PROGRESSUS: how to reduce the CO2 emissions in smart grids with AI


In support of Europe’s 2030 climate targets, the EU-funded PROGRESSUS project (ECSEL Joint Undertaking) aims to introduce a next-generation smart grid demonstrated by the application example of a smart charging infrastructure integrating seamlessly into current smart-grid architecture concepts. In PROGRESSUS, CTTC is developing a Federated Deep Reinforcement Learning (FDRL) framework to optimally control the peer-to-peer energy exchange of prosumers with photovoltaic-based energy generation panels and battery energy storage systems. Smart contracts and blockchain technology are employed to secure the trading transactions. This allows for a more accurate and privacy-preserving prediction of energy demand. The developed system minimizes the external energy supply and subsequently decreases the cost and CO2 emissions. The project has an overall budget of 19.6 M€ and the consortium counts with major industrial players on board like Infineon AG (project coordinator) or STMicroelectronics.