The transition from centralized energy systems, which have long dominated the energy supply infrastructure, to more sustainable and resilient alternatives has become a priority in Europe. These decentralized energy communities (DECOs) are local energy networks, capable of operating autonomously or in connection with larger networks, whose members are both energy producers and consumers (prosumers).

This R&D project has three main scientific contributions:

• The development of an AI-based tool to predict the consumption of different types of households, taking into account uncertainties.

• The development of an AI tool for rapid and accurate forecasting of photovoltaic energy production, based on historical and meteorological data.

•  The optimization of energy exchange within and between energy communities, with buy/sell recommendations using blockchain for users.

The rise of electric vehicles—84 million expected in Europe by 2030—and the massive rollout of residential and public charging stations are making battery aging management much more complex. The rise of bidirectional charging exacerbates these challenges by increasing the number of charge/discharge cycles and introducing interoperability issues between infrastructures.

The project aims to model this aging more accurately based on realistic usage profiles, to use empirical and machine learning approaches to predict their health status, and to integrate these models into control systems capable of adapting charging and discharging strategies according to the actual degradation of the batteries.