Glossary · Verification and validation
Calibration drift
In measurement and verification, calibration drift is the change in measurement accuracy or calibration status over time. It erodes the validity of a calibration long before the next due date if conditions are harsh.
- Verification
- Data and interfaces
In one sentence
Calibration drift is the gradual loss of measurement accuracy after calibration, which can undermine test results and safety thresholds.
Example
A force gauge used to check the closing force of power-operated guards reads increasingly low over months of use, so the measured values understate the real forces on the operator's hand.
How it applies
- Verification: Drift can make past test results unreliable. If an instrument is found out of tolerance at recalibration, assess which earlier results it affected.
- Maintenance: Calibration intervals should reflect observed drift, not only the manufacturer's default. Record as-found values so that trends become visible.
- Technical documentation: State calibration intervals and the conditions that shorten them (shock, temperature, heavy use) in the maintenance information.
- AI and retrieval: The same pattern appears elsewhere: AI drift and information drift describe a gradual gap between a model or a document and the reality it describes.
Calibration drift vs. sensor drift
Calibration drift concerns the calibration state of measuring equipment. Sensor drift describes the change in output of a sensor in operation, including sensors that are part of a safety function.