Measurement and traceability24 / 39

A lid sensor may show event timing; it cannot alone prove product temperature or the cause of a deviation. Match it with other records.

Agree shipment and device IDs, measurement point, period and file format in advance. Limit raw-data access to customer scope and permissions.

Spoilage in enzymes and cultures is invisible: the product looks the same while its activity falls. That makes the temperature record an output delivered alongside the goods.

Vaccines stay in the middle of the fridge at 2 °C to 8 °C. How a weekend excursion goes unnoticed, which device shows how long it lasted, and how the fridge log is matched with the ATS lot record.

4 °C is the most common storage temperature for environmental and water samples, yet pathogen, metal and some bacterial samples follow other conditions. Examples from EPA guidance.

Link plate batch, readiness check, route handoff and return condition in one shift record.

Do not assume the display is product temperature or a trip record. Verify its sensing point, device documentation and comparison with a reference.

A surface reading does not prove the core state. Log the measurement point, instrument, timing and manufacturer acceptance method together.

Before interpreting a curve, verify the model, load, ambient conditions, sensor location and test method.

Compare calibration status, sensor location and measurement time. Do not choose a reference until both devices measure the same item and point.

If the cable affects lid closure, the measurement setup has changed. Verify a manufacturer-defined pass-through or route it outside the seal.

Review battery performance by cell type, device and use conditions. If records have gaps, compare battery, temperature and communication logs.

A glycol bottle imitates a defined thermal response; do not assume product equivalence. Trial its fill, volume and placement against the load.

An infrared thermometer reads the radiation a surface emits. On a shiny surface it sees the reflection of its surroundings instead of the product's heat, so the acceptance decision rests on a probe placed between packs.

For readings near a limit, assess measurement point, instrument uncertainty and decision rule against the contract and procedure.

Assess sensor lag, sampling interval and mounting separately to test for brief peaks.