Energy
Predict and optimise the timing of energy use, storage and thermal production.
TECHNOLOGY FOCUS · INTELLIGENCE
AI becomes interesting when it can act on the real world.
We are exploring how sensing, data, controls and AI can move products from fixed rules toward systems that can understand conditions, make better decisions and adapt.
INTELLIGENT SYSTEMS
THE NEXT STEP IN CONTROL
From programmed systems to responsive systems.
Traditional controls are often built around predefined rules: turn something on, turn it off, increase the setpoint, reduce the setpoint. That approach still has a place, but modern systems are producing more data and facing more variables.
NIST is now researching AI-optimised building controls that can use intelligent agents to make control decisions, diagnose faults and support common goals such as lower energy cost, occupant comfort and grid flexibility.
OUR INTEREST
Make intelligence part of the product.
We are less interested in adding AI as a marketing label and more interested in where intelligence can remove friction, improve performance or unlock a use that fixed control could not handle well.
“What should the product do differently when the conditions change?”
SENSE → UNDERSTAND → ACT
SENSECollect useful signals.
UNDERSTANDFind patterns and context.
DECIDEChoose the next action.
ADAPTLearn from the outcome.
WHERE ELSE CAN WE USE IT?
Start with the problem, not the AI.
Buildings
Coordinate comfort, IAQ, equipment performance and occupancy rather than controlling each system in isolation.
Maintenance
Use operating data to detect unusual behaviour before a small problem becomes an expensive one.
THE EVOLUTIONFX QUESTION
What can the product know?
That question can lead somewhere more useful than simply asking where AI can be added. It moves us toward products that can measure their environment, understand what is happening and respond to change.