Optimization of HVAC energy consumption with artificial intelligence

The amount of available data about energy usage, indoor environment, occupancy from buildings is constantly growing due to the installation of more and more sensors driven by the IoT development. However, resources to analyze and work with the data are limited in the Danish municipalities so it is expected that there is an optimization potential that can be released by using artificial intelligence technologies and methods on the data.

The purpose for this project is to uncover and implement energy saving potentials, indoor environment improvements and proactive maintenance based on artificial intelligence and machine learning on available energy data and building data.

The project will run for 2 years and will aim at collecting energy data from up to 400 ventilation systems and combine this with data about the indoor environment and other public available information including data from the building register (BBR), location, energy classification registers etc. The data will be analyzed and continuously fed to self-learning algorithms through the project period.

The results from the algorithms will be handle in an energy management and monitoring center that will work together with the local maintenance staff to investigate the result and implement relevant optimization potentials in terms of adjustment in BMS systems, technical maintenance and changes in usage/behaviour.

The partners in the project are IQ Energy Nordic who will be responsible for the energy management part including collection the data from the buildings and sending them to the AI platform as well as running the energy management and monitoring center. Ento Labs will be responsible for analyzing the data on their AI platform. The municipality partners Høje-Taastrup, Rudersdal and Frederiksberg will provide access to buildings and ventilation systems. They represent different experiences and infrastructure with regards to data driven energy management and will as such contribute to uncovering of the potential for upscaling of the project afterwards.