Food safetyTemperature monitoring
Temperature sensor data gaps: record what is unknown
Missing sensor data should be shown as a gap, not a normal reading or a zero. Record the affected device, time window, investigation and any alternative evidence so the responsible person can assess the uncertainty under the site procedure.
At a glance
Missing sensor data should be shown as a gap, not a normal reading or a zero. Record the affected device, time window, investigation and any alternative evidence so the responsible person can assess the uncertainty under the site procedure.
Define the decision behind the record
A disconnected device and a device measuring an unexpected temperature are different problems. Avoid charts that join distant points without making the missing period visible. Store the last received reading separately from any assumption about the conditions during the gap.
Define the missing period before assessing its implications
Identify the sensor, measurement location, expected recording pattern and last available observation. Keep connectivity information separate from temperature information: a network outage explains a transmission problem only to the extent supported by evidence. It does not automatically establish what temperature the food or equipment experienced during that period.
Record the first missing expected point, restoration information and any separately obtained observations with their actual method and time. If buffered readings arrive later, preserve their event timestamps and distinguish them from receipt time. Do not draw invented values through the gap or treat a manual observation as though the sensor produced it.
Review which questions the recovered evidence answers and what remains unknown. Record the responsible assessment and any follow-up affecting equipment, data collection or stock. A useful test contains both a transmission delay with recovered observations and a genuine absence of observations. The display and export should make those cases distinguishable rather than showing the same continuous line.
Worked example
A reviewer notices no data arrived during a service period. The team records the last received time, checks device and connection status, and gathers permitted alternative observations. They document what those observations establish and what remains unknown.
Build a usable record
Use the fields below to capture the facts needed for this particular decision. Keep the actual observation separate from any later correction or approval.
- Device and monitored location
- Last received and next received timestamps
- Gap duration and investigation
- Alternative evidence and its limitations
- Decision, action owner and review
Example record and review decision
Fictional record example: SENSOR-03: last available reading at 09:00; GAP-01: no observations until 10:20; MANUAL-01: separately identified manual evidence.
Separate the period with no sensor data from evidence collected by another method. Identify what remains unknown and who assesses it.
Do not draw a continuous temperature line across the gap or substitute a later normal value for the missing observations.
Test the review, not just the form
Simulate a disconnected device in a test environment. Check the display, notification and exported record. A manager should be able to identify the gap even when later readings are normal. Confirm that manual additions remain identifiable.
A few useful answers
Should missing readings be filled automatically?
Do not present estimated or substituted values as actual measurements. Any alternative evidence must be labeled and reviewed.
Does a restored connection mean no data was lost?
Not necessarily. Check whether actual observations were buffered and recovered, or whether the period remains unobserved. Connection status, recovered timestamps and measurement evidence are separate parts of the investigation.
Sources and further reading
MarkdownExplore starter templates
Start with a useful checklist
English
