Based on the data acquired during the building exploitation, HEATwin will automatically create the most adequate predictive models of the internal energy production and consumption, which will enable simulation of the data chain for any hypothetic operation plan.
The results obtained from the simulations, along with all other micro-grid features and external factors, will be subjected to an optimization process in order to find the energy management pattern that will result with the most economical usage of energy under the given conditions. Creation of predictive models is performed by our existing automated machine learning (AutoML-MLops) platform, Blackfox. Daily optimization of operation plans relies on OSICE, our service for optimization in Cloud environment.
This solution will be offered to end users in the form of SaaS, thus enabling them to optimize energy savings without investment in scarcely available data science and optimization experts, or in the necessary computing infrastructure.
Usage of Standards for data interoperability:
HEATwin supports modern industry standards for the entire life cycle of generated models. Starting from tracking model versions, through storing them in the registry, all the way to serving over HTTP, the entire pipeline is compatible with MLFlow, de-facto industry standard for MLC. Moreover, model metrics are available both through the MLFlow interface and through Blackfox custom interface.
