Summarize the same deviation list first by container ID, then by vehicle ID. Show run denominators as well as counts; grouping alone does not establish a cause.
Compare both views
Calculate rates by container and vehicle separately. Check that records cover the same period, product regime and definition of a run.
Hypotheses related to the vehicle
- Match any body-temperature record by location and time; do not assume sun or ventilation conditions.
- Compare loading position and door proximity using observation records.
- Review door openings with stop count and elapsed time.
- Compare run duration and route legs with other runs for the same containers.
Test container and vehicle separately
If records cluster by vehicle, inspect different containers on that vehicle; if they cluster by container, inspect runs for that container on other vehicles. Where seal, lid or plate records exist, verify them against model and inspection records.
Match the comparison
When comparing vehicles, match regime, duration, stops, load and route conditions. If conditions differ, do not compare raw counts directly; differences may reflect confounders.
Resorting the same table creates an investigation question only. Complete run-ID, measurement-time and matching checks, then review exposure records.
What can be changed?
If records support a hypothesis about parking, loading position or door-open time, track one variable through the authorized process and compare suitable runs. Vehicle or container decisions depend on product specifications and the quality process.
Complete the finding record
Record vehicle ID, period, sample count, run denominator, matching criteria and open uncertainties. Repeat the method in a later period; a continuing pattern alone does not prove a fault.
