Can you trust a factory's measurements?
Trust factory measurements only after checking the instrument suits the tolerance, its calibration evidence matches the equipment used, and the method is agreed. Compare readings on the same identified parts before production approval. Ask Cambridge China Bridge to coordinate these checks with its own staff in China.
Define what is being measured
Start with the critical feature and its acceptance limits. The product specification guide covers what to specify; here, add how each feature will be measured. Identify the drawing revision, units, reference surfaces and whether the result describes an individual reading, an average or a maximum.
Agree how measurement uncertainty will affect acceptance near a limit. Uncertainty describes the doubt around a result, rather than extra tolerance for the supplier. Ask your technical owner to approve a decision rule, including when a result needs further checking instead of an immediate pass.
Check the instrument suits the job
Ask for the instrument type, model, equipment identifier, working range and accuracy information. Display resolution is the smallest change shown, not proof of accuracy. Check whether the instrument and measurement method can distinguish acceptable parts from unacceptable ones at your agreed tolerance.
Consider access, contact force and part behaviour. Calliper jaws can compress a soft part; a fixture can distort a thin component. Ask the operator to demonstrate the actual feature on your part. The factory capability guide covers checking whether the relevant equipment and process are really available.
Read the calibration evidence
Request the full calibration certificate, not just a sticker or a photograph of its cover. Match its equipment identifier to the instrument used. Review the calibration date, tested range, measured errors, uncertainty and reference standards. Ask whether any stated corrections are applied to production readings.
Check who performed the calibration and, where accreditation is claimed, verify that its scope covers the relevant calibration. Ask how the factory sets its recalibration schedule and records damage, repairs and intermediate checks. If calibration found the instrument outside its accepted limits, ask which earlier inspection results need reviewing.
Watch the method and keep the raw readings
Write a short measurement instruction with photographs: where to measure, how to support the part, how to prepare surfaces, how to set the instrument and what environmental conditions matter. Include conditioning, contact pressure or dwell time where relevant. Watch the operator follow it rather than accepting a posed photograph.
Measurement uncertainty can include random variation and systematic bias. Ask the operator to repeat measurements after removing and repositioning the part, then have another operator follow the same instruction. Keep individual readings, part identifiers, operator names and instrument identifiers. Repeating a biased method can produce consistent but misleading results.
Compare buyer and supplier results before release
Send the same marked, stable parts between buyer and supplier, protecting them against damage or changes in condition. Have each side record readings independently using the agreed method. Compare results part by part, including parts close to acceptance limits, and assess differences against the agreed uncertainty and decision rule.
Investigate disagreement before approving production or releasing affected goods. Check reference surfaces, units, fixtures, conditioning, corrections and instrument suitability. If the difference remains unresolved, agree an independent laboratory and method. Keep this measurement check alongside the batch plan in the quality inspection guide; agreement on selected parts does not establish that the whole batch passes.
Frequently asked questions
Is a calibration sticker enough?
No. Ask for the full certificate and match its equipment identifier to the instrument used. Check the tested range, errors, uncertainty and any corrections needed.
Does a digital calliper prove accuracy?
No. Display resolution does not establish accuracy. Check suitability for the feature and tolerance, calibration evidence, contact force and the measurement method.
What if my measurements differ from the factory's?
Compare the same identified parts using an agreed method. Check units, reference surfaces, conditions and corrections. Hold the affected acceptance decision until the difference is explained.
Should I accept a reading just inside the tolerance?
Apply the decision rule agreed with your technical owner. Consider measurement uncertainty and arrange further checking when the result cannot support a clear acceptance decision.