Quality control audit: testing whether inspection, test and release are doing what the quality plan says they do

A quality control audit is narrower than a quality management system audit and it is aimed at the sharp end: the checks performed on materials, work in progress and finished output, and the decisions made on the results. It asks what is inspected, against which specification, using what equipment, by someone with what training, recorded where, and what happens when the result is out of specification. That last question is the one that finds things, because reject and concession handling is where a quality control function either holds the line or quietly stops holding it under delivery pressure.

The specification is the first thing to test, not the inspection

An inspection can only be as good as the specification it checks against, so start there. Is there a current, approved specification for each characteristic being checked, does it match the customer requirement or the drawing revision in force, and is the version on the shop floor the version in the system. Inspecting diligently against a superseded drawing is a control that produces confidence without producing conformity, and it is common wherever revisions arrive by email rather than through document control.

Measurement equipment and the results it produced

Test the calibration status of the equipment used for the checks you sampled, on the dates those checks were made rather than today. Then ask the question people forget: when a gauge was found out of tolerance at its last calibration, what was done about the product measured with it since the previous one. A documented impact assessment is the answer. Silence is a finding, and it is one that reaches back through everything that gauge touched.

Nonconforming output, rejects and concessions

Follow a reject from identification to final disposition. Was it identified and segregated so it could not be used by accident, was the disposition decided by someone authorised to decide it, and if it was accepted under concession was the customer informed where the contract requires it. Then count concessions over the period. A rising concession rate with no corresponding corrective action is a quality control function documenting its own erosion, and it is visible from the log without interviewing anyone.

Who inspects, and whether they are allowed to say no

Competence records for inspectors, including any specific qualification the process needs. Then the structural question: does the person performing the check report to someone who is measured on output. Where they do, the audit should look hard at the pattern of results near shift end and near the end of the month. Quality control that is organisationally unable to stop production does not fail visibly, it fails gradually, and the records show it before anybody admits it.

Questions people ask about quality control audit

Is quality control the same as quality assurance?

Control is the checking of output against a specification. Assurance is the system that gives confidence output will conform, including the control activities themselves. A quality control audit is narrower and tests the checks.

How should inspection records be sampled?

Across the whole period, and deliberately including the busiest weeks. Sampling a quiet period tests the process when nobody is under pressure, which is not the condition the control exists for.

What about statistical sampling plans?

If a plan is used, the audit tests whether it is applied as written, including the switching rules people ignore, and whether the plan matches the risk. A plan chosen once and never revisited after a product change is a common finding.

Should the audit cover the customer complaints route?

Yes, at least as a cross-check. Complaints about characteristics that inspection is supposed to catch tell you the control is not effective, regardless of how tidy the inspection records look.

Sources

Related answers

Start Capanix ProKeep the findings, not the binder