Sustained temperature out-of-range.

Trigger, action plan, and escalation criteria.

Related metric

temp_compliance

Trigger

space_temp outside target_band for duration > dwell_threshold

Value drivers

Customer experience Avoided truck rolls

Preconditions

Expected band + dwell window + continuous data stream.

Human role

Primary — dispatch decision.

Action plan

(1) Validate data continuity (any gap in the stream?). (2) Cross-check Runtime and Cooling DeltaT to localize the issue (capacity, control, or load mismatch). (3) Attempt remote remediation (setpoint or schedule adjustment). (4) Dispatch if remote action does not resolve.

Escalation

(1) Dwell exceeds 4 hours during business hours → dispatch within SLA; comfort impact is likely. (2) Data gap suspected → escalate internally to the technical team; do not report the site as a real failure until the gap is resolved. (3) Same site recurrent > 3 events in 30 days → schedule a capacity assessment; the comfort-failure pattern may indicate undersized equipment. (4) Out-of-range correlated with a known schedule or setpoint change → roll back the change and reclassify as an operator-induced event.

Prevention

(1) Data-continuity monitoring at platform layer — gap detection at minute-level cadence prevents the gap from being misclassified as a comfort failure. (2) Dwell-window calibration per zone — too-short dwell windows fire on transient events that aren't real comfort issues; too-long windows hide real failures. (3) Site-change watchlist — any site with a recent remodel, equipment swap, or capacity change gets tighter monitoring for the following 30 days to catch undersizing early.