BIGGER PICTURE Energy & climate
Teaching an old factory when to start warming up
At Milan’s La Forgiatura campus, software reached into existing heating and cooling controls instead of replacing the plant.

- Reported saving
- 26.1% for controlled equipment
- Energy
- 183,300 kWh in the reported period
- Physical changes
- Existing building controls reused
How the idea works
Connect existing controls
A gateway links to the building management system.
Anticipate thermal demand
Models predict conditions and choose operating settings.
Adjust and meter
Start times and water temperatures change; energy is tracked.
A steelworks became a place to work
La Forgiatura in Milan is a former industrial site converted into a campus of offices and other workspaces. Its heating and cooling equipment already had a building management system. The project’s question was practical: could better decisions about that equipment reduce waste without first buying a new heating plant or undertaking another major renovation?
Warm up for the weather that is coming
BrainBox AI connected a gateway to the existing controls. One algorithm, Optimus, chooses when heating, ventilation and cooling should start or stop using predicted conditions. Another, Water-Clock, adjusts water-delivery temperatures and related control settings to expected demand. The physical idea is anticipatory: provide the temperature the building needs without unnecessary running time or excessive conditioning.
The result belongs to the equipment controlled
A REHVA Journal account by project participants reports 183,300 kilowatt-hours saved, or 26.1%, for controlled equipment over its 2022–2024 reporting period. Nine buildings were driven or partly driven by the system in that account. Those boundaries matter: the percentage is not a measured reduction in every form of energy used across the whole campus.
Sources: [1]
Compare periods before comparing percentages
A separate, undated installer page now describes ten buildings and a different average saving. It does not provide a directly comparable measurement period, so the figures should not be blended. The concrete inspiration is the retrofit method: make existing controls more responsive to anticipated conditions, then verify savings against a clearly defined equipment and time boundary.
The original sources
- REHVA Journal; project-team authorsResults from the application of smart-ready technologies
- R2M SolutionLa Forgiatura project
- R2M SolutionBrainBox AI building control technology