AIRAH Awards 2022 winner
Conserve It
Real-time model predictive control with digital twins and edge computing technologies
This comprehensive initiative researched, developed and productised machine-learning-based model predictive control to minimise the energy usage of chilled water plants.
The solution was crafted to operate fully on the edge on low-cost embedded controls and, as such, extensive research was conducted to develop machine learning and mathematical optimisation frameworks that can resolve quickly with limited computing resources. This allows for the elimination of security and reliability concerns that often arise with offsite controls, as well as ongoing cloud subscription fees.
The solution was deployed at numerous sites and delivered demonstrable energy savings of up to 18 per cent.