PLATONIC (deeP LeArning neTwOrks for building’s eNergy effICiency) uses cutting edge technologies, such as Deep Reinforcement Learning (DRL), to unlock up to 44% in cost reduction all while preserving user’s comfort.

In line with REACH goals, PLATONIC will establish a high-value data chain, collecting and exploiting data from CERTH’s indoor and outdoor energy consumption and health dataset, as well as weather predictions and Sensinov’s datasets to model building occupancy and propose a set of recommendations that will improve HVAC operation based on:

External factors, such as weather and building occupancy;
Component optimum operation points, which represent the actual (mechanical) components found within HVAC systems.
The combination of these two approaches should yield the highest impact on the efficiency & performance of HVAC systems.