Advancing building energy science at the intersection of machine learning, human behavior, and urban systems.
Modeling how people interact with buildings and move through cities — from individual window operation to urban-scale mobility patterns — to inform energy-efficient, resilient, and human-centered building design.
Deploying and benchmarking next-generation control strategies in real buildings — including reinforcement learning, model predictive control, data-driven predictive control, differentiable predictive control, and agentic AI frameworks for fault detection and autonomous operation.
Integrating buildings with the power grid through demand response, EV charging optimization, and PV-battery management. Exploring quantum annealing and quantum computing for real-time building-grid control at scale.
Developing physics-consistent neural networks and modular surrogate models that embed thermodynamic and system constraints — enabling faster, more accurate, and more deployable building simulation and control.
Measuring and modeling indoor air quality, infection risk, thermal comfort, and ventilation effectiveness — including airborne COVID-19 transmission and smart energy recovery ventilators for schools.
Researching low-carbon building envelope solutions — from mycelium-based composite insulation panels (MycoCore) to 3D-printed siding — and developing human-robot collaborative workflows for large-scale building retrofits.