Failed temperature recovery.

Trigger, action plan, and escalation criteria.

Related metric

refrigeration_temperature_recovery

Trigger

asset does not return to configured_band within approved_max_recovery_window after a valid disturbance

Value drivers

Product loss prevention Avoided truck rolls Customer experience

Preconditions

A valid disturbance can be identified; the approved maximum recovery window and the asset's baseline are configured; temperature, door, controller, and runtime feeds validated.

Human role

Primary for the dispatch decision, store communication, and any product-check recommendation; secondary review after AI on evidence correlation.

Action plan

(1) Confirm the failed recovery is real — the asset has not returned to band within the approved maximum window, and the reading is validated (not a sensor fault). (2) Correlate temperature, door, controller, runtime, and peer-asset evidence before recommending dispatch — a failure shared across a rack points upstream, a single asset points to that unit. (3) Do not assert a specific mechanical cause the metric cannot confirm; use the correlated evidence to scope the investigation. (4) Notify the store and recommend a product check on the affected category when the temperature is already out of band; record the outcome. (5) Dispatch within SLA with the correlated evidence attached so the field visit starts targeted. (6) Capture closure evidence with root cause on close, feeding recurring-failure detection.

Escalation

(1) Failed recovery on a high-value or high-risk product category with temperature already out of band → out-of-cycle customer-ops alert with a product-check recommendation. (2) Remote correlation cannot explain the failure and product is at risk → dispatch within SLA with a scoped diagnostic starting point. (3) Several assets on one rack or site fail recovery together → suspect a rack, power, or connectivity event; escalate the site, not the individual assets. (4) The same asset fails recovery repeatedly after a repair → flag for asset or controls review at the MBR; candidate for replacement.

Prevention

(1) Configure the approved maximum recovery window and the asset baseline per asset and product category before activating the metric. (2) Trend recovery so a slowing curve is caught by the slow-recovery SOP before it becomes a failed recovery. (3) Keep a per-asset recovery and failure log so a recurring failure pattern is visible and feeds replacement planning.