Currently, there are two active pilots in the project:
Energy intelligent grocery store: The purpose is to examine the AI (Artificial Intelligence) based energy management in two S-markets in Oulu. The aim is to inspect the applicability of the models of machine learning to the behavior prediction of the S-markets and the flexible components of the S-markets. In addition, the possibilities of the deep reinforcement learning in site-specific energy consumption optimizing is also inspected.
Heat pumps as a virtual power plant: The purpose of the pilot is to examine the applicability of different types of heat pumps to a virtual power plant. The data achieved through NIBE's Uplink interface from more than ten connected consumer on-line heat pumps and the data based machine learning models are used as a starting point. The heat pumps include geothermal heat pumps, air-water heat pumps and exhaust air heat pumps.