Since July 1, 2022, Inria has been coordinating and supporting Adra-e, a project funded by the European Commission, with the aim of facilitating synergies between the AI, data and robotics communities. Marc Shoenauer, head of the TAU project team at Saclay and deputy scientific director in charge of AI, explains the project's scope.
What is the origin and meaning of the project?
Adra (AI, data and robotics association) is the private partner of the European Partnership ADR (AI, Data and Robotics) which aims to strengthen collaboration and develop synergies between the three complementary technological communities of AI, data and robotics.
Adra-e is a European project that has been set up in close collaboration with the Adra association. It is the operational tool for the development of a European ADR by building bridges between communities while raising awareness and protecting the interests of EU citizens. Adra-e's objective is to participate in the construction of a trustworthy, sustainable and competitive ADR ecosystem that is able to raise private investments.
The project has six objectives:
- Connect the three communities and support the Strategic Agenda for Research Innovation and Deployment (SIRD) through its update and implementation.
- Map the current ADR landscape and its infrastructure to build connections between the communities.
- Raise awareness of trusted ADR technologies for all with the creation of an ADR Awareness Centre in particular.
- Increase capacity for innovation and adoption of ADR technologies.
- Supporting the deployment of standards and regulations in support of European sovereignty.
- Sustainability through close collaboration with Adra, the AI-on-Demand platform and the ADR partnership.
Who are the actors of the Adra-e project?
Inria is the coordinator of a consortium of leading industrialists and academics pooling their expertise in the three disciplines of AI, data and robotics. The constitution of this consortium has enabled us to achieve a balanced representation between the industrial and academic sectors, as well as between the three deeptech disciplines.
Adra-e has 13 partners
Inria, Universiteit van Amsterdam (UvA), Universiteit Twente (UT), ATOS SPAIN SA, Commissariat à l'énergie atomique et aux énergies alternatives (CEA), TRUST-IT SRL, Linkopings Universitet (LiU), Siemens Aktiengesellschaft, Dublin City University (DCU), Deutsches Forschungszentrum fur Kunstliche Intelligenz GMBH (DKKI), National University of Ireland Galway (NUIG), AI, Data and Robotics Association (Adra) et Hrvatska Udruga Za Umjetnu Intelligenciju (CroAI).
Adra-e will work closely with CSA AI4Europe which develops the AI on demand platform.
What are the scientific challenges of this European coordination?
First of all, it is necessary to clear up a misunderstanding: the challenges here are not directly scientific but rather cultural.
First of all, we have to understand that these three domains are today closely linked, and must be understood in their entirety: the recent rise of AI has only been possible because of the availability of data, and robotics (except for the mechatronics part) is developing today mainly through the increase in the number of sensors that each robot can have at low cost, generating masses of data whose efficient processing requires AI. But it is not a question of developing just any technology in these fields. It is essential not to dehumanize the sector, or else there will be a massive rejection of the possible advances and the threats that these technologies may pose to the fair and democratic society that France and Europe want.
Trustworthy human-centered AI, data and robotics are technologies that must be safe, inclusive and sustainable, respecting people's basic needs, while reflecting the values that all citizens in Europe enjoy or aspire to enjoy, even, and especially, if these values are not universally recognized and enforced (see EC White Paper on AI, 2020 + "Report on the Safety and Accountability Implications of Artificial Intelligence, the Internet of Things and Robotics").
Despite recent advances in regulation, standardization, and technology development, much work remains to make this vision a reality.
There is an opportunity for Europe to become a world leader in human-centric and trustworthy AI, data and robotics, but this ambition can only be achieved by mobilizing, in the face of global competition, all relevant stakeholders and the necessary resources. With its cultural diversity and strong commitment to a sustainable, human-centered transformation of our society and economy, Europe is already well positioned. However, the European AI, data and robotics ecosystem is still quite fragmented, slowing down the progress that is still needed.
Innovation in AI, data and robotics is not a linear process (increasing TRL from research to industrial implementation), but rather relies on a complex and circular dynamic of various actors (research, industry and policy). Some of them play an active role in the ADR partnership (mainly large companies and applied research), but not all. While young players (startups or unicorns) are unlikely to play an active role in the partnership, they are very important for disruptive innovation in the relevant fields in Europe. The partnership must therefore develop ways to include them and ensure that their needs are taken into account. This includes bridging the gap between the top-down activity of the partnership and the regional ecosystems that are often the birthplace of innovation.
What are the complex dynamics to which Adra (and thus ADRA-e) can contribute?
Standardization, regulation (both in its drafting and enforcement), and trust are critical to the adoption of European AI, data, and robotics technology.
Yet trustworthy AI, data, and robotics is a moving target that evolves over time due not only to scientific and technological advances, but also to opinion movements related to the big issues of the day. Standardization, regulation, and trust activities progress in parallel with technological advances, and the whole must be closely coordinated to achieve the set goals.
The socio-economic opportunities on the demand side of AI, data and robotics technologies are enormous. But this is a "winner takes all" area, as the success of the technology giants, now mostly non-European, shows. Europe has the opportunity to lead the development of trustworthy AI, but supply/demand coordination must be consolidated to ensure global competitiveness.
What will be the priority actions?
The project will be punctuated by annually recurring events as well as by 14 interdisciplinary workshops that will mark important stages in the development of synergies within the ADR community and interactions with the various actors in society: citizens, companies, administrations, etc.
Four types of events are planned:
- ADR Conference, which aims to bring together the ADR community;
- ADR Project Exhibition, in which the various projects of the ADR landscape will be able to present themselves and establish links with other projects that are thematically similar but do not necessarily belong to the same programs;
- ADR Awareness Day, aimed at a varied public: citizens, industry, commerce, administration;
- Convergence Summit, which will bring together high-level actors (business leaders, elected officials, senior civil servants, etc.) to discuss the challenges of the future.
Note that the first edition of the ADR Conference will be held in October 2022 in Brussels and proposes to bring together CSA in the field of AI, data and robotics.
In addition, a launch event for the 42 new projects funded by the EC will also be held on October 17.
Funded by the European Union. The views and opinions expressed, however, are those of the author alone and do not necessarily reflect those of the European Union or the European Commission. Neither the European Union nor the funding authority can be held responsible for them.