Aerate
Aerate provides AI-driven HVAC optimization for commercial buildings and data centers, continuously learning how each room and VAV box heats and cools to reduce energy use while maintaining comfort. The system adjusts start-stop times, leverages free outdoor air, and fine-tunes fan pressure and reheat, delivering up to 35% lower energy costs without new hardware. It retrains nightly and runs small, safe tests within comfort limits to keep improving savings over time.
- Clean Technology
- Artificial Intelligence
- Energy
- Internet of Things
- Software Only
Funding
Founders
Founder details are not available yet.
Product
Problem
Commercial buildings and data centers typically run on legacy building management systems (BMS) with static control sequences that fail to adapt to changing conditions, occupancy patterns, and equipment drift. This leads to significant HVAC energy waste—from overcooling empty rooms, reheating air that was just cooled, and running fans at excessive pressure—while occupants still experience comfort issues.
Solution
Aerate provides an AI-driven HVAC optimization platform that continuously learns how every room and VAV box in a building heats and cools, then runs the building on the least energy that keeps everyone comfortable. The system adjusts start and stop times, opens economizers to leverage free outdoor air, drops duct static pressure when boxes are under-utilized, and reduces reheat at the box level. It operates as a read-only overlay on existing building controls, requiring no new hardware installation. The AI retrains every night with the day's data, runs small safe tests within comfort limits to validate savings, and adapts to building changes like new tenants or stuck dampers without recommissioning.
Target Audience
Primary customers are commercial building owners, facility managers, and data center operators seeking to reduce HVAC energy costs without replacing existing building automation infrastructure.
Features
- AI model that learns per-room and per-VAV-box heating and cooling patterns to predict optimal setpoints
- Automated adjustment of start-stop times, economizer use, duct static pressure, supply air temperature, and reheat
- Sandboxed AI that can read points and send setpoints but cannot reach the network, files, or internet
- Hard-coded safety limits on room and inlet temperature and backup cooling that the AI cannot override
- Cloud deployment with a single outbound firewall destination, or on-prem deployment with an open-weight model and no data leaving the building
- Nightly retraining with each day's readings to sharpen predictions week by week
- Small, safe tests that nudge one setting at a time within comfort limits, measure savings, and keep what works