Low temperature excursion.

Trigger, action plan, and escalation criteria.

Related metric

refrigeration_temperature_compliance

Trigger

asset_temperature < configured_lower_band, sustained ≥ persistence_window (with sensor health and product category confirmed)

Value drivers

Product loss prevention Energy savings Avoided truck rolls

Preconditions

Temperature sensor operational and validated; lower band and persistence window configured per asset and product category; the asset's product category (refrigerated vs frozen) recorded so freezer behavior is not mis-scored.

Human role

Primary when a dispatch or store communication is required; secondary review after AI on classification and validation.

Action plan

(1) Validate the sensor and data quality first — a low reading is often a drifting or failed probe rather than a real over-cool; exclude invalid, stale, or suppressed intervals. (2) Classify the event: (a) excessive cooling or a control failure driving the asset below its band; (b) freezing risk for refrigerated product held too cold; (c) normal freezer behavior that remains inside its own configured band; (d) an invalid or drifting sensor reading. (3) Confirm the band matches the asset's product category — a frozen asset reading well below a refrigerated band is expected, not an excursion. (4) Where a controllable cause is found (a stuck control state, a defrost or setpoint issue exposed by the controller), attempt a remote correction first. (5) For genuine freezing risk to refrigerated product, notify the store so product can be checked or moved; record the outcome. (6) Scope a dispatch only when remote triage cannot resolve a real control failure, with the correlated evidence attached; capture closure evidence with root cause on close.

Escalation

(1) Freezing risk confirmed on refrigerated product of a high-value category → out-of-cycle store notification with a product-check recommendation. (2) A control failure that remote triage cannot correct → dispatch within SLA with a scoped diagnostic starting point. (3) The event traces to a sensor drift pattern across several assets → pause low-excursion alarms for the affected assets and reconcile the sensors before working them as excursions. (4) The same asset over-cools repeatedly after a repair → flag for controls review at the MBR.

Prevention

(1) Configure the lower band and persistence window per asset and product category, and record refrigerated vs frozen, so normal freezer behavior is never scored as a low excursion. (2) Maintain sensor-calibration and drift checks so a drifting probe is caught as a data-quality issue rather than dispatched as an over-cool. (3) Trend low-side behavior per asset so a control that is slowly over-cooling is visible before it freezes refrigerated product.