Available online at www.sciencedirect.

com
Available online at www.sciencedirect.com

Available ScienceDirect
ScienceDirect
online atonline
Available www.sciencedirect.com
at www.sciencedirect.com

ScienceDirect
Procedia CIRP 00 (2019) 000–000
Procedia CIRP 00 (2019) 000–000
ScienceDirect www.elsevier.com/locate/procedia
www.elsevier.com/locate/procedia
Procedia CIRP 00 (2017)
Procedia 000–000
CIRP 88 (2020) 163–166
www.elsevier.com/locate/procedia

13th
13th CIRP
CIRP Conference
Conference on
on Intelligent
Intelligent Computation
Computation in
in Manufacturing
Manufacturing Engineering,
Engineering, CIRP
CIRP ICME
ICME '19
'19

A
A holistic
holistic methodology for
28th CIRP Design
methodology development
forConference,
development of
of Real-Time
May 2018, Nantes, France Digital
Real-Time Digital Twins
Twins
A new methodology
Carmen to analyze
Carmen Constantinescu
Constantinescu
a,
theGiosan
a,*, Stefan
*, Stefan functional
Giosan a,b, Raul and
a,b
Mateiphysical
, Raul Matei architecture
a,b, Denis Wohlfeld
a,b
, Denis Wohlfeld
c
c of
existing products for an assembly
Fraunhofer
Fraunhofer
a
oriented
IAO, Nobelstraße
IAO, Nobelstraße
a product
12, Stuttgart 70569,
12, Stuttgart 70569,
Germany family identification
Germany
Technical University of Cluj-Napoca, Strada Memorandumului 28, Cluj-Napoca 400114, Romania
b
Technical University of Cluj-Napoca, Strada Memorandumului 28, Cluj-Napoca 400114, Romania
b
FARO Europe GmbH & Co. KG, Lingwiesenstraße 11/2, D-70825, Korntal-Münchingen, Germany
c
FARO Europe GmbH & Co. KG, Lingwiesenstraße 11/2, D-70825, Korntal-Münchingen, Germany
c
Paul Stief *, Jean-Yves Dantan, Alain Etienne, Ali Siadat
* Corresponding author. Tel.: +49 711 970-2282; E-mail address: carmen.constantinescu@iao.fraunhofer.com
* Corresponding author. Tel.: +49 711 970-2282; E-mail address: carmen.constantinescu@iao.fraunhofer.com
École Nationale Supérieure d’Arts et Métiers, Arts et Métiers ParisTech, LCFC EA 4495, 4 Rue Augustin Fresnel, Metz 57078, France

*Abstract
Corresponding author. Tel.: +33 3 87 37 54 30; E-mail address: paul.stief@ensam.eu
Abstract
The Digital Twins of product and production face the challenge that once the development process is closed, they does not reflect the real status
The Digital Twins of product and production face the challenge that once the development process is closed, they does not reflect the real status
of production where events as equipment failures, poor quality or missing compound parts happens continuously. For assuring production
of production where events as equipment failures, poor quality or missing compound parts happens continuously. For assuring production
Abstract
resilience a holistic methodology, combining a top-down with a bottom-up 3D scanning approach for capturing real-time production parameters
resilience a holistic methodology, combining a top-down with a bottom-up 3D scanning approach for capturing real-time production parameters
and embedding them in Digital Twins is developed. The paper presents the methodology and a motivation scenario for further validation in an
Inand embedding
today’s
innovative
them
business
set-up of
in Digital Twins
environment,
an automated
is developed.
themeasurement
trend towardscell,The product
more paper presents
varietythe
where state-of-the-atandmethodology
customization
technologies
and aunbroken.
motivation
as isautonomous
scenario
Due
AGV, thisfor
to mobile further validation
development, in of
the need
3D laser scanning
an
and
innovative
agile and set-up of an production
reconfigurable automated measurement
systems emergedcell,towhere
cope state-of-the-at
with various technologies
products and as autonomous
product AGV,
families. To mobile
design and3Doptimize
laser scanning and
production
automated processes are integrated.
automated well
systems processes
as to are integrated.
© 2019 as The Authors. choose
Publishedtheby
optimal
Elsevierproduct
B.V. matches, product analysis methods are needed. Indeed, most of the known methods aim to
© 2020
2019
analyze aThe Authors.
Authors.
product or Published
one product by Elsevier
family on B.V.
theB.V.
physical level. of
Different
Peer-review
This under
is an open responsibility
access of the scientific
article under CC BY-NC-ND committeelicense the 13thproduct families, however,
CIRP Conference may differ
on Intelligent largely ininterms
Computation
(http://creativecommons.org/licenses/by-nc-nd/4.0/)
of the number and
Manufacturing
Peer-review
nature under responsibility of the scientific
of components. This fact impedes an efficient committee of the
comparison 13th CIRP Conference
and choice on Intelligent
of appropriate Computation
product family in Manufacturing
combinations for the production
Engineering.
Peer review under the responsibility of the scientific committee of the 13th CIRP Conference on Intelligent Computation in Manufacturing
Engineering.
system. A new17-19
methodology is proposed to analyze existing products in view of their functional and physical architecture. The aim is to cluster
Engineering, July 2019, Gulf of Naples, Italy.
these products
Keywords: in new
Digital assembly
Twins; oriented
Real-time product families for the optimization of existing assembly lines and the creation of future reconfigurable
3D scanning
Keywords: Digital Twins; Real-time 3D scanning
assembly systems. Based on Datum Flow Chain, the physical structure of the products is analyzed. Functional subassemblies are identified, and
a functional analysis is performed. Moreover, a hybrid functional and physical architecture graph (HyFPAG) is the output which depicts the
similarity between product families by providing design support to both, production system planners and product designers. An illustrative
example
1. of a nail-clipper Twins
is used to explain the proposed methodology. An such industrialmaterials,
case study parts,
on twomachines,
product families of steering columns of
1. Real-Time
Real-Time Digital
Digital Twins approach
approach such asas materials, parts, machines, equipment,
equipment, tools,
tools, people,
people,
thyssenkrupp Presta France is then carried out to give a first industrial evaluation etc. of
fromthe proposed
the approach.
Factory’s Shop Floor. Simple factory objects
© 2017The guiding idea of the research team is to realise the Real- etc. from the Factory’s Shop Floor. Simple factory objects
TheThe Authors.
guiding ideaPublished by Elsevier
of the research team B.V.
is to realise the Real- endowed with smart sensors will be treated as smart factory
Time
Peer-review Digital Twins
under Twins concepts
responsibility [1,2],
of the scientific technologies and endowed with smart sensors will be treated as smart factory
Time Digital concepts [1,2], committee
technologies of theand
28th CIRP Design
objects [4]Conference 2018.
and the information thus collected will be referred
applications objects [4] and the information thus collected will be referred
applications and and their
their validation
validation in in operational
operational factory
factory to
to as the context
as the context [5] [5] of
of aa smart
smart factory
factory object.
object. The
The context
context
environment, towards the achievement of the
environment, towards the achievement of the factory
Keywords: Assembly; Design method; Family identification
factory consists
optimum, consists of physical information (location, movement, energy
of physical information (location, movement, energy
optimum, suchsuch called
called factory
factory resilience
resilience planned
planned to to produce
produce at at consumption,
optimum consumption, temperature,
temperature, pressure,
pressure, etc.),
etc.), logical
logical information
information
optimum [3],[3], and
and toto promote
promote it it as
as standard-based
standard-based method.method. (identity,
The (identity, privileges,
privileges, and and preference),
preference), interrelationships
interrelationships withwith
The proposed
proposed approach
approach (Figure
(Figure 1) 1) consists
consists of of aa Factory
Factory other
1.shop
Introduction other domains, and a history of the object
of the domains,
product and
range a
andhistory of the
characteristics along
along the
manufactured
object time
theand/or
time
shop floor populated with scalable networks of technical and
floor populated with scalable networks of technical and scale.
assembled in this system. In this context, the main challenge in
human sensors—an approach based on the anthropocentric scale.
human sensors—an
Due to the approach
fast sensors” based
development on the anthropocentric
in- providing
the domain of For
modelling example,
and the
analysis following
is now notcontext
only parameters
to cope with can be
single
“humans For example, the following context parameters can be
“humans as as universal
universal sensors” concept
concept -ofproviding context-
context- In
communication
data and an ongoing trend digitization and considered:
products,
considered: ID,
ID, Type
a limited Type (e.g.,
product Chiron),
(e.g.,range Status
Status (e.g.,
or existing
Chiron), In Failure),
product
(e.g., families,
Failure),
data onon primary
digitalization, primary factory
manufacturing
objects
objects (humans,
factory enterprises (humans,
area Space
facing of
machines,
machines,
important Date
but Time (e.g., 16.02.2016 13:25 CET), Position (i.e.,
equipment, tools, devices, parts, etc.) and Factory Date Time (e.g., 16.02.2016 13:25 CET), Positiondefine
also to be able to analyze and to compare products to (i.e.,
equipment,
challenges tools,
in devices,
today’s parts,
market etc.) and
environments: a Space of Factory
a continuing x=200,
new y=150,
product z=725),
families. It Relation(s)
can be observedto other
that factory
classical objects
existing
Real-Time Models which are shared and queried by factory x=200, y=150, z=725), Relation(s) to other factory objects
Real-Time Models which are shared and queried by factory
tendency towardsState-of-the-art
applications. reduction of product sensor development
systems, times and
wireless (e.g., Product:
(e.g.,
product families are
Product: 459XY,
regrouped
459XY, Production
in function
Production order: TTT56,
of clients
order: Order
or features.
TTT56, Order
applications.
shortened product State-of-the-art
lifecycles. In sensor
addition, systems,
there is an wireless
increasing priority:
However, Very high, Drill Tool T45V, Rotation speed: xyz
communication, and priority: Very high, Drill Tool T45V, Rotation speed:find.
assembly oriented product families are hardly to xyz
communication,
demand and mobile
of customization, mobile
being
device
device
at
technologies -- herewith
thetechnologies
same time in herewith
a global rot/min),
On the etc.
product family level, products differ mainly in two
called Cyber-Physical System Technologies (CPSTs) will rot/min), etc.
called Cyber-Physical
competition with System
competitors all Technologies
over the world. (CPSTs)
This will
trend, main characteristics: (i) the number of components and (ii) the
allow
allow the
the capturing
capturing of
of the
the current
current status
status of
of factory
factory objects
objects
which is inducing the development from macro to micro type of components (e.g. mechanical, electrical, electronical).
markets, results in diminished lot sizes due to augmenting Classical methodologies considering mainly single products
product varieties (high-volume to low-volume production) [1]. or solitary, already existing product families analyze the
To cope with this augmenting variety as well as to be able to product structure on a physical level (components level) which
identify possible optimization potentials in the existing causes difficulties regarding an efficient definition and
2212-8271 ©system,
production 2019 The Authors. Published toby Elsevier B.V. knowledge
2212-8271 © 2019 TheitAuthors.
is important
Published byhave a precise
Elsevier B.V. comparison of different product families. Addressing this
Peer-review©under
2212-8271 2020responsibility
The Authors. of Published
the scientific
bycommittee of the 13th CIRP Conference on Intelligent Computation in Manufacturing Engineering.
Elsevier B.V.
Peer-review under responsibility of the scientific committee of the 13th CIRP Conference on Intelligent Computation in Manufacturing Engineering.
This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Peer review
2212-8271 under
© 2017 the
The responsibility
Authors. Publishedofbythe scientific
Elsevier B.V.committee of the 13th CIRP Conference on Intelligent Computation in Manufacturing
Engineering,
Peer-review 17-19
under July 2019,ofGulf
responsibility of Naples,
the scientific Italy. of the 28th CIRP Design Conference 2018.
committee
10.1016/j.procir.2020.05.029
164 Carmen Constantinescu et al. / Procedia CIRP 88 (2020) 163–166
C. Constantinescu. et al./ Procedia CIRP 00 (2019) 000–000
Fig. 1. Concept of Space of Real-Time Digital Twins.
This way the factory will become a network within which
every device can be linked to the end of providing its
information to any user interested in these data. A very large
amount of context data will be captured by a wide range of
mature, commercially available, well understood and accepted
CPSTs. These can be combinations of tags (RFID, barcodes,
etc.), portable computing devices (e.g., tablet, PC), and smart
sensors (mobile 3D laser scanners) deployed in the factory
shop floor.
The models of smart factory objects such called context
and user-aware models will be available in a repository
named Space of Factory Real-Time Models. The Product,
Process, equipment, Building and Factory Data Models
will thus be extended by dynamic context data about the smart
factory objects, such as the objects’ location, movement,
status, etc. The models will be enriched by temporal aspects
respectively by attaching historical information to the context
data. These models are further on called Real-Time Digital
Twins.
The Space of Real-Time Models will build the foundation
of Factory Context-aware Applications [6], which will
provide decision support by constantly monitoring the status
of the factory smart objects in real-time. Thereby, the
applications will be able to act upon and adapt to changes in
the shop floor, and select and present information depending
on the applications physical context. They can share the wide
range of context models from the Space according to their
aim, e.g. real-time monitoring of energy use, material flow,
quality, machine conditioning, etc. Based on the enrichment
of the context models by temporal aspects, these applications
can query not only the current state of the factory objects, but
also states of the past or even predicted ones in the future, e.g.
through online simulation.
2. Holistic methodology for development of Real-Time
Digital Twins
2.1. Research procedure and motivation scenario
In order to implement the Real-Time Digital Twins in
operational manufacturing environment, the research group
approaches a bottom-up procedure, addressing the theme in
four steps. In the fist step, identification of at least two
motivation scenarios coping with the challenge of giving life
and real-time to the Digital Models of physical manufacturing
entities from factory shop floor, e.g. parts, components,
equipment, tools, devices, human workers, etc.; is done. The
validation of the motivation scenarios (Step 2) with state-of-
the-art enabling technologies is followed by (Step 3),
development of a generic approach for Real-Time Digital
Twins in manufacturing. the (Step 4), development of the
Road Map for migration of the generic approach in other
industries and use cases, e.g. machine tool/equipment
industry. Additionally, the employment of other context
capturing technologies, e.g. smart wireless sensors, will be
considered as well.
The motivation scenario presented in this paper consists of
the Measurement Cell, dominated by robotics and
autonomous vehicles (AGVs) and surveilled by 3D capturing
and sensing technologies. The overall objectives of building
and validating this scenario is the improvement of the quality
of the body in white components towards zero defect
manufacturing in automotive industry.
The created layout of the Measurement Cell, presented in
Figure 2, consists of the following main components,
highlighted in Figure by their numbers, as follow, (1)
Industrial robotic arm (KUKA, KR360 R2830);
 Carmen Constantinescu et al. / Procedia CIRP 88 (2020) 163–166 165
C. Constantinescu. et al./ Procedia CIRP 00 (2019) 000–000

2.2. Enabling technologies for Measurement Cell set-up

For the realization of the Measurement Cell, the research

group employed a number of state-of-the-art technologies in
the field of mobile/autonomous industrial robotics,
projection systems, 3D laser capturing, which are
comprehensively presented in the following paragraphs.
Industrial robots [3] are automated equipment, capable
of movement on three or more axis and they can be
programmed in various ways. In laser scanning quality
check [7], an industrial robot can follow a defined trajectory
with respect to the part that needs to be measured, multiple
times, with the same precision.
Automated Guided Vehicles: known as AGVs, are used
Fig. 2. Layout of the Measurement Cell.
in industry for logistic operations. There are multiple types
of AGV’s considering their uses and navigation capabilities.
(2) Laser scanner for measurement of body shells/attachments
Processes like towing, loading and transporting are performed
coupled on industrial robotic arm (Cobalt, KUKA KR360);
autonomous by following a cable, a tape, a laser or even a
(3) AGV for parts transport (KUKA, KMP1500 FTS); (4)
wireless interfaces configuration. Parameters like maximum
Mobile industrial collaborative robot (KUKA KMR IIWA);
load, battery autonomy, and maximum velocity, recharging
(5) Part, body in white (Mercedes S Class); (6) Safe camera
time, safety measures and position accuracy must be
system for monitoring safe rooms (PILZ SafetyEYE); (7)
considered when modelling an AGV.
Laser scanner for measurement and projection on surfaces
3D laser scanning: a relatively new technology, used to
(FARO VECTOR); (8) Laser scanning for measurement of
increase productivity and collect 3D data to optimize and
painted bodies (FARO FOCUS); (9) Imaging Laser Projector
simplify the work flow [8]. In the last years, there has been
(FARO TRACER); (10) Mobile 3D laser scanner (FARO
major progress with its implementation in industry, mainly for
ScanBot); (11) End of line measurement gate for buckling
quality check processes, but also for reverse engineering
dents (FARO Quality-Gate).
products and facility management. To have relevant data, the
The process flow of the performed activities in the
scanning parameters must meet the modelling demands.
Measurement Cell consists of the following steps, where the
Therefor the scanning rate, speed, resolution and distance
involved components from the Layout presented in Figure 2
must be considered in 3D industrial laser scanning.
are marked with numbers:
Safety camera systems: when working with automated
Step 1: The KUKA AGV (3) transports the Mercedes body
processes, defining a safety area is mandatory. Even if all the
in white (5) from outside of the layout area into the cell;
machines were set up correctly and installed properly, there is
Step 2: The KUKA industrial robotic arm (1) measures the
always a slim chance for an unfortunate accident, due to
body quality on one side of the body through the Cobalt laser
human error or incorrect software development. To solve this
scanner (2);
problem, the perimeter around the working station/cell, can be
Step 3: The KUKA mobile collaborative robot (4)
surrounded with fences. A more innovative way would be
measures all other sides of the body (5) where the fix KUKA
using a sensory based system. This way, the restricted area
industrial robotic arm (1) cannot reach the body;
can be defined in real time, offering flexibility and a higher
Step 4: The PILZ SafetyEYE (6) monitors the cell for
safety level.
safety issues, while the cell is open on the left and bottom
sides as in yellow and red colors is in Figure 2 is highlighted.
2.3. Holistic methodology and solution
The yellow area will trigger an alert and the red one will stop
for safety reasons all processes into the cell;
The holistic methodology and our solution for the
Step 5: The FARO VECTOR (7), FARO FOCUS (8), and
realization of the Real-Time Digital Twins is presented in
FARO TRACER (9) capture continuously the quality of the
Figure 3. The Figure represents on the bottom the factory
processes performed in the cell, on a fixed and top-down
shop floor where equipment, tools, devices, parts, etc. are
frame;
monitored in real-time by capturing their context through
Step 6: The FARO ScanBot (10) scans the whole cell in a
technologies like smart sensor “factory objects” is represented
mobile and bottom-up frame;
by: a) Type T, e.g. KUKA robot; b) Position x, y, z in factory
Step 7: After the measurement operation performed on the
layout; c) Status s, e.g. 0=not working, 1=working; d) Time
Step 2 and Step 3 are accomplished, the KUKA AGV (3)
t; e) Relations r, e.g. the KUKA robot processes measurement
transports the Mercedes body in white part (5) outside of the
of the Mercedes body in white with a high priority order and a
cell by passing throw FARO Quality Gate (11) where the final
accuracy up to 100µm. On the middle layer of the Figure the
high-precision quality measurement is performed.
context aware Models of the factory objects are realized by
The activities of the Steps 2-6 are performed
using technologies and systems from the upper layer, e.g.
simultaneously being followed by the final Step 7.
CAx for modeling the products, 3D laser systems for
capturing the real time context of the factory shop floor.
166 Carmen Constantinescu et al. / Procedia CIRP 88 (2020) 163–166
C. Constantinescu. et al./ Procedia CIRP 00 (2019) 000–000
Fig. 3. Holistic methodology and solution.
On the right in Figure 3, the Context-aware assistance
systems for Real-Time Digital Twins is represented as the
solution envisioned by the research group. It consists of
few components represented in Figure 3 in boxes and
detailed below as follows:
Step A: In the first step of the process flow, all factory
objects, e.g. industrial robotics arm, mobile AGV, 3D laser
scanning, parts, etc. are modelled by using digitalization
systems like CAD for the product design, Digital Factory
for layout creation, etc.;
Step B consists of data acquisition through e.g. wireless
sensors and/or laser scanning, using a bottom-up approach.
This process is also known as context capturing;
Step C: Simulation and analysis process represents the
next step performed after enough data in real-time is
captured. Using 3D captured information and having the
production history data, an optimization algorithm based on
machine learning technology can be applied to develop
further optimization solutions or measures aiming at
continuously improvement of key production parameters,
like in our case the quality of the product;
Step D: Elaboration of measures for optimization of the
performed processes represents the final step, accomplished
by documentation and visualization. The last can be
performed through a report or even by a laser projector
right in the factory floor. Additionally, some mobile devices
as iPad, iPhone can be used for sending triggers, alerts to
the worker operating into the production area, in our case
outside of the cell. Based on these measures the semi-
automatic or self-configured optimization of the cell is
realized.
3. Conclusions and future work
The performed work started with a new concept on
giving Real-Time feature to the Factory Digital Twins,
followed by definition and set-up of a motivation scenario.
Next activities will focus on defining the second motivation
scenario in the field of Digital Twins for carbon fibre 3D
printing process. The validation of both motivation
scenario, the development of a generic approach for Real-
Time Digital Twins in manufacturing and the development
of the Road Map for migration in other industries and use
cases are planned as well.
References
[1] Boschert S, Heinrich C, Rosen R. Next generation digital twin.
Proceedings of TMCE 2018, 7-11 May, 2018, Las Palmas de Gran
Canaria, Spain Edited by: I. Horváth, J.P. Suárez Rivero and P.M.
Hernández Castellano, ISBN 978-94-6186-910-4.
[2] Madni AM, Madni CC, Lucero SD. Leveraging Digital Twin
Technology in Model-Based Systems Engineering. Systems Journal,
Systems 2019, 7(1), 7; https://doi.org/10.3390/systems7010007.
[3] Morar L, Westkämper E, Abrudan I, Pisla A, et al. Planning and
operation of Production Systems. Ed. Fraunhofer IRB, 2007. p. 387-
409, p. 463-507.
[4] Datta SPA. Emergence of Digital Twins - Is this the march of reason?.
Journal of Innovation Management Datta, JIM 5, 3 (2017) 14-33. ISSN
2183-0606. P. 15.
[5] Lucke D, Constantinescu C, Westkämper E. Context Data Model, the
Backbone of a Smart Factory. In: Tichkiewitch, Serge (Chairman) u.a.;
CIRP: "Sustainable Development of Manufacturing Systems": 42nd
CIRP Conference on Manufacturing Systems, Grenoble, June 3-5,
2009.
[6] Lucke D, Constantinescu C, Westkämper E. “Smart Factory - A Step
towards the Next Generation of Manufacturing". In: "Manufacturing
Systems and Technologies for the New Frontier", Editors: Mitsuishi,
Ueda, Kumura, the 41st CIRP Conference on Manufacturing Systems,
2008, 26-28 May, Tokio, Japan, ISBN 978-1-84800-266, Springer-
Verlag, London, S. 115-118.
[7] Schubert S, Neubert P, Protzel P. How to build and customize a high-
resolution 3D laserscanner using o-the-shelf components. Towards
Autonomous Robotic Systems (TAROS) 2016. Springer via
http://dx.doi.org/10.1007/978-3-319-40379-3 33.
[8] Wulf O, Wagner B. Fast 3d scanning methods for laser measurements
systemshttps://pdfs.semanticscholar.org/633d/42113180b581e64ac04b
846cb55c6b9bffd2.pdf.