AI Digital Twin, Cool Smarter with Human Feedback, Targeting Carbon Zero
SmartCool
is an AI digital twin using DEMMFL & RL-MPC. It optimizes chiller supply and room settings (temp/fan/mode) via user-feedback devices, ensuring a closed-loop balance between energy efficiency and occupant comfort.
Key Features & Advantages
Closed-loop, self-optimising AI-embedded HVAC control for optimal efficiency in real-time
Multi-modal predictive engine to forecasts cooling demand
Live interactive informative digital twin
Significant energy and cost savings for building management team
Urban sustainability and carbon neutrality: Minimise overcooling and discomfort based on actual demand from users
SmartCool is an advanced AI-powered digital twin platform designed to mitigate the prevalent issue of urban building overcooling. It employs a dynamic, closed-loop optimisation framework that continuously refines HVAC operations for optimal performance and enhanced comfort.
The platform integrates two sophisticated AI modules: the Dynamically Engineered Multi-Modal Feature Learning (DEMMFL) predictive engine, which delivers precise cooling demand forecasts based on real-time data, and a Reinforcement Learning-enhanced Model Predictive Control (MPC) system that determines optimal equipment setpoints. Its interactive 3D digital twin dashboard offers comprehensive, real-time visualisation of system performance, enabling facility managers to efficiently balance energy use and operational costs.
