Download as pdf or txt
Download as pdf or txt
You are on page 1of 3

Multiple UAVs-based Surveillance and

Reconnaissance System Utilizing IoT Platform


Jong-Hong Park, Sung-Chan Choi, Il-Yeop Ahn, and Jaeho Kim
IoT Platform Research Center
Korea Electronics Technology Institute (KETI)
Gyeonggi-do, Republic of Korea
jonghong@keti.re.kr, csc@keti.re.kr, iyahn@keti.re.kr, jhkim@keti.re.kr

Abstract—In recent years, as interest in Unmanned Aerial the situation that open source based UAS needs to build a
Vehicles (UAVs) has increased, the development of UAV-based platform environment that can be easily serviced, we propose a
services in traffic monitoring, military defense, disaster global standard based IoT platform environment that has
prevention, logistics, and agricultural area has begun to be of undergone many development stages. Considering UAVs as
interest. The main characteristic of these services is that it is well as sensors and devices mounted on the UAVs as things,
effective to provide based on a plurality of UAVs rather than a users and developers can easily get open source platform
single UAV. In addition, in order to expand the service provision, service. Furthermore, we implemented an automated UAV-
different types of UAVs, not the UAVs of a particular kind, based surveillance system as an example of a service using this
should be able to operate in the same service together. In order
open source platform.
for different types of UAVs to receive the same services
efficiently, a standardized platform is needed and services need to
be provided on that platform. In this paper, to address this issue, II. UAV SERVICE WITH IOT PLATFORM
the platform based on the oneM2M global standard, which The most important requirement of the UAV platform is to
provides the Internet of Things (IoT) service layer for globally provide the common functions required by the service
interoperable IoT solutions, is considered. Based on the global
architecture. In addition, various providers should be able to
standard-based platform, we present UAV management and
control system. Furthermore, we implemented an automated
manage services and provide convenience to customers
UAV-based surveillance system as an example of a service using (developers, service users) who will use the services. In other
the platform. words, in order for the platform to be widely used, the platform
should provide developers with the functionality they need, and
Keywords—Unmanned Aerial Vehicle(UAV); Unmanned be able to provide a variety of services at a low cost.
Aircraft System (UAS); drone; oneM2M; Internet of Things(IoT); Furthermore, the platform should be able to make it easier for
platform; surveillance and reconnaissance customers to utilize the various services developed. In this
regard, the IoT platform, which is already being developed or
I. INTRODUCTION under development, is a well-prepared candidate for providing
UAV services. From the point of view of IoT, an UAV itself
Unmanned aerial vehicles (UAVs), also known as drones, can be regarded as a thing, and various sensors mounted on the
are expected to provide diverse military, civilian, commercial, UAV can also be regarded as things. Furthermore, the concept
and governmental services. Along with this expectation, the of an UAS as a service is possible [2].
development of various protocol-based communication devices
and flight control computer (FCC) providing sophisticated There are various IoT platforms currently being
navigation that can be mounted on the UAVs is creating an standardized [3]. Among them, platforms that comply with the
environment where unmanned aircraft system (UAS) operators oneM2M standard have particular advantages in providing
can receive more sensitive services. Due to this expectation and UAV services compared to existing service provider and
the development of devices, many UAS operators and application dependent platforms [4]. The advantages of
developers are demanding a system that allows different types oneM2M standard are to prevent fragmentation of the IoT
of UAVs to control and manage in the same services platform, thereby reducing development and operating costs
simultaneously. This requirement has led to the necessary to and thereby boosting the industry. Based on these advantages,
build a platform for service operation, and many service oneM2M IoT common service standard platform has been
providers have built their own platforms to provide their developed. The oneM2M common platform is a layer
service to users. DJI's Fighthub[1] is a representative UAS architecture style that can be represented as a common service
platform. Flighthub provides flight log and statistics service layer to provide network services and to provide common
through management of flight data of many UAVs and real- functionality to various IoT-based applications [5]. All the
time work environment through pilot management. However, entities in the oneM2M reference architecture are divided into
since Flighthub provides the service for UAVs developed by three tiers, and the function of each entity is defined as follows.
DJI, UAVs developed based on open source that most The application entity (AE) provides the application logic for
operators currently use cannot receive the service. Therefore, in the end-to-end IoT solution, such as cargo tracking, remote
Fig. 1. The resource browser of Mobius platform.

Fig. 3. Scenario of demonstration

IPE is a specialized oneM2M application and it serves any


legacy systems glued with oneM2M platform. Therefore, any
non-standard legacy systems can be accessed and managed
Fig. 2. The user interface of ground control station. through the oneM2M standard interfaces. Due to the functions
provided by the oneM2M standard, interworking is possible
between a plurality of heterogeneous UAVs in one service, and
monitoring, remote meter reading and control, etc. The they can serve as a mission trigger by notifying and subscribing
common service entity (CSE) is a platform made up of to each other.
functions that are commonly used by various application
entities of IoT. A network service entity (NSE) provides III. IMPLEMENTATION OF UAV SERVICE
network services to a common service entity. Based on the IoT platform mentioned above, we
The CSE incorporates an instantiation of a set of common implemented a surveillance and reconnaissance service system
services functions (CSFs). The oneM2M CSFs are defined as a using UAVs. In our implementation, two DJI MATRICE 100
total of 12 functions including registration, discovery, security, are used for patrol and a DJI MAVIC is used for tracking. DJI
group management, and subscription & notification, etc. provides software development kit (SDK) for developers, and
Among these functions, subscription & notification is a useful we used the onboard SDK for MATRICE and the mobile SDK
function for exchanging data between applications. In the CSE, for MAVIC [7]. The interworking IPE between UAVs and
many data and services are stored in the form of resources, and oneM2M-based UAV management system is running on the
applications can use such resources to read and write data. If an NVIDIA Jetson TX2 [8] on the patrol UAVs. For the oneM2M
application is interested in a particular resource and the platform, we used ‘Mobius’ which is an oneM2M-compliant
resource is automatically changed to a status change, you can IoT server platform provided by OCEAN (open alliance for
use the subscription functionality provided by the CSE. At this IoT standards) [9]. In Mobius platform, users can check the
time, the information that can be subscribed includes an access oneM2M standards-based resource mapping structure through
history of another entity to the resource to be subscribed, a the resource browser as shown in Fig. 1.
property change of the resource, and a change of the child In our case, the AE resource type is used for mapping each
resource. Similar to the search function, it is defined to set UAV to resource, and container resource type is used for
notification event criteria to notify only the changes satisfying mapping status, control, detection, and streaming data for
a certain condition. UAV to oneM2M resource. In status container, the values of
In addition, the oneM2M provides Interworking Proxy the flight data (latitude, longitude, altitude, speed, heading) and
Entity (IPE) for interwork with non-oneM2M systems [6]. The the status data (standby on ground, takeoff, in air, landing,
finishing landing) received from the corresponding UAV are IV. CONCLUSION
stored. The control container is used to store the UAV control In this paper, we present a high-level view of a UAV-based
command (takeoff, land, return to launch (RTL), waypoint) IoT platform and implement an example of multiple UAVs
received from the user. In detection container, information services utilizing the global standards based IoT platform.
about objects detected by the UAV during flight is stored. In While numerous vendors or developers are developing their
streaming container, the address of the streaming server, which own platforms, the presence of a standardized platform is an
is the destination to which the video is transmitted through the essential part of controlling and monitoring multiple
camera mounted on the UAV, is stored. heterogeneous UAVs in terms of linking with other legacy
platforms and spreading UAV services ecosystem. As a key
driving force for the fourth industrial revolution, it is expected
that the value of UAVs will be expanded not only for military
and reconnaissance purposes but also for various services such
as logistics, agriculture and entertainment. In addition, as
UAV-based services become important as future growth
industries, it is important to secure competitiveness through
convergence with IoT platform.

ACKNOWLEDGMENT
This research was supported by Unmanned Vehicles
Advanced Core Technology Research and Development
Program through the National Research Foundation of Korea
(NRF), Unmanned Vehicle Advanced Research Center
(UVARC) funded by the Ministry of Science and ICT, the
Fig. 4. Image capture of the actual demonstration Republic of Korea (2017M1B3A2A01049996) and
(2018M1B3A1A02937507).
In addition, we developed Mobius-based Ground Control
REFERENCES
Station (GCS) which can manage and control the UAVs
utilizing the system User Interface (UI) as shown in Fig. 2. The [1] DJI Flighthub – Drone Operations Management Soultion
(https://www.dji.com/flighthub)
GCS subscribes to the status resources of all registered AEs
and shows the location and direction of the UAVs on the map. [2] S-C. Choi, N.-M. Sung, J.-H. Park, I.-Y. Ahn, and J. Kim, “Enabling
Drone as a Service: oneM2M-based UAV/Drone Management System,”
Also, when click on a specific UAV, a user can give a takeoff, in proc. 2017 Ninth International Conference on Ubiquitous and Future
land, or RTL command by using the button, and can give Networks (ICUFN), Jul. 2017.
waypoint mission by dragging to certain position. When the [3] J. Guth, U. Breitenbücher, M. Falkenthal, F. Leymann, and L. Reinfur,
Play button is pressed, the GCS receives the address of the “Comparison of IoT Platform Architechtures: A Field Study Based on a
streaming server from the resource and displays the live Reference Architecture,” in proc. 2016 Cloudification of the Internet of
streaming screen. The left side of the screen shows information Things (CIoT), Nov. 2016.
about objects detected by the camera mounted on the UAVs. [4] A. Kunz, A. Prasad, K. Samdanis, S. Husain, and J. Song, “Enhanced
3GPP system for Machine Type Communications and Internet of
All of the functions provided by GCS are available because the Things,” in proc. 2015 IEEE Conference on Standards for
GCS subscribes to resources for all AEs registered on the Communications and Networking (CSCN), May 2015.
implanted system. [5] J. Swetina, G. Lu, P. Jacobs, F. Ennesser, and J. Song, “Toward a
Standardized Common M2M Service Layer Platform: Intoduction to
In our implementation, the demonstration scenario is shown oneM2M,” IEEE Wireless Communications, vol. 21, no. 3, pp. 20-26,
in Fig. 3. First, two patrol UAVs subscribe to the each other's Jun. 2014.
detection resources, and a tracking UAV subscribes to [6] J. Yun, R. C. Teja, N. Chen, N.-M. Sung, and J. Kim, “Interworking of
detection resources of both Patrol UAVs. The two patrol UAVs oneM2M-based IoT Systems and Legacy Systems for Consumer
move according to the waypoint mission as the user Products,” in proc. 2016 International Conference on Information and
commanded through the GCS and continue to perform the Communication Technology Convergence (ICTC), Oct. 2016.
object recognition algorithm. When one of the patrol drones [7] DJI - DJI Matrice 100: The Quadcopter fir Developers
(https://www.dji.com/matrice100?site=brandsite&from=nav)
recognizes a person, the new data is updated in the container
[8] NVIDIA-Embedded Systems Developers Kit & Modules
instance of the detection resource. At this time, since both of (https://www.nvidia.com/en-us/autonomous-machines/embedded-
the other UAVs are subscribing to detection resource, they will systems/)
be notified of new data update. This will cause the remaining [9] OCEAN Developers (http://developers.iotocean.org/)
patrol UAV to perform landing commands, and the tracking
UAV will takeoff and move to the position of detecting patrol
UAV automatically. Then, the tracking UAV will perform
active tracking mission following the recognized person. Fig. 4
captures the actual demo video for each step.

You might also like