TACTFUL

Trustworthy AI for safe & secure traffic control in connected & autonomous vehicles

 

Scope:

Connected and Autonomous Vehicle and System Technologies will be the product of a rapidly developing Artificial Intelligence (AI)-centric breakthrough that will most likely transform the very way transport is perceived, mobility is serviced, travel eco-systems ‘behave’, and cities and societies as a whole operate. At least on paper this breakthrough promises some critical safety, mobility and sustainability rewards spanning from accident prevention, reduced traffic congestion and lessened greenhouse gas emissions to energy savings, improved surveillance, increased ease of use, and improved traffic management and control. However, these promising benefits are not without significant technological challenges. On the one hand, there is the need to ensure the autonomous driving capabilities of individual vehicles. On the other hand, the complex machine-led and interconnected dynamics of a high-tech mobility paradigm built around Connected and Autonomous Vehicles (CAVs) make our transport futures more susceptible to data exploitation and vulnerable to cyber-attacks, increasing the risks of privacy breaches and cyber security violations for road users.

There is a wealth of literature addressing trustworthy issues and cyber security threats and vulnerabilities on the technology and operational level in terms of how CAVs can be compromised and how the threat can be overcome and mitigated. However, these studies are often performed in a ‘bottom-up’ isolated manner. I.e., they focus on specific aspects of the technology that can be compromised without considering why the technology might be exploited and for what purpose within the smart traffic infrastructure. This has resulted in discrete research studies lacking a joined-up perspective of their adversarial use and how they can be mitigated. In the meanwhile, the risk of a substantial cyber-attack on smart transport infrastructure is continuously increasing.

This workshop aims at providing a venue to present approaches related to any aspect of autonomous driving and on the use of CAV/AV functionalities for traffic control, including driving algorithms, security vulnerabilities, exploit potential, and how to mitigate them by leveraging on AI to increase resilience and robustness of intelligent transport systems.


Topics:

Topics include, but are not limited to:

  • Human factors in autonomous driving
  • AI ethics in AVs and CAVs
  • The role of AI interpretability in traffic control with CAVs
  • Vulnerabilities associated with underpinning technology and connectivity
  • Threat modelling in CAVs and smart traffic control
  • Transfer learning, simulation to real-world, meta-learning, multi-task learning
  • Cyber Threat Analysis in urban traffic control and mobility
  • Cyber Threat Analysis for Avs/CAVs
  • Intersection between safety and security
  • Security by design in urban traffic control and mobility
  • AI and Traffic Control
  • AI and CAVs
  • Learning vehicle dynamics at high-speeds and in unstable regimes
  • Autonomous driving datasets, simulation, evaluations, and metrics
  • Fleet optimisation and resilience
  • AI approaches for Urban Traffic routing
  • Knowledge representation and reasoning in autonomous driving
  • Smart cities and interaction with CAVs
  • Policy developments for safety and secure CAVs and smart traffic control

Submission:

Two types of papers can be submitted. Full technical papers with a length up to 7 pages (including references) are standard research papers. Short papers with a length between 2 and 4 pages can describe either a particular application, or focus on open challenges. All papers should conform to the ECAI style template. The submission is done via EasyChair: https://easychair.org/conferences/?conf=tactful23

The reviewing process will be double blind, therefore the authors are required to anonymise the papers for submission.
Accepted papers will be made available via the workshop website.

xxx

Selected papers will be invited for inclusion in dedicated proceedings published in the LNCS series by Springer.

Final version of the papers: https://drive.google.com/file/d/1Hqb1HPyLE-0GRxRCyv-jD2q_qdgz5CW2/view?usp=sharing
 


Workshop Dates:

● Submission deadline: 30.06.2023  20.07.2023
● Decisions: 31.07.2023 10.08.2023
● Camera Ready Deadline: 14.08.2023


Schedule:

October 1

  • 9.20-10.30    session1 
    • 9.20-9.30: Opening (10 mins)
    • 9.30-9.50: Michal Welna and Joanna Jaworek-Korjakowska: Safe & secure traffic control: Deep learning based architecture for the detection of traffic anomalies based on recordings from street cameras
    • 9.50 - 10.10: Lorenz Klampfl and Franz Wotawa: Identifying Critical Scenarios in Autonomous Driving During Operation
    • 10.10-10.30: Daniel Attard and Josef Bajada: Autonomous Navigation of Tractor-Trailer Vehicles through Roundabout Intersections

 

  • 10.30-11.00    coffee break

 

  • 11.00-12.30    session 2
    • 11.00 - 11.20: Dimitrios Manolakis and Ioannis Refanidis: Finding time optimal routes for trains using basic kinematics and A
    • 11.20 - 11.40: Josef Bajada, Joseph Grech and Therese Bajada: Deep Reinforcement Learning of Autonomous Control Actions to Improve Bus-Service Regularity
    • Closing and final remarks (10 mins)

Accepted papers:

1. Michal Welna and Joanna Jaworek-Korjakowska: Safe & secure traffic control: Deep learning based architecture for the detection of traffic anomalies based on recordings from street cameras

2. Dimitrios Manolakis and Ioannis Refanidis: Finding time optimal routes for trains using basic kinematics and A*

3. Josef Bajada, Joseph Grech and Therese Bajada: Deep Reinforcement Learning of Autonomous Control Actions to Improve Bus-Service Regularity

4. Daniel Attard and Josef Bajada: Autonomous Navigation of Tractor-Trailer Vehicles through Roundabout Intersections

5. Lorenz Klampfl and Franz Wotawa: Identifying Critical Scenarios in Autonomous Driving During Operation


Invited Speakers

The workshop will feature several prominent names in the field of autonomous driving, including:

TBA


Organisers:

Paweł Skruch, Univ. Professor (skruch@agh.edu.pl - contact person)
AGH University of Science and Technology in Cracow, Poland, Department of Automatic Control and Robotics
Aptiv Technical Center in Krakow.

Marek Długosz, PhD Eng.
AGH University of Science and Technology in Cracow, Poland, Department of Automatic Control and Robotics

Moi Hoon Yap, Professor

Department International Lead and Lead of Humen-Centred Computing, Manchester Metropolitan University, UK 

Joanna Jaworek-Korjakowska, Univ. Professor
Director Center of Excellence in Artificial Intelligence, Deputy Head of the Department of automatic Control and Robotics, AGH University of Science and Technology in Cracow, Poland

Alexandros Nikitas, Reader
University of Huddersfield, Huddersfield Business School,
Huddersfield, West Yorkshire, HD1 3DH, United Kingdom

Mateusz Orłowski

Aptiv Company, Advanced Machine Learning Lead Engineer, Aptiv Technical Center in Krakow.

Simon Parkinson, Professor
University of Huddersfield, School of Computing and Engineering,
Huddersfield, West Yorkshire, HD1 3DH, United Kingdom

Mauro Vallati, Professor
University of Huddersfield, School of Computing and Engineering,
Huddersfield, West Yorkshire, HD1 3DH, United Kingdom


Contact

Paweł Skruch, Univ. Professor (skruch@agh.edu.pl)

Mauro Vallati, Professor (m.vallati@hud.ac.uk)