Imagine a giant, safe playground where energy companies can test their AI software before letting it run the real power grid. Instead of risking a blackout, they use digital copies of cities and real-world hardware like wind turbines and batteries to see if their code works. It is like a flight simulator for the energy transition, ensuring AI is safe and reliable before it hits the market.
Companies developing AI for energy face high costs and risks when testing software on real grids. There is a lack of safe, standardized environments to validate 'trustworthy AI' before commercial launch.
A Testing Experimentation Facility (TEF) combining digital twins and physical assets (wind, solar, hydrogen, buildings) with an AI development environment and a risk database.
If you are a farm operator dealing with unpredictable power surges — this project developed a Testing Experimentation Facility that allows you to validate AI control software using real-world assets like wind power and photovoltaic systems. This reduces the risk of equipment failure during deployment.
If you are a manager dealing with inefficient heating and cooling in residential blocks — this project developed a high-fidelity digital twin of multi-apartment buildings in Riga. You can test AI energy-saving apps in a virtual environment before installing them in physical buildings.
If you are a startup dealing with high costs of testing new electrolysers — this project developed access to hydrogen testing platforms at CEA LITEN, CARTIF, BER and CIEMAT. This allows you to validate your AI-driven storage solutions without building your own expensive lab.
Based on available project data, the project is designing a self-sustainable business model that will include a subscription/plan framework and billing for users.
Yes, the facility is designed to support national AI regulatory sandbox initiatives and integrates 5 large-scale European testing facilities.
Based on available project data, the facility will be driven by a new entity and provides an open environment for innovators to bring products to market, though specific licensing terms are not detailed.
It provides an Acceptance Environment and a common open AI risks database for cybersecurity and trustworthy AI assessments.
The project period runs from 2025-01-01 to 2027-12-31, indicating the development and rollout happen within this window.
The consortium is well-balanced for commercialization, featuring 16 partners across 8 countries. With an industry ratio of 44% (7 industrial partners, 7 of which are SMEs), there is a strong focus on market needs. The presence of 4 research entities and 5 other organizations ensures a bridge between academic AI development and real-world energy infrastructure.
Contact the Research Center for Communication and Computing Network (REC) in Greece
Contact us to find out how to access the EnergyGuard testing environment for your AI energy product.
