QM BenchQuality engineering tools

Quality inspection tools

Inspection decides which parts go to the customer. It only works when the sample is right, the gauge can see the difference and the inspectors agree. These tools cover sampling, measurement systems and the plan that ties them together.

By Ľubomír Blažek · Process quality team leader, automotive · updated 2026-11-14

Inspection planning

What is checked, with which gauge, how often, and how the first part is approved.

Sampling

How many parts to check and when to accept the lot.

Measurement systems

Make sure the gauge and the inspector can see what you want to control.

Dimensional and process results

Evaluate measured results against the drawing and the process.

Approval and follow-up

Part approval and the handling of nonconforming product.

Learn and practise

Inspection is a process too

Every inspection has its own errors: a sample can miss a bad lot, a gauge can scatter more than the tolerance, and two inspectors can judge the same part differently. Studies of visual inspection regularly find that even careful 100 % inspection misses a noticeable share of defects. That is why inspection needs the same discipline as production: a defined method, a capable measurement system and checked agreement between people.

The inspection chain

  • Plan: the control plan names the characteristic, gauge, sample size, frequency and reaction plan
  • Sample: AQL or a statistically sized sample decides how many parts to check
  • Measure: a gauge with a proven type 1 study and Gage R&R, or inspectors with proven agreement
  • Decide: accept, reject or sort, against the plan and the drawing
  • Learn: feed results back: control charts, capability, Pareto of defects, corrective actions

Inspection or SPC?

Inspection sorts good from bad after the fact; statistical process control keeps the process from making bad parts. For stable, capable processes the inspection effort can be reduced, while special characteristics and unstable processes need more. The capability of the process and the risk of the characteristic decide the mix.

Coming to this section

  • · Import of characteristics from the control plan
  • · Defect catalogue with limit samples
  • · Inspection effectiveness tracker

Missing a tool you need? Write to info@qmbench.com.

Frequently asked questions

What is AQL?

The acceptance quality limit: the worst process average that is still considered acceptable for a sampling scheme. It is not a promise about a single lot. Lots at the AQL are accepted most of the time; worse lots are rejected more often.

How many parts do I need to inspect?

For lot acceptance, ISO 2859-1 gives the sample size from the lot size and inspection level, and the accept and reject numbers from the AQL. For a statistical study, the sample size depends on the confidence and the difference you need to detect.

What is the difference between a type 1 study and a Gage R&R?

A type 1 study checks the gauge alone: bias and repeatability on one reference part. A Gage R&R checks the whole measurement system on real parts with several operators. The type 1 study comes first.

How do I check visual inspectors?

With an attribute agreement analysis: several inspectors judge the same set of good, bad and borderline parts several times, and kappa measures their agreement with each other and with the standard.

Does 100 % inspection guarantee zero defects?

No. People and machines both miss defects, especially rare ones and on long shifts. Error prevention in the process and capable processes are more reliable than more inspection.

Other areas:LeanSix SigmaShop floor managementSupplier qualityAPQP