The Decision Support Systems Laboratory of ECE of the National Technical University of Athens (EPU-NTUA) is among the authors of the new European roadmap for the development of Artificial Intelligence (AI) foundation models for electricity grids, a strategic initiative for the future of the European Union’s energy system.

The roadmap, entitled Plan4EUAI, was discussed and consolidated during the “Workshop on the European AI Foundation Model for Energy Grids”, organised on 8 December 2025 in Brussels by the European Commission (DG ENER). The workshop was moderated by EPU-NTUA, Fraunhofer FIT and INESC TEC, and brought together key stakeholders across Europe, including electricity transmission and distribution system operators, research organisations, European AI Factories, sectoral associations and representatives of the European Commission.

The workshop aimed to explore how AI foundation models can support the digitalisation, resilience and operational efficiency of European electricity grids. Discussions focused on the technical, regulatory and organisational challenges associated with the development of large-scale AI models for the energy sector, including data access, computational requirements, interoperability and secure information sharing.

Building on these exchanges, EPU-NTUA, INESC TEC and RWTH Aachen jointly presented Plan4EUAI, a structured roadmap proposing concrete steps towards the development of a European foundation model capable of supporting applications, such as forecasting, optimisation and digital twins of electricity grids. The plan outlines a sequence of activities covering the full lifecycle of the model, from the identification of priority use cases and datasets to large-scale training, validation and real-world evaluation with European operators.

EPU-NTUA’s contribution to Plan4EUAI focuses on the scientific and technical development of the foundation model itself. EPU-NTUA will lead the Working Group responsible for the design, training and validation of the first Pan-European AI Foundation Model for electricity grids. This work will address key challenges related to model architecture, learning strategies, robustness and performance assessment, ensuring that the resulting model can scale across diverse European power systems while meeting the sector’s reliability and security requirements.

EPU-NTUA was represented at the Brussels workshop by Dr. Elissaios Sarmas (Senior Researcher), who is a co-author of the Plan4EUAI roadmap. In addition to contributing to the drafting of the plan, he co-moderated workshop sessions dedicated to defining “quick-win” use cases and outlining the short-term framework for AI foundation models, facilitating dialogue between system operators, researchers and European institutions.

The workshop concluded with a shared commitment to move forward with concrete next steps, including the mobilisation of interested operators, the refinement of priority use cases and the establishment of a coordinated European structure for data, model development and governance.