News

On Monday 24th March, the 3rd FAME stakeholders forum was successfully organised in Brussels. Together with other CCAM-tagged projects, AI4CCAM was invited to reflect on challenges and lesson learned from project execution.

Arnaud Gotlieb, Project Coordinator, reflected on the challenges of developing Trustworthy AI technologies and tools for automated driving (AD) application while Trustworthy AI has multiple interpretations in Europe – Its application to AD is not trivial in the sense that the perception/user acceptance of AI in AD is linked to cultural differences (languages, history, wording, etc.) and the technology is evolving so fast that its perception/acceptation is constantly moving. Involving EU citizens in consultations about AD systems to foster user acceptance, or reduce worries and misunderstanding is thus crucial. Also, techniques such as Metamorphic Testing, Generative AI, Scene understanding, Virtual Reality based testing, VRU trajectory prediction, despite their complexity are extremely relevant to create robust, transparent, trustworthy AI models.

Marc Eynaud from BVA Xsight participated to a large workshop dedicated to societal acceptance of AD, where the dissemination of Trustworthy AI to citizen was heavily debated. It was the occasion for him to examples taking into account the need for transparency and clarity, from Lisa Dixon’s Autonowashing to Citizens Convention on AI held at municipality level (example of Montpelier). More particularly, the need to discuss the adoption of AI through tools used for participative democracy will promote this type of innovation, which was AI4CCAM is trying to do via its participatory space.