MQ Flow runs a building's heating, ventilation and humidity from a physics model of human comfort, then learns what the people in each room actually prefer. Occupants never set a temperature. They tap warmer or cooler, and the system does the rest.
We build the AI and the software. Our HVAC partner Mifimi brings the domain knowledge that makes it work in a plant room. This is a work note, not a case study: the building has not yet produced the numbers a case study needs.
The problem
A thermostat controls air temperature, but people feel radiant heat, humidity and air movement as well. So buildings run warmer or colder than anyone needs, and the heat pump does work that fresh air or a humidifier could have done for less. MQ Flow starts from what people feel, not from a setpoint.
How it works
- Baseline: Physics first. A Fanger PMV comfort model (ISO 7730) computes each room's comfort target from temperature, radiant heat, humidity, and air movement. It works from day one, before any learning.
- Personalise: Learns each room. Per-room reinforcement learning turns warmer and cooler taps into a personal comfort offset, tuning to the actual people in the room rather than a statistical average.
- Optimise: Cheapest levers first. The same comfort can be reached several ways. The optimiser adjusts fresh air and humidity before touching the heat pump, targeting at least 30% lower energy use.
What we chose, and why
- Comfort is computed, not guessed: PMV weighs radiant heat, humidity and air movement, not air temperature alone.
- Occupant taps are the training signal for the reinforcement learning, so the system improves with use.
- Built to drop into existing buildings: BACnet integration and one API toward any PLC.
- The control loop runs on hardware in the building, per-room thermostat units on our edge stack plus a weather station, so it keeps working without a cloud connection.
Where it stands
The first batch of 28 thermostat units is going into a faculty building in Belgrade, where data collection and model use are being evaluated on the live building.
Verified numbers
None yet. At least 30% lower energy use is the project target, not a result. We publish measured figures when the evaluation closes, and not before.
Mifimi is named with written permission. MQ Flow is looking for design partners with a building to instrument; if that is you, book a call.