2nd Hi-Drive SUMMER SCHOOL: Navigating the Future with Advanced and Safe Automated Driving 

The 2nd Hi-Drive Summer School on ‘Navigating the Future with Advanced and Safe Automated Driving’, held on September 25–26 in Vouliagmeni, Greece, brought together 60 participants from across Europe to discuss the latest developments in automated driving (AD) technologies. Proudly organised by the EU project Hi-Drive and the Institute of Communication and Computer Systems (ICCS) of the National and Technical University of Athens, the event featured four sessions with 11 insightful presentations and seven research posters. Volkswagen Group’s Aria Etemad, coordinator of the Hi-Drive project, and Dr. Angelos Amditis, Research and Development Director at ICCS, inaugurated the two-day event, which centered on the most cutting-edge advancements in automated driving.

Key Insights from Hi-Drive Summer School 2024

The Hi-Drive Summer School 2024 offered profound insights into the complexities of AD technologies, with experts highlighting both technical advancements and human-centric considerations. Among the key takeaways was the recognition that driving behaviours and automation level definitions vary significantly across cultures, with Japan offering a distinct perspective. ASAM OSI was identified as a promising first step toward the standardisation of interoperable sensor models, while user prompts in multimodal foundation models demonstrated potential for improving predictions of other road users’ movements—a critical challenge for AD.

Cybersecurity and functional safety emerged as closely intertwined areas requiring parallel attention, while the robustness of deep learning models for situational awareness continues to demand further research. Experts also emphasised that testing prediction and planning modules for mixed-traffic scenarios necessitate sophisticated new simulation tools and safety metrics tailored to diverse road users.

Other advancements discussed included gated cameras, which are showing significant promise in poor weather conditions, and V2X-enabled functions, though the latter remains contingent on overcoming networking challenges. Driver gaze tracking was proposed as a powerful tool for better-predicting transitions of control in autonomous systems, and standardized HMI elements were suggested as a way to improve user understanding of AD functionalities. Additionally, participants agreed on the urgent need for structured, interdisciplinary research to support the teleoperation of autonomous vehicles. Moving forward, the focus will be on refining models, enhancing user acceptance through simulation-based training, and achieving human-like AV behaviour through more advanced interaction models. These efforts hold promise for the responsible deployment of AD/ADAS systems in the near future.

As the Hi-Drive Summer School 2024 ended, it was clear that the event offered invaluable insights into the future of automated driving, seamlessly blending technical progress with human-centred approaches.

Below you can find all the presentations from this year’s Summer School, and for a more comprehensive view, you can also check out the 2023 Hi-Drive Summer School presentations here: https://www.hi-drive.eu/events/summerschool_1/.

Agenda

>>>  The detailed agenda with all featured speakers is available HERE

DAY 1 | 25 September
Session A: Advancements in Perception and Motion Prediction Planning and Testing

• Fascinating sensor behavior models: enabling the disruption from ADAS to AD || Stefan Schneider, Professor for Advanced Driver Assistance Systems, University of Applied Sciences Kempten
• The key to proactive traffic interaction: Challenging sequential integration of prediction and planning – a survey perspective || Steffen Hagedorn – PhD Student AI for Automated Driving at Bosch
• Enhancing vehicle vision for automated driving in low visibility conditions with Artificial Intelligence: insights and solutions from the European research project AI-SEE || Werner Ritter, Researcher and Senior Expert, Mercedes-Benz GA 

Session B: Advancements in Collective Perception Design and Testing

• Microclouds for Cooperative Perception in Vehicle Platoons || Claudio Ettore Casetti, Full Professor, Department of Control and Computer Engineering – Politecnico di Torino,  
• Cybersecurity Design for Collective Perception || Krishna Ashok, Senior Engineer at Volvo Cars Corporation

DAY 2 | 26 September

Keynote: 7 Foundational Principles for Automated Driving || Alexandre Massoud Alahi, Associate Professor & Director of the Visual Intelligence for Transportation (VITA) Laboratory at Ecole Polytechnique Fédérale de Lausanne (EPFL)

Session C: Human-in-the-loop and human factors 

• Motion Comfort in AVs: Understanding, Modelling and Enhancing it || Georgios Papaioannou, Assistant Professor on Motion Comfort in AVs, Delft University of Technology 
• Estimating driver takeover performance using gaze data || Rafael Goncalves, Research Fellow, University of Leeds 
• Teleoperation: research needs and human factors challenges || Lena Plum, Psychologist, Federal Highway Research Institute – BAST 

Session D: Human Factors Research for Autonomous Driving Safety Argumentation

• Human factors research for standards and regulation || Ilse Harms, Senior Advisor Human Factors & Vehicle Automation, RDW
• Human road user behaviour modelling to support vehicle automation || Gustav Markkula, Professor & Chair in Applied Behaviour Modelling, University of Leeds

Poster Sessions

• Stochastic Model Predictive Control of an Autonomous Vehicle Approaching Unsignalized Crosswalks with Pedestrians ||  Branimir Škugor, Assistant Professor, Department of Robotics and Automation of Manufacturing Systems, University of Zagreb
• On-Road Autonomous Vehicle Navigation In a Dynamic Environment Using Deep Reinforcement Learning, Towards Fuel Consumption Optimisation || Elpida Kelesi, PhD. candidate, Intelligent Systems and Software Engineering Labgroup – Aristotle University of Thessaloniki
• Occupant’s Postural Stabilisation While Being Driven In Automated Vehicles || Chrysovalanto Messiou, PhD. candidate, Technical University of Delft
• Predictive Models Based On Machine Learning Techniques In Connected Systems To Improve Traffic Flow And Road Safety || Gerard Franco-Panades, PhD. candidate, Servei Català de Trànsit (GENCAT-SCT), CARNET, Technical University of Catalonia (UPC) 
• Robust Multimodal and Multi-Object Tracking for Autonomous Driving Applications || Nil Muntè, ADAS & AV Software Engineer, IDIADA 
• Modelling Interactions Between Cyclists And Motorists at Unsignalised Intersections || Ali Mohammadi, PhD candidate, Chalmers University of Technology
• Pedestrian Intention Prediction Using Computer Vision and Machine Learning || Mohsen Azarmi, Doctoral Candidate Institute for Transport Studies, University of Leeds

#HiDrive_Summerschool2024