Central plantCentral Plant/Chiller
Increases overall plant plus individual chiller efficiency through improved heat rejection at the cooling tower(s).
TechnologiesFacil.AIAgents and gateway
Equipment-specific agents, one gateway, your existing BMS.
How Facil.AI's Physical AI connects to the plant you already run: the equipment each agent covers, how the software gateway works, and the systems it is built to work with.
The agents
One family across central plant, air side and refrigeration.
Each agent is built for one equipment family. Together they cover central plant, air side and refrigeration, using the same learning and control method.
- Central plant
Air sidePortfolio RTU
Increased comfort and reduced costs across hundreds of RTUs in a portfolio.
Air sideAHU
Dynamically manages air distribution and thermal conditions.
Air sideAir Source
Balances outside versus conditioned air dynamically.
RefrigerationRefrigeration Rack
Increased energy efficiency through AI enhanced “floating” control.
RefrigerationDefrost
Dynamic scheduling of defrost cycles, using AI store traffic analysis.
The gateway
A software gateway, not a hardware project.
Facil.AI connects to the plant through a plug-and-play software gateway that talks to your BMS in the protocols it already uses. Nothing is ripped out and nothing new is bolted to the equipment.
- Connect
The gateway connects to the BMS and the controllers it already manages.
- Read
Live equipment data flows to the agents, around the clock.
- Run
The agents continuously adjust system behaviour, based on multiple live variables, through the existing controls and within each machine's operating envelope.
- New hardware
- None
- Capital cost
- No CapEx
- Connection
- Software gateway
- Integrations
- 30+
Compatibility
Built for the BMS you already run.
The gateway works across 30+ BMS and control integrations. These are examples of the ecosystems it is built for. If your site runs something else, ask us on the intro call.
CarrierBuilding controls
SiemensBuilding controls
HoneywellBuilding controls
Johnson ControlsBuilding controls
Schneider ElectricBuilding controls
ABBBuilding controls
TraneBuilding controls
Automated LogicBuilding controls
DanfossBuilding controls
TridiumBuilding controls
Mitsubishi ElectricBuilding controls
Andover ControlsBuilding controls
DaikinBuilding controls
Delta ControlsBuilding controls
EmersonBuilding controls
RobertshawBuilding controls
OPC UAProtocol
ModbusProtocol
BACnetProtocol
SiebeBuilding controls
The marks identify example BMS, control-system and protocol ecosystems for fit discussions. Names and marks belong to their owners and do not indicate partnership, endorsement or customer status. Individual site, controller and configuration fit requires assessment.
Equipment map
Where each agent sits in the building.
Each agent is equipment specific. They run 24/7 as "relentless, unattended supervisors" babysitting your system and sending precision commands back to your BMS every 5 minutes, compounding gains.
01 Equipment families on site
Central plant
Air side
Refrigeration
02 Your building automation system
03 Connection
04 Physical AI
Equipment-specific agents, running unattended, 24/7- Observe Live tons, kW and kW/ton
- Learn How this plant behaves
- Act A precise command back to the BMS
Illustrative simulation, not customer data
Track record
One central plant, measured.
48%Central-plant efficiency · one campus
California State University, Channel Islands: a chiller plant on an existing building automation system.
In Facil.AI's published case study, the campus central plant improved its efficiency by 48% within 45 days, working through the controls already in place. The figure belongs to that plant. Results elsewhere depend on the equipment and how it runs today.
- Campus
- 21 buildings
- Plant efficiency
- 1.15 to 0.6 kW/ton
- Time
- 45 days
Resources
Read further before the call.
Published by Facil.AI. Each opens on facil.ai in a new tab.
Case studyHow AI improved central plant efficiency by 48% at a California State University campusThe full central-plant story, from the starting point to the measured result.
Open on facil.ai (opens in a new tab)
WhitepaperVirtual closed loop control for adaptive energy management in retail buildingsHow closed-loop control adapts energy use to real-time conditions in retail environments.
Open on facil.ai (opens in a new tab)
LibraryAll Facil.AI resourcesWhitepapers, case studies and videos from the Facil.AI team.
Open on facil.ai (opens in a new tab)
Start with the overview.
Facil.AI overview Why Physical AI, where it fits and how Cool Estate supports the rollout in Canada.
Bring your BMS details. We'll map the agents.
Twenty minutes with Cool Estate to check your controls, your plant and where Facil.AI could start.