Battery-informed Resilient Energy Scheduling Framework for EMS-BMS Coordination in Smart Buildings

Objective

Buildings account for ~40% of final energy use and 36% of CO2 emissions in the EU, making them critical for achieving the European Green Deal targets. Today, Energy Management Systems (EMS) optimise energy costs while Battery/Building Management Systems (BMS) ensure safety, but they work separately. This leads to inefficient energy use, faster battery degradation, and higher emissions. As EU flexibility needs are expected to more than double by 2030, there is an urgent need for intelligent EMS–BMS coordination that integrates renewable generation, storage, and digital optimisation. The BRAIN fellowship (Battery-informed Resilient Energy Scheduling Framework for EMS-BMS Coordination in Smart Buildings) will develop the first integrated, AI-driven EMS–BMS coordination framework for smart buildings. The approach combines: (i) multi-task deep learning forecasting of load, PV, and battery health, (ii) EMS optimisation using hybrid Model Predictive Control and Deep Reinforcement Learning with explicit battery health constraints, (iii) privacy-preserving synthetic datasets enabling reproducible and GDPR-compliant research, and (iv) validation on real-world data. By embedding battery health into EMS decisions, BRAIN will reduce household energy bills by 10–15%, extend battery lifetimes by 20%, and increase PV self-consumption by up to 20%. Scientifically, it will deliver 7 peer-reviewed journal and conference papers, 2 open FAIR datasets, and an open-source EMS–BMS toolkit. Societally and economically, the project will contribute to lower emissions, reduced e-waste, and affordable clean energy, while strengthening Europe’s leadership in the €250B global smart building market. BRAIN directly supports the European Green Deal, the MSCA Green Charter, and A Europe Fit for the Digital Age, creating measurable impact in science, industry, and society.