Project objectives, and Third-party Financing
ENFIELD will create a unique European Centre of Excellence that excels the fundamental research in the pillars of Adaptive, Green, Human-Centric, and Trustworthy AI that are new, strategic and of paramount importance to successful AI development, deployment, and acceptance in Europe and will further advance the research within verticals of healthcare, energy, manufacturing and space by attracting the best talents, technologies and resources from world-class research and industry players in Europe and by carrying out top-level research activities in synchronisation with industry challenges to reinforce a competitive EU position in AI and create significant socio-economic impact for the benefit of European citizens and businesses.
The foundational research activities will be implemented across carefully chosen four horizontal topics or so-called pillars :
Green AI it is time to transition to Green AI by developing novel approaches to ML that minimize carbon footprint while maintaining performance.
Adaptive AI Adaptive AI systems are necessitated in situations where rapid changes in the external environment or changing enterprise goals require an optimized response.
Humancentric AI ENFIELD aims at designing novel techniques aimed at enhancing the explainability of AIS systems, providing a comprehensible description of the internal procedures used while the model is trained or makes a (sequence of) decisions, explaining a particular result to both the users and the designers, providing context-aware explanations for different stakeholders with different user profiles.
Trustworthy AI The following advances of the state-of-the-art will be developed within ENFIELD: Explicit formal approaches to develop trustworthy AI tools: Design formal approaches to abstraction and modelling of uncertainty in AI tools, advance research on specifications, robust learning approaches and on data quality in learning tools.
The latter two stem out of Europe AI strategy, Adaptive AI is key to AI dissemination in real world applications, Green AI stem out of Europe Digital and Green priorities.
The research aspects will be further enhanced crosswise along four critical infrastructure sectors (verticals) whose virtual and physical assets, systems, and networks are considered vital in Europe and their disruption would have a debilitating effect on society. These verticals are:
Energy The renewable energy and energy storage industry also applied AI in use cases such as short-term forecasting and predictive maintenance, dispatch of energy batteries. Recent breakthroughs in AI (e.g., deep learning, transfer learning, reinforcement learning, automated ML) are boosting once again the AI potential for the energy sector.
Healthcare Recent advances in health digitalization, sensory data, simulated systems biology data, lead towards clinical Big Data, given the growing volume of data as well as their inherent complexity the analysis and inference become rather demanding.
Manufacturing The key question the manufacturing vertical is going to address is: how to incorporate Green, Adaptive, Trustworthy and Human-centric AI in the new Industry 5.0 paradigm and along the dimensions of Smart Autonomous, Product-Service and Hyperconnected Factories of the Future?
Space This project aims to tackle the main challenges for the adoption of AI technologies in the Space and AAM industry, especially for operational phases.
In May 2022, the EC has presented the REPowerEU1, aiming to transform Europe's energy system to end the EU's dependence on Russian fossil fuels and tackle the climate crisis. Therefore, the energy vertical has been selected due to its significant impact and political priorities and further reasoned by potential contributions to adaptation to climate change. The healthcare vertical is strongly related to fairness, privacy, explainability, ethics and legal aspects of AI; these are the primary concerns for responsible and transparent use of AI technologies, through a human-in-the-centre approach, while manufacturing is crucial for economic operation in Europe and the efficient use of AI can greatly boost production and reduce costs. Space is becoming a domain of high importance at the global scale, thus, has been chosen to continue to position the EU as a leader in this field and ensure that the EU can access and use space with a high level of autonomy. Green and Adaptive AI research results can be directly applied to the space domain.
Further, the Open Calls to be implemented within ENFIELD will have a tremendous impact to its outputs
- 76 researchers will be selected (15 per each pillar and 4 per each vertical) following Third-party Exchange Scheme (TES) through the Open Calls. TES is designed to attract the top-level researchers to conduct foundational research activities related to specific scientific and technological challenges raised by the relevant pillar or vertical, to disseminate the most advanced knowledge and contribute to the ENFIELD network creation and expansion to European AI labs and countries that are not presented in the consortium.
- 18 small-scale projects will be funded (4-5 per each vertical) following Third-party Innovation Scheme (TIS) through the Open Calls. TIS is designed to conduct applied research activities in energy, healthcare, manufacturing, and space verticals in the context of the use cases proposed by the ENFIELD industry players, while these use cases link to the pillars. The selected small-scale projects will also contribute to networking activities and disseminate the latest knowledge in Europe. This scheme is applicable for R&D intensive establishments (research centres, start-ups, SMEs and large companies).