.edu.ec
L. Tello-Oquendo (B)
Universidad Nacional de Chimborazo, Riobamba, Ecuador
e-mail: luis.tello@unach.edu.ec
R. Andrade · J. Torres
Escuela Politécnica Nacional, Quito, Ecuador
e-mail: roberto.andrade@epn.edu.ec
J. Torres
e-mail: jenny.torres@epn.edu.ec

https://doi.org/10.1007/978-981-13-9155-2_40
508 A. Cadena et al.
Keywords Security incidents · Metrics · Indicators · Security incident

management · Key performance indicators · Empirical study
1 Introduction
According to Miloslavskaya, “Information security (IS) incidents have become not

only more numerous and diverse but also more damaging and disruptive. Preventive
controls based on the IS risk assessment results decrease the majority but not all the
IS incidents” [1]. One way to control incidents is to manage them through proper
identification and analysis of key metrics and indicators. Therefore, it is necessary
to emphasize all those metrics and indicators of computer security incidents that can
help organizations in the quest to safeguard their intellectual capital from threats.
In the literature, several related works can be found on this matter. For instance, in
[2] the authors present a method that evaluates the ability in which the organization
handles security incidents; in addition, they perform the analysis of various security
metrics based on detection, response, and cost of incidents and focus on early warning
indicators based on resilience. It is important to visualize from the technical aspect,
where the experiences obtained from the field of action are reflected when dealing
already in organizations with the management of computer security incidents; a clear
example of this is [3] in which the authors propose a methodology based on “goal,
question, metrics (GQM)” that establishes steps for the detection and resolution of
security violations; they also recognize metrics already proposed by ISO 27004:2009
and NIST 800-55. “What is not defined cannot be measured. What is not measured
cannot be improved. What cannot be improved, is always degraded” [4]. Hence, the
importance of discovering key metrics and indicators for CSIM to help measure the
effectiveness and efficiency of such management.
Under this perspective, the purpose of this work is to conduct a systematic mapping
study (SMS), which allows organizations to guide in the definition of metrics and
indicators of CSIM. To achieve this, an empirical study was carried out in three
databases: IEEE, ACM, and SpringerLink to later analyze the articles found and
translate, according to the criteria of the authors of this study, metrics, and indicators
of computer security incidents. The results obtained emphasize those focused on
cost, quality, and service.
The remainder of this paper is structured as follows. Section 2 documents the
SMS process, while Sect. 3 presents the results of the systematic mapping. On the
other hand, Sect. 4 introduces a proposal for CSIM, specifically in higher education
institutions. Finally, Sect. 5 draws the conclusions and future work.
Metrics and Indicators of Information Security Incident … 509
2 Systematic Mapping Study Process (Research

Methodology)
Petersen defines a SMS as a secondary study aimed at constructing a classification

scheme and structuring a software engineering field of interest [5]. To carry out the
study, he suggests a five-step process as follows: (1) Define the research questions;
(2) search the relevant documents; (3) select the primary studies; (4) analyze the
abstracts and extract keywords and data; and (5) map the selected primary studies.
Each of the steps in the process has an outcome, with the systematic mapping
being the final outcome of the overall processes. The steps mentioned above were
followed, and a protocol is constructed for the purpose of avoiding bias [6].
2.1 Research Questions
In order to frame the research on the problematic raised, research questions (RQs)
were established, which were defined following the objectives of this study: to ana-
lyze and define metrics and indicators of information security incidents employing
this SMS. In this way, we follow the respective guidelines leading to the RQs pre-
sented below:
RQ1: What are the metrics related to information security incidents in organizations?
RQ2: Are there indicators related to information security incidents?
RQ3: Are there Key Performance Indicators (KPI’s) related to cost, quality, and
service?
RQ4: In most metrics, what standards apply to information security incident man-
agement?
RQ5: Are there studies on SMS aimed at Information Security Incidents?
This information will lead to obtaining the metrics and key indicators of informa-
tion security incidents as well as a formal study that validates all the data shown.
2.2 Inclusion and Exclusion Criteria
For the selection of the studies, the corresponding titles and abstracts were taken into
account in which the following inclusion criteria will be denoted:
– Studies in the field of computer security;
– Studies in the field of information security;
– Articles published since 2010;
– Scientific articles and conference papers.
Within the exclusion criteria, the following parameters were considered:
– Studies that were not included in the selected databases;

– Duplicate studies;
– Course articles, books, or early access articles;
– Studies that did not present contents in English or Spanish.
2.3 Digital Libraries and Search String
The following digital libraries were considered for the literature review: IEEE Xplore,
ACM Digital Library, and Springer Link. The reason for selecting them is because
they cover a wide range of publications in the field of engineering science maintain-
ing a complete and consistent database. For conducting the initial search, we used
keywords and search terms closely related to:
– “Security information management”;
– “Security incident metrics and indicators”;
– “Key performance indicators or KPIs.”
Then, to determine the optimal search string for research, several options have
been considered taking into account that they conform with the terms as mentioned
above. The proposed inclusion and exclusion criteria were also applied generating
diverse possibilities. The first search approach can be seen in Table 1.
Finally, once the search string was optimized, we selected the one that contains
the most significant amount of relevant information and that returned a manageable
number of documents for the research. The optimized search string is shown in
Table 2.
Table 1 Search strings or chains, databases, and results

Suggested search string Academic database result
((((Information Security Incident) OR Security Information IEEE Xplore 2077
Management) AND Metrics) AND Indicators OR Key ACM 893
Performance Indicators)
Springer Link 24,126
(((Information Security Incident) OR Security Information IEEE Xplore 391
Management) AND Indicators) ACM 13,277
Springer Link 7102
((Security Information Management) AND (Informatic Security IEEE Xplore 3
OR Security Incidents) AND (Metrics OR Indicators) AND (Key ACM 2509
performance Indicators OR KPIs))
Springer Link 9
((((Security Information Management) AND Informatic Security IEEE Xplore 2180
OR Security Incidents) AND Metrics OR Indicators) AND Key ACM 27
performance Indicators OR KPIs)
Springer Link 23,941
Table 2 Selected search string

((Information Security Incidents) AND (Indicators OR Metrics OR Key Performance Indicators))
3 Results of the Systematic Mapping Study
Aiming at refining the information obtained from the search string, an information
crossing was performed in two ways: manually and using the academic databases.
By doing so, the amount of literature obtained for reviewing was reduced to 10
relevant articles as shown in Table 3. This result can also be visualized in Fig. 1
which illustrates the number of articles selected from each academic database for
conducting this study. In addition, Fig. 2 depicts the mechanism for determining the
relevance of each study. This was done by means of validation of the researchers, by
simple selection and equitable weights selection.
Note that we also analyzed the years in which the articles were indexed. This
information can be seen in Fig. 3. It reflects that, in the year 2015, a greater emphasis
was made in the generation of scientific articles focused on the analysis of metrics
and indicators of CSIM in the organizations. Then, it decreases until the year 2018
in which the tendency to study this subject increases again.
Table 3 Metrics and indicators of information security incidents: systematic mapping study
Item study title
S1 [7] “A practical experience on evaluating intrusion prevention system event data as
indicators of security issues”
S2 [2] “Forewarned is forearmed: indicators for evaluating information security incident
management”
S3 [8] “A comprehensive survey on machine learning for networking: evolution,
applications and research opportunities”
S4 [9] “Adoption of security as a service”
S5 [1] “Security operations centers for information security incident management”
S6 [10] “Information security considerations for protecting NASA mission operations
centers (MOCs)”
S7 [11] “Establishing national cyber situational awareness through incident information
clustering”
S8 [12] “The β-time-to-compromise metric for practical cybersecurity risk estimation”
S9 [13] “Integration of IT frameworks for the management of information security within
industrial control systems providing metrics and indicators”
S10 [14] “Application of security metrics to instrument systems that use distributed
processing”
2
SpringerLink
19
0
ACM
3
8
IEEE Xplore
44
0 10 20 30 40 50
Selected Articles All Articles
Fig. 1 Results of the research string application
Fig. 2 Relevance of the S1

literature considered for this 100%
study S10 80% S2
60%
S9 40%
S3
20%
0%
S8 S4
S7 S5
S6
3.5
3 3
Number of publications
2.5
2 2 2
1.5
1 1 1 1
0.5
0
2010 2011 2012 2013 2014 2015 2016 2017 2018 2019
Publication year
Fig. 3 Distribution of the publications by year

3.1 Answers to Research Questions
The selected documents, result of the SMS, were used to answer the research ques-
tions posed in Sect. 2.1, which are the focus of this research.
Regarding the research questions RQ1 and RQ2, several metrics and indicators
related to information security incidents were analyzed. However, a significant num-
ber of articles focused solely on information security were found through the search;
therefore, an exhaustive analysis was carried out, and the results are presented in
Table 4. It is worth noting that the metrics and indicators considered by the studies
analyzed are based on standards such as [14]. On the other hand, [13] is based on
the metrics proposed by other authors as [15–18]; in the same way, [12] is based on
[19], and [9] is based on indicators proposed by various authors. Conversely, articles
were found that support their indicators or metrics throughout their research process
as is the case of [1, 7, 8, 10]. However, there are several studies that use indicators
based on the REWI method [20] (e.g., [2]) and that present metrics in the form of
facts (e.g., [11]).
Regarding the research question RQ3, cost, quality, and service are three entities
of great importance in organizations, which is why the work of [1, 2, 13] focuses
on these attributes. With regard to the research question RQ4, both the metrics and
indicators of information security incidents analyzed in the standards were proposed
by their authors based on ISO/IEC 27035 [21] or NIST SP 800:61 [22], among others.
However, the indicators presented in [14] are based on the ANSI/ISA-99.01.0 1-2007
[23] and ISA-99.02.01-2009 [24] standards.
Finally, regarding the research question RQ5, there are studies on SMS, as is the
case of [2], where the authors present a method to evaluate an organization’s capacity
to manage security incidents. They mention that this method is based on the form of
response to change (resilience) and describe how to identify, select, and implement
early warning indicators for the management of information security incidents.
4 Proposal
Several institutions worldwide have presented standards for CSIM, highlighting

the standards ISO/IEC 27004:2009 (International Organization for Standardiza-
tion/International Electro-technical Commission) [25] and NIST SP 800-55 Revi-
sion 1 (National Institute of Standards and Technology) [26], as guidelines that
provide guidance and advice on the development and use of measures to evaluate
the effectiveness of the information security incident management. However, there
is the problem that as long as organizations continue to focus on isolated technology
solutions without the respective balance between processes and management, uncer-
tainty will continue to be generated as to why the increase in incidents despite the
investments made in information security. Organizations dedicated to IS manage-
ment, as is the case of higher education institutions (HEIs), propose to carry out an
Table 4 Metrics and indicators of information security incidents
514
Study Metrics Indicators

[7] – Number of alerts per week No indicators
– Number of (distinct) attackers per week
– Number of (distinct) objectives per week
– Number of (distinct) signatures per week
– Number of alerts per attacker per week
– Number of attackers per target per week
[2] – Presence and fulfillment of plans (Based on resilience)
– Incident statistics – Risk awareness
– Detection and response statistics – Responsiveness
– Consequences – Support
– Incident management cost and performance
– Culture and learning aspects
[1] No metrics – Unauthorized user on the network or shared credentials
– Unauthorized access to sensitive data, personally identifiable
information (PII), and financial data
– Unauthorized connection of the internal host (client or server) to the
Internet
– Excessive access from a single or multiple internal host to the
external malicious Web site (from known blacklists)
– Detection of malware and user activity outside of the hour (at night or
on weekends)
– Multiple logins with the same ID from different locations in a short
time
– Multiple alarms from a single host or duplicate events on multiple
computers on the same subnet over a 24-h period
– Repeated attack from a single source or on a single host
– Access to the Internet service account or to an unauthorized device,
among others
(continued)
A. Cadena et al.
Table 4 (continued)
[8] Performance metrics for precision validation [8] No indicators
– True positive rate or recall
– False positive rate
– True negative rate
[9] No metrics Indicators focused on vulnerability and threat
[10] – Number of new {incidents | vulnerabilities (discovered) | plan of No indicators
actions and milestones (opened)} in the past 30 days
– Number of open {incidents (i.e., under investigation) | vulnerabilities
(not yet mitigated) | plan of actions and milestones (not yet closed)}
in the past 30 days
– Number of closed {incident investigation(s) | vulnerabilities
(mitigated) | plan of actions and milestones} in the past 30 days
[11] Action No indicators
Damage euro
Damage score
Incident date
Incident ID
Metrics and Indicators of Information Security Incident …
System
Time until discovery
Time until report
Vulnerability date
Business area
[12] Time-to-compromise No indicators
(continued)
515
Table 4 (continued)
516
[13] Project metrics and indicators specified by PMI-PMBOK

– Program of progress of all projects in the portfolio
– The active risks and the amount of money used compared to that
budgeted for that date
Management of ITIL services related to processes, services and
technology:
– The number of closed and pending tickets for each of the processes of
the service operation stage
– The number of closed and pending tickets for the incident process
– The number of tickets closed and the total pending
– The response times of each of the Ethernet ICS nodes per month
– The percentage of uptime of each of the Ethernet ICS nodes per
month
– The number of closed tickets per day for each area administered by
ICS
– The ticket seniority times for each service associated with the
established SLA
– The percentage of compliance of each ICS engineer’s priorities for
the week associated with the established SLA
– The average time taken to complete a ticket
– The average cost of a ticket
[14] – Number of times [in a specified time interval] an error condition No indicators
occurs
– Total number and number of unsuccessful connection attempts to
authenticate devices communicating via untrusted networks per [unit
of time determined by the asset owner]
– Current number of untrusted network connections
– Number of access retries associated with a source address recorded
by each access point
A. Cadena et al.
objective evaluation of the incident records in order to find useful information that
feeds their knowledge. Based on the above, according to a study presented in 2017
by Consortium for Advanced Internet Development (CEDIA) in its article “State of
Information and Communication Technologies in Ecuadorian Universities,” in the
area of ICT security, 70% of HEIs are supported by CEDIA’s CSIRT, 8% have their
Committee, and 22% none. Similarly, 54% of the HEIs surveyed rely on CEDIA’s
vulnerability analysis tools, 22% on their own, and 24% none [27]. This shows evi-
dent shortcomings in HEIs concerning CSIM. Besides, there is a lack of studies
on information security incidents in HEIs and how these could benefit management
decision-making and the continuous improvement of the IS.
For this reason, it is proposed to perform a study on the incidents aiming at achiev-
ing cognitive learning that allows analyzing the problem from another perspective.
On the one hand, the generation of dashboards of the preeminent indicators related
to security incidents is highly advised. Besides, the discovery of behavioral patterns
of incidents through data mining techniques will provide relevant information. This
information will be transformed into knowledge that will drive better business strate-
gies that will translate into a competitive advantage for any organization. It is for
this reason that some research is planned where metrics and key indicators will be
established in the management of security incidents in HEIs in order to improve their
management, using benchmark surveys (KPI benchmark surveys). Additionally, a
business intelligence model required to evaluate these metrics and indicators will be
implemented through the “cross-industry standard process for data mining” (CRISP-
DM) or through Ralph Kimball’s methodology and where the effectiveness of the
model will finally be evaluated, leveraged with the action research methodology, to
determine the confidence level of the results.
5 Conclusions
The objective of this study was to conduct a systematic literature review of academic
articles focused on security incident management, in order to identify and describe
metrics and indicators of security incident management in organizations. To achieve
this, an empirical study was carried out on three indexed databases: IEEE, ACM,
and Springer.
The articles found were then analyzed along with their metrics and indicators.
The methodological procedure consisted of a systematic mapping study with their
respective research questions, inclusion and exclusion criteria, selection of digital
libraries, and analysis of the respective search strings. Among the main findings
were metrics and key indicators focused on cost, quality, service, products, incident
management, resilience, and early warning indicators. It should be noted that some of
them were paired with the following international standards: ISO/IEC 27035, NIST
SP 800:61, ANSI/ISA-99.01.0 1-2007, and ISA-99.02.01-2009. This information
could serve for the analysis of their behavior as input for decision-making in the
context of information security incident management in organizations. All this will
result in reducing threats, vulnerabilities, and computer risk, which will influence
the increase in the level of information security in institutions. As a future work,
we have planned to create a knowledge database to design and develop a decision
support system of the management of security incidents.
Acknowledgements The authors would like to thank the financial support of the Ecuadorian Cor-
poration for the Development of Research and the Academy (RED CEDIA) in the development of
this work, under Research Team GT-II-Cybersecurity.
References
1. Miloslavskaya, N.: Security operations centers for information security incident management.
In: 2016 IEEE 4th International Conference on Future Internet of Things and Cloud (FiCloud),
pp. 131–136 (2016)
2. Bernsmed, K., Tondel, I.A.: Forewarned is forearmed: indicators for evaluating information
security incident management. In: 2013 IEEE Seventh International Conference on IT Security
Incident Management and IT Forensics, pp. 3–14 (2013)
3. Hajdarevic, K., Allen, P.: A new method for the identification of proactive information security
management system metrics. In: 2013 IEEE 36th International Convention on Information &
Communication Technology Electronics & Microelectronics (MIPRO), pp. 1121–1126. (2013)
4. Thomson, W., Kelvin, L.: Baltimore Lectures. CJC a. Sons, Ed., London (1904)
5. Petersen, K., Vakkalanka, S., Kuzniarz, L.: Guidelines for conducting systematic mapping
studies in software engineering: an update. Inf. Softw. Technol. 64, 1–18 (2015)
6. Elberzhager, F., Münch, J., Nha, V.T.N.: A systematic mapping study on the combination of
static and dynamic quality assurance techniques. Inf. Softw. Technol. 54(1), 1–15 (2012)
7. Miani, R.S., Zarpelao, B.B., Sobesto, B., Cukier, M.: A practical experience on evaluating
intrusion prevention system event data as indicators of security issues. In: 2015 IEEE 34th
Symposium on Reliable Distributed Systems (SRDS), pp. 296–305 (2015)
8. Boutaba, R., Salahuddin, M.A., Limam, N., Ayoubi, S., Shahriar, N., Estrada-Solano, F.,
Caicedo, O.M.: A comprehensive survey on machine learning for networking: evolution, appli-
cations and research opportunities. J. Internet Serv. Appl. 9(1), 16 (2018)
9. Senk, C.: Adoption of security as a service. J. Internet Serv. Appl. 4(1), 11 (2013)
10. Takamura, E., Mangum, K., Wasiak, F., Gomez-Rosa, C.: Information security considerations
for protecting NASA mission operations centers (mocs). In: 2015 IEEE Aerospace Conference,
pp. 1–14 (2015)
11. Skopik, F., Wurzenberger, M., Settanni, G., Fiedler, R.: Establishing national cyber situational
awareness through incident information clustering. In: 2015 IEEE International Conference on
Cyber Situational Awareness, Data Analytics and Assessment (CyberSA), pp. 1–8 (2015)
12. Zieger, A., Freiling, F., Kossakowski, K.P.: The β-time-to-compromise metric for practical
cyber security risk estimation. In: 2018 IEEE 11th International Conference on IT Security
Incident Management & IT Forensics (IMF). pp. 115–133 (2018)
13. Bustamante, F., Fuertes, W., Díaz, P., Toulkeridis, T.: Integration of IT frameworks for the
management of information security within industrial control systems providing metrics and
indicators. In: 2017 IEEE XXIV International Conference on Electronics, Electrical Engineer-
ing and Computing (INTERCON), pp. 1–4 (2017)
14. Munro, J.K.: Application of security metrics to instrument systems that use distributed pro-
cessing. In: Future of Instrumentation International Workshop (FIIW), 2011, pp. 5–8 (2011)
15. Stouffer, K., Falco, J., Scarfone, K.: Guide to industrial control systems (ICS) security. NIST
Spec. Publ. 800(82), 16 (2011)
16. Rose, K.H.: A guide to the Project Management Body of Knowledge (PMBOK® Guide)—Fifth
Edition. Proj. Manag. J. 44(3), e1–e1 (2013)
17. Lloyd, V.: ITIL Continual Service Improvement (Best Management Practices). The Stationery
Office (2011)
18. ISACA: COBIT 5: A business framework for the governance and management of enterprise
IT. ISACA (2012)
19. McQueen, M.A., Boyer, W.F., Flynn, M.A., Beitel, G.A.: Time-to-compromise model for cyber
risk reduction estimation. In: Quality of Protection, pp. 49–64. Springer (2006)
20. Øien, K., Massaiu, S., Tinmannsvik, R.K.: Guideline for implementing the REWI method;
Resilience based Early Warning Indicators. SINTEF report A 22026 (2012)
21. Information Technology—Security Techniques—Information Security Incident Management.
Standard, International Organization for Standardization, Geneva, CH (2011)
22. Cichonski, P., Millar, T., Grance, T., Scarfone, K.: Computer security incident handling guide.
NIST Spec. Publ. 800(61), 1–147 (2012)
23. ANSI/ISA: Security for Industrial Automation and Control Systems Part 1: Terminology, Con-
cepts & Models. Tech. rep., American National Standards Institute/International Society of
Automation (ANSI/ISA) (2007)
24. ANSI/ISA: Security for Industrial Automation and Control Systems: Establishing an Industrial
Automation and Control Systems Security Program. Tech. rep., American National Standards
Institute/International Society of Automation (ANSI/ISA) (2009)
25. ISO/IEC: Information Technology—Security Techniques—Information Security Managemen-
t—Measurement (ISO/IEC 27004: 2009). ISO/IEC (2009)
26. Chew, E., Swanson, M., Stine, K., Bartol, N., Brown, A., Robinson, W.: NIST Special Publica-
tion 800–55 Revision 1. Performance Measurement Guide for Information Security, National
Institute of Standards and Technology, US Department of Commerce. Computer Division,
Gaithersburg, MD 20899, 8930 (2008)
27. Verdugo, R.P.: Estado de las tecnologías de la información y la comunicación en las universi-
dades ecuatorianas. CEDIA (2017)
Virtual Environment for Remote Control
of UGVs Using a Haptic Device
F. Roberto Segura, Pilar Urrutia-Urrutia, Z. Andrea Sánchez,

C. Tomás Núñez, T. Santiago Alvarez, L. Franklin Salazar ,
Santiago Altamirano and Jorge Buele
Abstract This paper presents a virtual reality environment designed for military
training personnel, focused on the remote control of unmanned land vehicles. The
environment design has been made in the V-REP software, where a prototype of
an explorer robot based on the kinematic model of a unicycle is presented. This
vehicle is attached with proximity sensors to detect obstacles and thus be able to
avoid them. Instead of operating with a conventional joystick that only allows the
use of push buttons, a haptic device with force feedback is used with which the
user experiences a more realistic immersive situation. In this context, the person can
manipulate the unmanned vehicle direction and perceive when there is a collision
with a nearby object as if it were on the site. To link the input device (Novint
Falcon) with the virtual interface, the device mathematical modelling is carried out,
and through MATLAB, the respective processing and the implementation of the
P. Urrutia-Urrutia · Z. Andrea Sánchez · L. Franklin Salazar · S. Altamirano · J. Buele (B)

P. Urrutia-Urrutia
Z. Andrea Sánchez
e-mail: ap.sanchez@uta.edu.ec
L. Franklin Salazar
S. Altamirano
e-mail: santiagomaltamirano@uta.edu.ec
F. Roberto Segura · T. Santiago Alvarez
Instituto Tecnológico Superior Guayaquil—Ambato, Ambato 180205, Ecuador
e-mail: rsegura@institutos.gob.ec
T. Santiago Alvarez
e-mail: salvarez@institutos.gob.ec
C. Tomás Núñez
CELEC EP, Baños 180250, Ecuador
e-mail: leandro.nunez@celec.gob.ec

https://doi.org/10.1007/978-981-13-9155-2_41
522 F. Roberto Segura et al.
proportional–integral–derivative (PID) control algorithm for the displacement are

made. The after-scenario questionnaire (ASQ) test is used, and a general average of
1.78/7 is obtained. Being a value close to 1, it shows the acceptance that the system
has for the users.
Keywords Haptic interface · Force feedback · Training · Teleoperation ·

Unmanned ground vehicle · Virtual reality
1 Introduction
In recent years, the advance of technology has brought with it several tools that facil-
itate the daily tasks of the human being [1]. He wants to acquire more knowledge,
explore, and innovate, and thanks to this, science expands covering more fields of
application [2]. This is how robotics was born, a topic of interest that has allowed us
to carry out several multidisciplinary researches [3]. The main interest of researchers
is to design robots with autonomy that can be used in health care, space explo-
ration, military, and rescue applications, among others [4]. Within mobile robotics,
the unmanned ground vehicles (UGV) stand out, and they can be teleoperated by
specialists and thus develop dangerous activities, where it would be risky to enter a
person [5]. Time and effectiveness, in addition to safety, of these robotic systems is
highly compared to actions developed by a human being where the aim is not intended
to supplant it, but complement it with an important instrument [6]. Additionally, by
attaching sensors and video devices to a robot, reconnaissance and exploration tasks
can be performed in known and unknown territories. The knowledge of space can
then be defined in two parts: structured spaces and unstructured spaces. On the one
hand, the path to be followed and the objects that must be avoided are already defined;
on the other hand, several characteristics that can change must be contemplated and
therefore must be studied to obtain satisfactory executions [7].
Whether directing a real or a simulated robot, the operator’s learning is very
important for the integration of new knowledge and skills. Traditionally, these vehi-
cles have been operated by control levers such as joysticks which only have push
buttons, but do not give the feeling of contact with the environment in question [8].
This is the reason why the use of haptic devices with force feedback allows the user
to feel the effect of being inside the environment where actions are executed through
a slave robot. Another important technological tool that is very popular is virtual
reality (VR). It consists in a platform that facilitates the implementation of interac-
tive environments, allowing the user to relate to objects and situations that simulate
the real world. Several areas of application of the VR are come to combine including
entertainment, education, research, rehabilitation, military applications, and others.
For example, in [9] a study is carried out in which the relationship between presence
and performance is evaluated, while a psychomotor task is executed in a virtual envi-
ronment. Although virtual simulation has been validated empirically to be effective
in the transfer of skills to the live environment, what is still unknown is the effect of
Virtual Environment for Remote Control of UGVs … 523
the user sense of presence in the simulation performance. In this way, it can be known
in a better way the behavior of the user before an immersive experience. For their part
in [10] using a VR application, war veterans with post-traumatic stress disorder are
trained so that they can obtain a job after serving in their military work. Preliminary
results showed that participants who attended 95% of the sessions showed improve-
ment in their abilities when facing a job interview and in their self-confidence.
In parallel, by merging graphic computing with other areas of knowledge, staff
can be trained to perform tasks that require hand–eye coordination, such as surgery,
maneuvers in dangerous spaces, or teleoperation. In medical area, there are some
researches like [11], where the application of virtual reality is described to optimize
the acquisition and retention of military medical skills. This system not only devel-
ops the surgical skills of the person, but also incorporates an intelligent mechanism
that recognizes the deficiencies of the skills and generates an optimal program of
adaptive training. In the proposal developed by Ibari et al. [12], a VR system is
presented for the remote control of a unicycle robot through an intuitive multimode
man–machine interface. This system manages the tracking of a path by the real robot
through the Internet and a wireless network. The result analysis shows the system
performance in terms of precision, stability and convergence, and the ability to run
from any machine connected to the Internet. In this context, the prior research car-
ried out by Tikanmäki et al. [13] presents a teleoperation system, where the user
can control semiautonomous mobile robots simulated by a joystick, replacing the
conventional video screens with an immersive virtual environment. The operator can
see and inspect the situation by collecting data from an appropriate perspective with-
out suffering motion sickness due to latency in traditional teleoperations. The user
experiences are positive when highlighting the possibility of having multiple virtual
screens to display images of the camera without experiencing dizziness and without
requiring prior knowledge. In the work of Lwowski et al. [14], two experiments were
carried out with unmanned ground vehicles (UGV). The first one is manually driven
in square form with a Vive Tracker where an estimated position error in average
odometry is visualized. Second experiment has a PID controller used to drive the
UGV in two different test patterns: a square and an eight figure. The usage of virtual
reality as a simulation environment shows as result the usefulness of the tool for
systems researchers.
Related works have shown the use of a joystick to drive simulated and real vehi-
cles but without obtaining a complete immersive experience. In this context, this
paper seeks to develop the skills and abilities of military personnel to maneuver
exploration UGVs in a semi-structured environment. To achieve this goal, a virtual
reality environment is presented, where objects that the vehicle must evade are found
in order to carry out the desired trajectory and task. In addition to having a visual
perception, it is required that the user determines the exact force that must be applied
to command a robotic mechanism based on the kinematic model of a unicycle, and
therefore, the Novint Falcon haptic device is used. For the connection between the
physical and virtual media, the MATLAB software is used, where through the kine-
matic study of the Falcon Master Robot its behavior can be simulated. Through a
PID control algorithm, the desired trajectory is tracked. The experimental tests and
results demonstrate the validity of the proposal, since the control actions are executed
efficiently, thus representing an important teaching tool.
This project is divided into six sections, including the introduction in Sects. 1
and 2 presenting the description of system. Section 3 shows the elements of system.
Section 4 talks about the implementation; Sect. 5 shows the results; the conclusions
are presented in Sect. 6.
2 Methodology
The methodology used to solve the problems encountered follows a set of hierarchical
steps shown in Fig. 1. In this figure, all the sections that will contain the methodology
to solve the problem are detailed. First, a mathematical analysis of the equations
representing the Robotic Master System is considered, analyzing the kinematics of
the device. All the necessary libraries and drivers are also considered to connect
the haptic device to the computer, so it must have to be installed the corresponding
packages for the MATLAB and V-REP software. Mathematical modeling and the
implementation of a control method allow to simulate the behavior of the device in
MATLAB, prior to the execution of real tests. Also, the control method allows the
accomplishment of tasks by the unicycle-type mobile robot. This scenario contains
primary elements and/or equipment that represents a training scenario for the skill
of the user, where the acquired information is converted to forces. The conversion
to forces allows to fulfill the purpose of this work, which focuses on presenting a
clearer perspective of the interaction between the slave robot and the environment
with which it interacts.
3 Modeling
3.1 Geometric Analysis of the Novint Falcon Kinematic

Model
Geometrically, the mechanism of the haptic device can be analyzed as a parallel

robot with a triangular shape at the base and at the end, the joints. Three reference
systems are required for analysis, where A is the reference system of the base of the
three arms, C is the reference system of the workspace, and B1−2−3 are the reference
systems relative to each arm of the haptic device.
Fig. 1 Methodology to solve the problem
3.2 Analysis of the Proposed Controller
In order to achieve the goal of teleoperation, a control law is proposed to notify, by

force feedback, the type of excitations that users must do with the haptic device. In
order to feedback an adequate force, the kinetic model of Novint Falcon is used as
shown in Eq. (1) and using inverse kinematics the Eq. (2) is obtained.
ṖC = R −1 J¯q̇(t) (1)
q̇ = R J¯−1 ṖC (2)
In this context, the control law is shown in Eq. (3).

q̇ = R J¯−1 ṖCd + PID P̃C (3)
where the expression PID represents the control algorithm to compensate for the
error that occurs in carrying out the displacement task, and it is defined in (4).
t ∂
PID( P̃C ) = kp P̃C + ki P̃C dt + kd P̃C (4)
∂t
0
Fig. 2 Experimental execution of the desired trajectory
3.3 Device Simulation
The simulated object representing the Novint Falcon allows to validate both the
modeling and the proposed PID controller. Next, the operation of the controller
is demonstrated by tracking the trajectory described by: xd = r ∗ cos( p), yd =
(2 ∗ π ) ∗ 4, z d = π ∗ sin( p), components that form a circumference. To perform the
validation of the experimental control with the haptic device is considered a control
gain of k p = 0.1, and this gain is applied for the control of the errors in x, y, z.
Figure 2 shows the simulation of the stroboscopic movement of the robotic system
to fulfill the desired trajectory, where the path to be followed and the execution that
the operative end executes are shown.
4 Develop of the Virtual Environment
4.1 Virtual Environment
The robot interaction with an environment limited by walls can allow the increase
in the dexterity of an operator located at a remote site. In this way, the use of V-REP
properties is proposed to create a room, in which multiple objects are located in
different positions. These are chairs, tables, columns, and different furniture that can
interfere with the path of the robot and should be avoided by the user.
Fig. 3 UGV simulated in the V-REP software
Created the simulation environment, a differential robot is incorporated that moves

horizontally on the plane X-Y. Its design is based on the kinematic model of a unicycle.
In Fig. 3, the configuration of this device located on the work space is indicated,
denoting the speeds that produce a variation in its displacement. These speeds are
two: linear μ (forward displacement) and angular ω (counterclockwise rotation),
when modified, allow intuitive control of the robot’s direction. To incorporate the
design into the simulation environment, the V-REP libraries of a base prototype
(Pioneer 3-DX) are used. This one has two motors (each motor is associated with a
wheel that allows it to move according to its dynamics), and ultrasonic sensors are
also incorporated to determine the presence of nearby objects.
4.2 MATLAB and V-REP Interaction
In the application development, the main step is to adapt the configuration of the
simulation software and the mathematical software. Through this configuration, both
packages are intercommunicated to test the operation of the haptic device in the
control process. To maneuver the direction of the vehicle using the Falcon Novint,
the sending and receiving of feedback data is constant and its configuration is detailed
in Fig. 4.
4.3 Control of the Mobile Robot
In order to avoid collisions between the robot and the elements of the environment,
forces are fed into the operating end of the master robot when the proximity to an
obstacle exceeds a threshold. In order to carry out the proposed task, the position
differences between the robot and the neighboring elements are calculated through the
incorporation of a virtual laser. This information allows taking actions for evasion.
The implementation of a PID control algorithm allows to determine the feedback
force to the haptic device so that the operator can avoid collisions. This prevents
damage to the robotic mechanism through the injection of proportional speeds to
Fig. 4 Flowchart of the

interaction between
MATLAB, Falcon Novint,
and V-REP simulator
each wheel (depending on the location of the obstacle in the plane). This process is
supervised by visual feedback, since a camera is placed on the front of the robot to
visualize the space where it moves, allowing to validate the feedback obtained by
the haptic device.
5 Experimental Tests
The aim to conduct a UGV within a semi-structured environment is the main require-
ment in order to carry out the teleoperation experimentation through the use of the
Falcon. Considering that the robot is in an initial position, it is proposed to perform
an exploration task and take it to a desired point. To do this, the operator must apply
linear and angular velocities through the haptic device, while experiencing force
feedback. If an object is close and a collision will be made, a force is generated,
informing the operator in a direct way, so that it makes the required corrections and
keeps getting closer to the desired point. The used hardware to perform these tests
has an Intel Core i7-7500U at a 2.7 GHz speed, with 16 GB RAM and an Ether-
net network device: NIC Fast Ethernet family RTL8101E, wireless network device:
Realtek RTL8187B supporting the Wireless 802.11g 54 Mbps. For the optical device,
it has been used the Novint Falcon with 3D Touch Workspace of 4 × 4 × 4, with
2 lbs. of force capabilities and 400 dpi of position resolution. Figure 5a shows the
system in execution, at the starting point, and in Fig. 5b, the user is shown performing
teleoperation tasks.
For the validation of this proposal, 10 tests were carried out with three different
users. The experimental data taken from the teleoperation task of the mobile vehicle
Fig. 5 a Virtual environment presented to the user, b Experimental tests performed by a user
are within a considerable margin of error; in each test, it is seen that the average
errors for x, y, z oscillate depending on the movement and position of the mobile.
Depending on the distance between the robot and nearby obstacles, the user modi-
fies the return forces to avoid collision. In Fig. 6, the correction force injected into the
haptic device is presented as a response to the interaction between the robot and the
environment on the X, Y, Z axes. Based on these variations, various errors in each of
the axes are calculated, as shown in Fig. 7. After users have completed the task, they
are given the ASQ test developed by Lewis [15] which consists of three questions,
with a rating of 1–7 each. The lower the selected score, the higher the satisfaction
of the usability of the user with his/her system. This is presented in Table 1. Results
are determined by calculating the average (arithmetic mean) of the three questions.
In this way, the test of this work obtains 1.33, 2, and 2, respectively, for three people,
so the general average was 1.78/7.
Fig. 6 Feedback forces on X, Y, Z axes as a response between the robot-scenario interaction

Fig. 7 Control errors obtained when executing the application
Table 1 ASQ test

Question Score
1. Overall, I am satisfied with the ease of completing the tasks in this scenario 1–7
2. Overall, I am satisfied with the amount of time it took to complete the tasks in this 1–7
scenario
3. Overall, I am satisfied with the support information (online help, messages, 1–7
documentation) when completing the tasks
6 Conclusions and Future Works
Through the use of a powerful mathematical software and a 3D application develop-

ment engine, it has been possible to present an intuitive and interactive technological
tool for the development of skills of military personnel. In this prototype is integrated
mathematical modeling, autonomous control, the use of a haptic device, and a sim-
ulated environment. The simulation of robots provides an optimization of economic
resources and the opportunity to change its characteristics according to the needs of
the final application. The V-REP software allows, based on a virtual environment, to
teleoperate the displacement of a UGV while the haptic device performs the feedback
forces that provide a complete immersive experience. For this, it is recommended to
work on 32-bit-based operating systems to improve compatibility with the libraries
of the device, or to develop libraries that allow to generate *.dll files that are easily
executable by 64-bit computer packages. When applying the ASQ test, the average
values obtained were part of the best range of the evaluation; it shows the virtual
environment acceptance for the user, time, complexity, and supplementary material
of the application.
After the use of this system, weaknesses were detected in the learning process, so
that let the authors to propose teaching strategies, improving the skill and reflexes of
the student. The users were often practicing and having fun, which contributes to the
stress reduction to which they are constantly subjected. As the tests are developed,
the dexterity and self-motivation increase, which is evidenced in lower error values
when maneuvering the vehicle. Based on this, the authors are already working on
future research, where the first instance is to create new environments, which include
several types of obstacles, to prevent monotony. In addition, it is priority to use the
same principle to develop more military and industrial applications and thus cover
more areas.
Acknowledgements To the authorities of Universidad Técnica de Ambato (UTA), Dirección

de Investigación y Desarrollo (DIDE), Instituto Tecnológico Superior Guayaquil—Ambato and
CELEC EP., for supporting this work and future research.
References
1. Mosey, S., Guerrero, M., Greenman, A.J.: Technology entrepreneurship research opportunities:
insights from across Europe. J. Technol. Transfer 42(1), 1–9 (2017)
2. McCarthy, J., Wright, P.: The Enchantments of Technology. Human–Computer Interaction
Series, pp. 359–373. Springer (2018)
3. Zieliński, C., Winiarski, T., Kornuta, T.: Agent-based structures of robot systems. In: The 19th
Polish Control Conference, KKA 2017, pp. 493–502. Springer (2017)
4. Kunze, L., Hawes, N., Duckett, T., Hanheide, M., Krajník, T.: Artificial intelligence for long-
term robot autonomy: a survey. IEEE Rob. Autom. Lett. 3(4), 4023–4030 (2018)
5. Choi, S.Y.: Agent-based human-robot interaction simulation model for the analysis of operator
performance in the supervisory control of UGVs. Int. J. Precis. Eng. Manuf. 19(5), 685–693
(2018)
6. Aguilar, W.G., Rodríguez, G.A., Álvarez, L., Sandoval, S., Quisaguano, F., Limaico, A.: On-
board visual SLAM on a UGV using a RGB-D camera. In: 10th International Conference on
Intelligent Robotics and Applications, ICIRA 2017, pp. 298–308. Springer (2017)
7. Arbanas, B., et al.: Decentralized planning and control for UAV–UGV cooperative teams.
Auton. Rob. 42(8), 1601–1618 (2018)
8. Wang, H.: Adaptive control of robot manipulators with uncertain kinematics and dynamics.
IEEE Trans. Autom. Control 62(2), 948–954 (2017)
9. Stevens, J.A., Kincaid, J.P.: The relationship between presence and performance in virtual
simulation training. Open J. Model. Simul. 3(2), 41–48 (2015)
10. Smith, M.J., et al.: Virtual reality job interview training for veterans with posttraumatic stress
disorder. J. Vocat. Rehabil. 42(3), 271–279 (2015)
11. Siu, K.C., Best, B.J., Kim, J.W., Oleynikov, D., Ritter, F.E.: Adaptive virtual reality training to
optimize military medical skills acquisition and retention. Mil. Med. 181(5), 214–220 (2016)
12. Ibari, B., Ahmed-Foitih, Z., Reda, H.E.A.: Remote control of mobile robot using the virtual
reality. Int. J. Electr. Comput. Eng. 5(5) (2015)
13. Tikanmäki, A., Bedrník, T., Raveendran, R., Röning, J.: The remote operation and environment
reconstruction of outdoor mobile robots using virtual reality. In: 2017 IEEE International
Conference on Mechatronics and Automation (ICMA), pp. 1526–1531 (2017)
14. Lwowski, J., Joordens, M., Majumdar, A., Benavidez, P., Prevost, J.J., Jamshidi, M.: The
utilization of virtual reality as a system of systems research tool. In: 2018 IEEE 13th Annual
Conference on System of Systems Engineering (SoSE), pp. 535–540 (2018)
15. Lewis, J.R.: IBM computer usability satisfaction questionnaires: psychometric evaluation and
instructions for use. Int. J. Hum. Comput. Interact. 7(1), 57–78 (1995)
Author Index
A Degani, M. H., 363

Abreu, António, 39 de O. Gonçalves, Mariana C., 425
Aguilar, Wilbert, 205 Domínguez, Miguel A., 397
Altamirano, Santiago, 231, 521 Dornellas, Rafael M., 425
Amorim, Pedro H. O., 425 dos Santos, Lúcio Agostinho Barreiros, 451
Andrade, Roberto, 507 Duque, Juan F., 397
Andrango, Raúl, 439 Dutra, Max Suell, 307
Andrea Sánchez, Z., 231, 521
Aroca, Katherine, 371, 383 E
Augusto, Maria Fernanda, 87 Edisson, Jordan-H., 439
B F
Barbosa, Jorge, 29, 65 Fernández, Rafael Caballero, 161
Barbosa, Joseane R., 425 Fernández-Villacañas Marín, Manuel A., 293
Barbosa Junior, José Ailton L., 149 Fonseca, Miguel, 321
Benavides, Eduardo, 51 Forero, Sergio Iván Rueda, 345
Brandão, Ana Paula, 257 Franklin Salazar, L., 231, 439, 521
Brito-Filho, Francisco A., 149 Fuertes, Walter, 51, 507
Buele, Jorge, 231, 439, 521
G
C Garcés, Andrés Gordón, 135
Cadena, Alyssa, 507 Gonçalves, António, 475
Canchignia, Christian, 205 Gonçalves, Rafael, 281
Carrera, Enrique V., 411 Gualoto, Franklin, 507
Carrillo, Juan M., 397 Guarda, Teresa, 87, 497
Carrillo, Sandra, 439 Guerra, Lesslie, 363
Carvalho, João Vidal, 39
Clery, Arturo, 487 H
Coelho, Maria Manuela Martins Saraiva Hallo, Vicente, 371, 383
Sarmento, 185 Huertas, Harley Lovato, 135
Correia, Anacleto, 321, 475
J
D Jácome, Richard Navas, 135
Daniel Tenezaca, B., 205 Jakomin, R., 363
de Barros Paes, Carlos Eduardo, 333 Jiménez, Arturo, 371, 383

https://doi.org/10.1007/978-981-13-9155-2
534 Author Index
K Reinoso, Cristina, 439

Kaczynska, Manuela, 281 Reis, João, 245, 281
Rey, Carina, 487
L Ribeiro, Constantino G., 307
Lopes, Isabel, 87 Roberto Segura, F., 231, 521
Lopes, Isabel Maria, 497 Rocha, Álvaro, 39, 87
López, Victoria M., 439
Loureiro, Nuno Alberto Rodrigues Santos, 451 S
Sanchez, Manuel, 51
M Sanchez, Sandra, 51
Maialle, M. Z., 363 Santiago Alvarez, T., 231, 521
María, Guillermo Santa, 487 Santos, Lúcio Agostinho Barreiros dos, 185
Marco Antonio, Zurita C., 123 Sarkar, Dilip, 15
Mendoza, Dario, 205, 371, 383 Semaan, Felipe S., 425
Menezes, Sofia, 245, 281 Silveira, João Almeida, 269
Molina, Lilian, 87, 487 Soria, Angel, 439
Mourão, R. T., 363 Souza, P. L., 363
Moura, Ricardo, 321
T
N Tapia, Freddy, 507
Narváez, Celio Humberto Puga, 161 Téllez, Osvaldo Fosado, 161
Neto, Valdemar Vicente Graciano, 333 Tello-Oquendo, Luis, 507
Tomás Núñez, C., 231, 521
O Torres, Jenny, 507
Oliveira, Pedro, 497
U
P Urrutia-Urrutia, Pilar, 231, 439, 521
Palomeque, David, 397
Paredes, Manolo, 411 V
Penello, G. M., 363 Valverde, José Mª, 397
Pereira, Pedro, 363 Velasco, Nancy, 371, 383
Pereira, Robson P., 425 Vélez, Alex Fernando Jimenez, 161
Pérez-Aloe, Raquel, 397 Viana, Nayana L. M., 149
Pérez, Bruno, 397 Victor, Avelino, 39
Pinto, L. D., 363 Victor, José Avelino, 87
Pires, M. P., 363 Victor Xavier, Enríquez C., 123
Procel, Patricia Constante, 135
W
R Wilbert Geovany, Aguilar C., 123
Ramamoorthi, Lokesh, 15
Ramírez-Cabrales, Fabián, 345 Z
Raptopoulos, Luciano Santos Constantin, 307 Zacarias, Marielba, 475

2020 Book DevelopmentsAndAdvancesInDefen PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

2020 Book DevelopmentsAndAdvancesInDefen PDF

Uploaded by

Copyright:

Available Formats

Smart Innovation, Systems and Technologies 152

More information about this series at http://www.springer.com/series/8767

Developments and Advances

ISSN 2190-3018 ISSN 2190-3026 (electronic)

Álvaro Rocha, University of Coimbra, Portugal

Lakhmi C. Jain, University of Canberra, Australia

Robson Pacheco Pereira, Instituto Militar de Engenharia—IME/RJ, Brazil

Local Organizing Chair

Luiz Eduardo Pizarro Borges, Instituto Militar de Engenharia—IME/RJ, Brazil

Abderrazak Sebaa, University of Bejaia, Algeria

Gabriele Virzi’ Mariotti, University of Palermo, Italy

Maria Manuela Martins Saraiva Sarmento Coelho, Military University Institute,

Veronica Rossano, Department of Computer Science—University of Bari Aldo

Rio de Janeiro, Brazil Álvaro Rocha

Part I Cybersecurity and Cyberdefense

Part II Computer Networks, Mobility and Pervasive Systems

Part III Defense Engineering

Part IV Health Informatics in Military Applications

Part V Leadership and e-Leadership

Part VI Planning, Economy and Logistics Applied to Defense

Part VII Simulation and Computer Vision in Military Applications

Part VIII Strategy, Geopolitics and Oceanopolitics

Part IX Safety and Maritime Protection

Part X Wearable Technology and Assistance Devices

Part XI Analysis and Signal Processing

Part XII Chemical, Biological and Nuclear Defense

Part XIII Information and Communication Technology

Álvaro Rocha holds a Habilitation (postdoctoral degree) in Information Science,

Robson Pacheco Pereira holds a degree in Chemistry (2000) and a doctorate in

Ailton Santos Filho, Ricardo J. Rodríguez and Eduardo L. Feitosa

Keywords Anti-instrumentation · Analysis-aware · Malware · Dynamic binary

The number of software specially crafted with malicious intentions (commonly

A. S. Filho (B) · E. L. Feitosa

© Springer Nature Singapore Pte Ltd. 2020 3

instance, Kaspersky (a well-known anti-virus company) stated that they analyzed

3.1 Intel Pin DBI Framework

PinVMShield [21] is a tool to detect and circumvent evasion techniques used by

Fig. 1 Pin’s software Pin same address space

Operating System (OS)

4 Evasive Techniques and Countermeasures

4.1 Evasive Techniques

An exhaustive list of techniques to detect DBI frameworks is discussed in [22, 24].

1 See https://github.com/ailton07/eXait_Plugin_PinDetectionByDEPNeglect, https://github.com/

For the development of the proposed countermeasures, we used the tool

cannot be accessible by the instrumented application. In this regard, Pin provides

4.3 Efficacy of the Countermeasures and Performance

These countermeasures were incorporated in PinVMShield to evaluate their effec-

5 Conclusions and Future Directions

Malicious software able to detect an analysis environment is a potential threat, since

Table 1 Overhead introduced by PinVMShield with our countermeasures

Lokesh Ramamoorthi and Dilip Sarkar

Abstract Organizations provide multiple web-based services to its users. To manage

Keywords Information security · Access control · Federated identity

L. Ramamoorthi (B) · D. Sarkar

© Springer Nature Singapore Pte Ltd. 2020 15

1.1 Key Components of SSO Implementation

Observing users’ tendency of accessing multiple Web sites simultaneously by open-

1 For instance, federated identity management is a widely used SSO mechanism.

Fig. 1 Browser-centric SSO

Fig. 2 A tabbed browser

To provide an enhanced personalized web browsing experience, the service providers

Table 1 Standard browser storage options

4 Process Sequence in Sign-in, Sign-out Activities