Anomaly Detection & Facility-Wide Insights

When three zones flag at once, it's probably not three separate problems.

Cross-zone correlation analysis runs alongside per-zone diagnostics — surfacing a likely shared cause instead of treating every zone as an isolated incident.

A concrete walkthrough

How a shared cause gets surfaced

08:12 — Z-04 flags

Health dips on Z-04. On its own, this looks like a routine zone-level watch alert.

08:40 — Z-05 and Z-09 flag

Two more zones show correlated early-stage stress within the same window.

08:41 — pattern recognized

The three zones share an air-handling loop. The platform clusters them rather than paging three separate alerts.

08:41 — probable cause

Flagged: possible HVAC drift on that loop. One investigation, not three wild-goose chases.

Facility-wide insight

Correlated anomaly across 3 zones (Z-04, Z-05, Z-09). Shared factor: air-handling loop 2. Probable cause: HVAC drift. Confidence rises as more zones on the same loop confirm the pattern.

Common shared causes detected

  • HVAC / air-handling drift across a loop
  • Water-source contamination affecting fed zones
  • Lighting-system fault over a bank of racks
  • Feed-batch or dosing error across a room
Why it matters

Fewer alerts. Better ones.

Reporting every zone independently buries the signal. Clustering correlated anomalies into one probable cause is a harder — and far more valuable — problem to solve.

Stop chasing symptoms across zones.