This is a concept/proposal submitted to a competition or in preparation, not an active project.
Cloud Technologies · DataCompetition closed
An edge-AI platform that combines solar generation, biogas, storage, and EV charging into a single controllable flexibility resource for distribution grids with a high share of renewables.
Research and development together with students and the Center's laboratory.
Overview
About the project
Edge AI Multi-Source Power Balancing Service for EV Charging in RES-Rich Distribution Grids. The platform has three layers: an Edge Energy Optimizer — PV and biogas generation forecasting with LSTM/XGBoost and battery dispatch with deep reinforcement learning (Deep RL), running inference on NVIDIA Jetson Orin Nano; a Smart Charging Controller — managing charging stations via OCPP 2.0.1 Smart Charging Profiles; and an integration layer built on the open NGSI-LD / Orion-LD data models and TMForum Open APIs. Target results (TRL 6→7) include lower charging cost and peak consumption, edge-inference latency under 200 ms, and a replicable template for "PV + storage + biogas + EV" sites. The project is carried out within the COP-PILOT programme (Horizon Europe) together with INFOCOM (a manufacturer of charging stations and storage systems, a Siemens Industrial Partner) and the ZNU HPC/ML laboratory; the pilot site is Preveza, Greece.