Automated Fault Detection & Diagnostics (FDD)
Most HVAC and building system faults waste energy silently for months before anyone notices. We build automated fault detection and diagnostics models that catch them in days.
A specialised application of machine learning, not generic analytics
Fault Detection and Diagnostics (FDD) is a well-established but underused discipline: applying rule-based and machine-learning models to continuous BMS and sensor data to automatically flag abnormal equipment behaviour — a stuck damper, a failed sensor, simultaneous heating and cooling — long before it shows up as an occupant complaint or an inflated energy bill.
This is a deliberately narrower, more technical service than our general building data science work: it's specifically about continuous, automated fault surfacing across HVAC and building systems, rather than broader forecasting or reporting.
- ✓Catch faults in days, not months — Continuous monitoring flags abnormal behaviour automatically, rather than waiting for a complaint or bill spike.
- ✓Prioritised by impact — Faults are ranked by estimated energy and cost impact, not just flagged in a long list.
- ✓Works with existing BMS data — Models are built on your existing trend data — no new hardware required in most cases.
- ✓Reduces reactive maintenance — Shifts maintenance effort from reactive firefighting to targeted, planned fixes.
Sectors we support
Commercial real estate
Portfolio-wide fault surfacing across multi-tenant buildings.
Local authorities
Automated oversight across public building estates with limited facilities staff.
Healthcare & industrial facilities
Early fault detection where downtime or discomfort has high operational cost.
Universities & research estates
FDD deployment across complex, high-diversity campus buildings.