Was the Gauge in Cal When the Gage R&R Ran?

You had a clean study. Three appraisers, ten parts, three trials, and a %GRR of 8.4% against tolerance — comfortably inside the AIAG threshold, ndc of six, nothing to argue about. The PPAP package went out. Then the customer SQE came back with one line: "Please provide the calibration record for the gauge on the date the study was performed." You pull the certificate. The gauge was calibrated in March, due in September, and the R&R data sheet is stamped in October. The number was fine. The gauge was overdue the day you collected it.

That is how a passing Gage R&R gets voided — not on the statistics, but on the evidence underneath them.

What the R&R number actually assumes

A Gage R&R quantifies the measurement system’s own variation — repeatability (the same appraiser, same part) and reproducibility (between appraisers) — and expresses it as %GRR against tolerance or against total study variation, with ndc telling you how many distinct categories the system can reliably resolve. Under the AIAG MSA reference manual the working bands are familiar: under 10% is acceptable, 10–30% is conditional depending on the application and the cost, and above 30% is unacceptable. ndc of five or more is the usual expectation.

Every one of those figures rests on a silent assumption: that the instrument was a known, traceable measuring device at the moment the data was collected. %GRR characterizes variation; it says nothing about bias or accuracy, and it cannot tell you whether the gauge was reading true. That assurance comes from calibration and traceability — not from the study. If the gauge was out of calibration on the study date, you have not measured the measurement system. You have measured an unknown.

This is exactly the seam IATF 16949:2016 addresses in two clauses that work as a pair. Clause 7.1.5.1.1 requires measurement systems analysis on the equipment used to accept product. Clause 7.1.5.2 requires that the same monitoring and measuring equipment carry demonstrable metrological traceability — calibrated against standards traceable to national or international references, with calibration status identifiable. PPAP element 8 is where the two meet on paper: the MSA study you submit is only as good as the calibration record for the specific instrument that produced it. An auditor or customer SQE reads both records together and expects them to agree on dates.

The two records that have to line up

The failure is almost never a bad study. It is two true records that quietly disagree.

• The MSA study record says: this study, on this gauge, on this date, produced this %GRR.
• The calibration record says: this exact serial was in calibration from date A, due on date B.

If the study date falls after date B, the study is unsupported no matter how good the number is. And the gauge’s identity has to be the exact serial, not "a 0–150 mm digital caliper." A quality system that tracks calibration by equipment type rather than by unique instrument cannot answer the auditor’s question at all, because it cannot prove which caliper — the calibrated one or the spare in the drawer — was actually in the appraiser’s hand during the trials.

The objection worth taking seriously

"But the gauge passed as-found at its next calibration, so it was clearly fine during the study." This is the honest defense, and it is not worthless — an as-found result within tolerance is real evidence that the instrument had not drifted. But it is retrospective inference, and it does not close the clause. Two problems remain.

First, as-found tells you the state at the next calibration, which may be months after the study; drift, a drop, or a shock in between leaves no trace in that data point. Second, and more practically, the auditor’s finding is written against traceability and calibration status at the time of use, not against your ability to argue backward from a later certificate. IATF 16949 7.1.5.2 wants the status to have been identifiable and valid when the equipment was used — a control, not a reconstruction. "We think it was probably okay" is a discussion you do not want to be having during a PPAP review or a stage-two audit.

The clean answer is preventive: the gauge simply was not overdue when the study ran, and you can show it in one record.

Where equipment management does the work

None of this asks a calibration system to compute your statistics. PrüfDesk does not calculate %GRR, ndc, bias or linearity — those belong to your MSA method and the AIAG manual. What it manages is the equipment, the calibration, and the evidence that make the study defensible:

• Per-serial calibration status and due date. Every micrometer, caliper, bore gauge and torque wrench is registered as a unique instrument with an auto-calculated calibration due date, so "was this exact serial in cal on the study date?" is a lookup, not an investigation.
• Configurable overdue blocking. Set a hard block per tenant and an overdue gauge cannot be checked out for a study in the first place — the disagreement between the two records never gets created. Where a hard block is too blunt, a soft warning still puts the status in front of the technician before they collect data.
• Check-out and allocation audit trail. The possession record ties the specific serial to the appraiser and the dates, so the instrument named on the R&R sheet is provably the one that was in calibration — closing the "which caliper" gap.
• Reusable calibration procedure templates. The same procedure applies across equipment of one type, with per-unit overrides, so as-found/as-left results are captured consistently and are there when the as-found question comes up.
• Audit-ready PDF reports. Certificate, traceability chain, calibration history and status at a point in time export as one package that sits beside the MSA study in PPAP element 8 and reads back cleanly under customer scrutiny.

A short pre-study checklist

Before an appraiser touches the first part, confirm:

• The exact serial to be used is identified — not just the instrument type.
• Its calibration due date is after the planned study completion date.
• The instrument is not overdue or blocked at check-out.
• The certificate and traceability chain are on file and current.
• The check-out record ties this serial to the study dates and appraisers.

Run that list and the auditor’s one-line question has a one-line answer.

Close

A good %GRR is necessary; it is not sufficient. The study is only as defensible as the calibration record standing behind the exact gauge that produced it — and the cheapest place to fix a mismatch is before it exists. If your MSA evidence and your calibration records live in separate systems that don’t share a date, it may be worth seeing how PrüfDesk keeps a gauge’s status pinned to the day a study ran. No number is worth defending if the instrument behind it was overdue.