AI4CCAM interviewed Ganesh Gowrishankar, CNRS (Centre National de la Recherche Scientifique), leader of the WP2 of the project, working on “Advance AI-driving CCAM sense-plan-act predictive models”.
Ganesh Gowrishankar is a Senior Researcher (Directeur de Recherche), Interactive Digital Human group;
CNRS-UM Laboratoire d’Informatique, de Robotique et de Microelectronique de Montpellier (LIRMM).
In this interview, Ganesh tells us more about the interaction between humans, and their behavior, and automated cars.
As leader of WP2, can you tell us about the research directions explored in the project?
WP2 is the scientific WP of the project. In the WP we are specifically interested in VRU prediction, which is a major challenge for automated vehicles. The WP aims to develop a more explainable and trustworthy AI framework to predict VRU movements. We plan to do this by developing a ‘hybrid AI model’ that integrates traditional end to end data based AI models with human behavioral models developed using neuroscientific psychophysical experiments and techniques. WP2 involves DEEPS and AKKODIS who will provide AI models for VRU prediction. VIF and SKODA will help develop the scenarios to be tested and augment data using GAN to help train these models. CNRS will provide a behavioral model of VRU that will be integrated with the AI model/s. SIMULA will develop techniques and test the explainability of the developed model/s.
You are specialist of human-machine interactions and especially the role of neuro-science in this important field of research, can you summarize what the research challenges of the project in this area?
When we are driving and see a pedestrian near the road for example, we are able to get a good prediction of their next moves by just looking at his/her physical features and the environment. We will predict differently for a kid, compared to an adult for example, and predict differently if an adult is walking alone compared to in a group. To efficiently interact with humans, automated cars need to do the same.
However, this is a major challenge for automated cars (and machines and robots that interact with humans) because humans behaviors are complex. Human behaviors both with their environment and with other humans, are characterized by complex dynamics that change with an interacting individual’s physiology, age and pathology, and also depend on emotional factors like fear and anxiety. Furthermore, human behaviors are determined by their current observations as well as by predictions of behavioral models they possess of their environment and of the agents they interact with (often investigated as Theory of mind), which themselves are continuously adapted with day to day experiences.
Due to this complexity, and diversity of behaviors across humans, VRU behaviors are very difficult for AI systems to predict. We will therefore utilize behavioral experiments to get a better insight into these aspects of VRU movement prediction. Using virtual reality, we will develop experiments in which participants will be put in daily situations of interaction with cars (in the virtual reality) and evaluate how their future behaviors can be predicted from their current behaviors and environmental conditions. We will try to integrate this model with the AI model to improve overall VRU prediction behavior.
What are the main research breakthroughs that can be achieved by the project and how will they impact the future in a 5-years horizon?
Ideally, we would be able to develop a behavioral model that will be integrated with the current state of art AI models of VRU prediction to develop a hybrid model of VRU prediction. Such a model can improve VRU prediction, while being more explainable due to its neuroscientific parts.
AI4CCAM was present at Safecomp 2023, the 42nd International Conference on Computer Safety, Reliability and Security, taking place in Toulouse, France, from 19 to 22 September 2023.
The Conference was established in 1979 by the European Workshop on Industrial Computer Systems, Technical Committee 7 on Reliability, Safety and Security (EWICS TC7).
AI4CCAM was involved in the “Software testing & Reliability” session with the project coordinator Arnaud Gotlieb, Simula Research Laboratory, holding a speech on Constraint-guided Test Execution Scheduling: An Experience Report at ABB Robotics.
AI4CCAM was also present in the exhibition area.
During the Conference, a contact was established with the Japanese AI2X Co-evolution project, which advocates for a human-centered AI framework in the automated driving safety context: a potential future internationalization development for AI4CCAM.
AI4CCAM, on 6 September, took part in the workshop “Driven by Data Workshop No 4: AI in research and innovation for data-driven mobility”, represented by the project coordinator Arnaud Gotlieb, Simula Research Laboratory. The event was organised by the German Federal Ministry for Digital and Transport (BMDV).
Four research projects, including AI4CCAM, shared their experience and thoughts about the usage of ML and AI within their research and innovation projects. The workshop enabled networking and provided a forum for all participants to discuss challenges and solutions for the cities of the future. Also part of the discussion was possible borders and ethical boundaries for the use of this relatively new technology.
The AI4CCAM presentation focused on AI in Automated Vehicles: Opportunity or Threat? Also, the role and objectives of AI4CCAM were presented, leading to a Trustworthy AI for Connected, Cooperative and Automated Mobility. Finally, the AI4CCAM road ahead was explained.
For additional information on the event, click here
For AI4CCAM presentation, go to our library
On Monday, the 18th September, 2023, the AI4CCAM Ethical and Scientific Advisory Board (ESAB) held its first meeting from 12:00 to 17:15.
The ESAB is led by Prof. Jean-Gabriel Ganascia, Professor of Computer Science at Sorbonne University. The ESAB also involves Margriet Van Schijndel-de Nooij, Program director Responsible Mobility bij Technische Universiteit Eindhoven; and Jerome Perrin, Recherche scientifique et développement industriel dans les domaines énergie, environnement, mobilité – Associations caritatives – Ethique et Théologie.
The scope of the Advisory Board is to maintain an accurate body of knowledge on AI in CCAM operations, enabling the continuous consideration and analysis of ethical aspects and also facilitating interactions with relevant associations and policy-makers.
The meeting took place in the centre of Paris and triggered interesting discussions in between the Coordination Team (CT) and the ESAB members. The CT was represented by: Louise Boland and Arnaud Gotlieb, Simula Research Laboratoty; Marc Eynaud, BVA group; Loic Cantat, SystemX.
Especially, the ESAB provided recommendations of three specific questions raised by the CT:
- How can we maximize AI4CCAM impacts (Scientific, Societal, Industrial)?
- How to better liaise with the CCAM Ecosystem?
- What is a human-readable explanation for AD scene?
A follow-up progress report is expected on these and other questions that were raised by the ESAB members.
The next meeting will be held in March 2024.
SAFECOMP2023, the 42nd International Conference on Computer Safety, Reliability and Security, will take place in Toulouse, France, from 19 to 22 September 2023.
On 20 September, AI4CCAM will be presented at the “Software testing & Reliability” session by the project coordinator Arnaud Gotlieb, Simula Research Laboratory, holding a speech on Constraint-guided Test Execution Scheduling: An Experience Report at ABB Robotics.
AI4CCAM will also be in the exhibition area: meet us there and pick a project flyer to find out more!
Since it was established in 1979 by the European Workshop on Industrial Computer Systems, Technical Committee 7 on Reliability, Safety and Security (EWICS TC7), SafeComp has contributed to the progress of the state-of-the-art in dependable application of computers in safety-related and safety-critical systems. SafeComp is an annual event covering the state-of-the-art, experience and new trends in the areas of safety, security and reliability of critical computer applications. SafeComp provides ample opportunity to exchange insights and experience on emerging methods, approaches and practical solutions. It is a single-track conference without parallel sessions, allowing easy networking.
For further information on the event and for participating, click here
AI4CCAM is pleased to invite you to a new technical webinar on 15 September 2023: Detection of the Unknown Unknowns or how to make your Perception Safe.
The webinar, from 16.30 to 17.30 (CEST), will be held by Ralph Meyfarth, Deepsafety.
Deepsafety is an AI Safety company, solving the fundamental safety problem of Deep Learning AI. The output of such AI systems is undefined for data the AI has never trained before. This is holding back many industries, especially autonomous driving.
Deepsafety is working on advanced AI models for trajectory predictions and residual risk calculation. In AI4CCAM, they will develop a multi-model architecture for predicting movements of Vulnerable Road Users (VRUs) including awareness, uncertainty, and risk measures as input for the subsequent planning instance of the Autonomous Driving System.
Who is Ralph Meyfarth.
Ralph Meyfarth studied computer science at TU Darmstadt and TU Munich. In 1992, he received his doctorate from the Chair of Robotics and Automation at the TU Munich. He then began his career at Siemens, first as an engineer and later as a project manager at Siemens Management Consulting. He was responsible for many projects (e.g. benchmarking, restructuring and innovation) and reported to the Siemens Managing Board. In 1999, he founded the deep-tech startup lesswire for location-based services, which won several innovation awards. Later, he successfully developed lesswire into an automotive Tier1 for connected car services until this business was sold. In 2015, he moved to Chinese Tier1 company Hangsheng, where he was responsible for the German subsidiary and built the R&D center for car connectivity from scratch. In 2020, he joined Deepsafety as co-founder, where he is now COO.
On Wednesday 13 September, AI4CCAM will be participating at the AUGMENTED CCAM 1st Pan-European Workshop to discuss with you the challenges and opportunities of PDI (Physical, Digital and Communication Infrastructure) deployment for CCAM in Europe.
AI4CCAM will be represented by Ioanna Fergadiotou (INLECOM), discussing on “The Role of CCAM in Seamless Multimodal Transport”.
AUGMENTED CCAM is a Horizon Europe project, that aims to understand, harmonise and evaluate innovative Physical, Digital and Communication (PDC) infrastructure support solutions in order to advance its readiness for large scale deployment of Connected, Cooperative and Automated Mobilitity (CCAM) solutions for all. Within this project eleven PDI support solutions will be developed and evaluated at seven test sites in Latvia, France and Spain.
During the event, AI4CCAM flyers will be distributed as well.
For further information on the workshop and to participated, click here
The first AI4CCAM newsletter is out!
A welcome from our project coordinator, Arnaud Gotlieb, Simula Research Laboratory and an overview of the project mission; the latest news and the upcoming event; our project video to find out more about us.
The newsletter is the best way not to loose any news and update on AI4CCAM.
AI4CCAM interviewed the project coordinator, Arnaud Gotlieb, Chief Research Scientist / Research Professor at Simula Research Laboratory. Arnaud is also IRT SystemX Scientific and Technological Council Member, and Head of the Validation Intelligence for Autonomous Software Systems Dept. (VIAS).
In this interview, Arnaud explains more about the context and the innovation behind the AI4CCAM project, also focusing on the possible exploitation in the coming five years.
In which context (normative, social etc.) is AI4CCAM moving?
The AI4CCAM project is dedicated to the development of Trustworthy AI methods for connected automated vehicles. In that respect, the project has to deal with a double complexity axis: On the one hand, the exploitation of Artificial Intelligence in Europe is the subject of intensive discussions especially since the AI Act has been adopted by the European Parliament in June 2023 and the with rise of large language models such as GPT; On the other hand, the development of AI-based automated systems in personal cars explode with the recent commercialisation of the first level-3 system in Europe. Level-3 means that for a particular subsystem of the car, the machine takes the decisions instead of the driver. Therefore, in this context, it is crucial to strengthen the research in the development of responsible and robust AI for automated vehicles and that’s exactly the objective of the AI4CCAM project.
What’s the innovation behind the project?
The project innovates in different directions: First, intensive research are conducted to develop scenarios which involve ethical issues with vulnerable road users such as pedestrians or cyclists. Second, scene understanding from sensors embedded in the car is crucial to improve the transparency of automated decisions as well as gaze analysis when it comes to anticipate actions of the pedestrians. The project is conducting research on scene understanding using qualitative reasoning, which is a very hot topic and on pedestrian behaviour analysis by using virtual reality and neuroscience which shall lead to an interesting demonstrator. Third, on the side of responsible AI, the project is conducting social science studies related to user acceptance of automated vehicle in urban conditions. This shall lead to a better understanding on how automated cars can be inserted in the city and can interact with all the road users. Finally, the research results obtained in the project will be confronted in three use cases which involve realistic simulations. I can’t wait to discover how these research results will improve the trustworthiness of AI-based systems embedded into automated cars.
In 5 five years from now, how do you see the results of the AI4CCAM project being exploited?
We are building a comprehensive body of research results which will be communicated and disseminated in the CCAM ecosystem, dedicated to European citizens. In that respect, the project has a huge responsibility and ambition. As I see it, in 5 years from now, the project will have paved the way for the adoption of AI in automated vehicles in a responsible and EU-compliant way. AI is a formidable opportunity for automated driving, provided that it is developed in a way that respect humans by all means. That’s what AI4CCAM is all about…
