Heating, Ventilation and Air Conditioning (HVAC) is responsible for keeping us warm and comfortable indoors. In a modern building, data such as weather, occupancy, and equipment use are collected routinely and can be used to optimise energy use. Reinforcement learning (RL) is a good fit as it can learn patterns in the data and identify strategies to control the system so as to reduce energy consumption to up to 20%. In this session we will be training a reinforcement learning algorithm with Amazon SageMaker using an open source simulator, EnergyPlus, to train an agent to optimise energy usage. We will then publish this model in AWS Marketplace for other customers to reuse your machine learning solution.
Aparna Elangovan, Solutions Architect, Amazon Web Services