A systematic review of 66 peer-reviewed studies published between 2023 and early 2026 examines how large language models are being tested in building HVAC operations. Deployment readiness remains severely bounded: 63 papers stay research-only, four reach pilot testing, and none demonstrate sustained real-world operation. While high-frequency control and short-horizon numerical forecasting remain firmly with conventional MPC and reinforcement learning, the review identifies pragmatic near-term value for LLMs as semantic and workflow layers—normalising sensor point names, aiding energy modelling, and supporting human operators.
No heat snapshots are available in the last 24 hours.