Buildings are one of the major sources of global energy consumption. In 2021, residential and commercial buildings were responsible for around 39% of total U.S. energy consumption and 74% of total U.S. electricity consumption. Consequently, research on the operation of building Heating Ventilation, and Air Conditioning (HVAC) systems can lead to significant energy savings and carbon emission reduction.

Our goal is to construct a platform that empowers researchers with access to extensive datasets, the ability to construct reinforcement learning simulations, and the capability to autonomously validate new approaches using the datasets provided on the platform.

We anticipate that this platform will stimulate innovation, promote collaboration, and expedite discoveries in building research.

Bear-Data: real-world building

We introduce BEAR-Data, an open dataset that captures the dynamics of a multi-zone building environment by providing measurements of zone temperature and corresponding HVAC (Heating, Ventilation, and Air Conditioning) control actions for over 80 thermal zones.

Bear-Simulator: physic-principled simulation

Bear-Simulator is a physics based Building Environment for Control and Reinforcement Learning. Bear Simulator allows researchers to benchmark both model-based and model-free controllers using a broad collection of standard building models in Python without co-simulation with other building simulators.