Pilot project shows potential of AI energy management in non-residential buildings

30.04.2026

Source: Deutsche Energie-Agentur GmbH (dena)

A pilot project by the German Energy Agency shows how energy consumption in non-residential buildings can be efficiently controlled through the use of artificial intelligence.

An office location of the Fraunhofer IEE was examined as part of the SET pilots.

Interaction of AI, smart meters and dynamic tariffs

Photovoltaics, a heat pump and charging infrastructure for electric vehicles were used at the site. The system was controlled by an AI agent that took into account dynamic electricity tariffs, time-variable grid charges and grid-related requirements.

The result: the technical and regulatory infrastructure already enables reliable, market- and grid-oriented control. For example, electric vehicles were specifically charged at times of low electricity prices and high PV generation.

Potential for the energy system and operators

The project shows that AI-supported energy management systems can increase energy efficiency, reduce costs and contribute to grid stability at the same time. In the future, flexibility from non-residential buildings could be bundled and marketed on the electricity market via virtual power plants.

According to the project participants, the prerequisite for widespread application is a suitable regulatory framework, particularly with regard to grid fees and incentive systems.

You can read the entire press release on the website of the German Energy Agency (dena)