4. 4
Problem
• Adoption of civilian drones will be challenging due to the complexity of physical environments
where they evolve.
• Unmanned aerial vehicles (UAV) only add value to the user if there are ways to process data
quickly and without putting additional efforts into this process.
• The faster, the more accurate, and the easier the data can be evaluated, the better.
Answer
• Combining drones and artificial intelligence seems to be the answer.
• AI in relation to Machine Learning, Deep Learning, and Motion Planning are the hottest topics.
Unlocking potential value of drones
“Drones technology can only unlock its full potential when data acquisition and
data analytics happen at the highest degree of automation.”
5. 5
Drones and AI – Motion Planning
Machine Perception (MP)
UAV must be able to perceive and absorb
the environment or objects in some way
(electro-optical, stereo-optical, and LiDAR)
Computer Vision (CV)
Automatic extraction, analysis, and
understanding of useful information from
one or more images
Machine/Deep Learning (ML/DL)
To optimize differentiable parameters,
techniques of Machine Learning can be
applied. ML algorithms are designed in
such a way that they can learn and
improve over time when exposed to new
data. DL is a subset of ML in which neural
networks learn from vast amounts of data.
Inspect & Avoid the blades of the vertical wind turbine
6. 6
Drones & AI - Drone Data Analytics
“Actionable” data. One of the most
important goals of AI in the drone industry
is to make efficient use of large data sets
which are collected by the drone.
• Availability of high-resolution images
used for various tasks such as
maintenance, surveying, mapping and
monitoring is increasing.
• There are already many established
software companies on the market
offering intelligent data analysis
solutions to make unstructured drone
data “actionable” and gain meaningful
insights without time-consuming
manual analytics.
Making unstructured drone data “actionable”
7. 7
COMP4DRONES Project
Running Time 01/10/2019 – 30/09/2022 ECSEL JU* under GA N° 826610
Start date:
Duration:
Total costs:
Total Funding (60%)
Number of participants:
Number of countries:
01 October 2019
36 months
29.76 M€
17.88 M€
49
08
The ECSEL Joint Undertaking - the Public-Private Partnership for Electronic
Components and Systems – funds Research, Development and Innovation
projects for world-class expertise in these key enabling technologies, essential
for Europe's competitive leadership in the era of the digital economy.
8. 8
Four primary safety concerns
1. The "inability to recognize" and "avoid other aircraft" and "airborne objects" in a manner similar to
manned aircraft.
2. A lack of "technological" and "operational standards" needed to guide safe and consistent
performance of UAVs
3. Vulnerabilities in the command and control. (e.g. "GPS-jamming, hacking", and the potential for
"cyber-terrorism")
4. A lack of comprehensive "government regulations" necessary to safely facilitate the accelerated
integration of UAVs into the national airspace system
Source: EASA, SESAR, GAO, UK House Of Lords
Safety issues surrounding UAVs
9. 9
Easing the
integration and
customization
of drone
embedded
system
Enabling drones
to take safe
autonomous
decisions
Ensure the
deployment of
trusted
communications
Minimizing the
design and
verification
efforts for
complex drone
applications
Ensuring
sustainable
impact and
creation of an
industry-driven
community
Project Objectives
COMP4DRONES will develop key technologies to provide secure and
autonomous drones for complex applications in the fields of transportation,
construction, logistics, surveillance and agriculture.
10. 10
Enabling drones to take safe autonomous
decisions
Design and develop a safe and reconfigurable components for the
navigation subsystem for enhancing the capability to autopilot a drone
with novel perception capacities and smart algorithms
Sensors State Prediction Conflict Detection
Conflicts
Resolution
Evasion Maneuver
Cooperative
(communicate with other
aircraft)
Non-cooperative
(sensing without
communication)
Support of Artificial Intelligence to Motion Planning
11. 11
Trusted communication
Availability
&
Performance
Confidentiality &
Integrity
Safety
Communication &
Security Monitoring
& Control
Validation
through
Attack Scenarios
Joint use of mesh
and infrastructure
technologies
Communication
safety
Ensure robust and efficient drone communications even in presence of malicious
attackers and taking into considerations intrinsic platform constraints.
• Distributed Intrusion Detection System with in-drone
Machine Learning – based detection probes
• Embedded autonomous management system for
drones, backed with artificial intelligence and machine
learning capabilities
13. 13
UC3 Logistics - METIS®
Demonstrator 1: Deployment of an Autonomous Communication
System in hard-to-access areas
• Selecting and Managing an heterogeneous fleet of
autonomous vehicles
• Using a communication infrastructure with redundant,
secure, robust, dissimilar and deterministic abilities
• Navigating and sensing at the landing or dropping
zone with a high positioning accuracy and a guarantee
of absence of objects, people or animals
• Detecting and considering dynamically of aircrafts in
the mission area and integrating vehicles of the
system in air traffic management
• Reducing risks and complexity on interactions
between system operators and autonomous vehicles
METIS® - https://www.youtube.com/watch?v=mp-9-mXny58
14. 14
UC3 Logistics
Demonstrator 2: Logistics in 5G urban environment: Clinical Sample
delivery in Hospital campus
• Develop a solution which will satisfy hospital requirements (special
challenge is safety and security requirements).
• LMT will work on establishing and benchmarking 5G communication
with special attention to security of communications.
• IMCS will work on integrating project results regarding object
recognition & avoidance and online safety monitoring in a ROS
node.
• Atechsys and SOBEN will work on coupling between a transport
drone and a logistics droide in a periurban area
15. 15
Conclusion - Expected Impact
• Strengthen artificial intelligence integration in drones and its
applications
• Reinforcing the ecosystems of drones industry by providing
methodology and a reference software architecture framework
that meets performance and safety requirements
• Improving innovation capacity and the integration of new
knowledge. A structuring aspect of COMP4DRONES is the
adoption of a "safe-by-design" approach
• Enabling and easing delivery of new services using drones in
Europe.