01
Lot definition
Define which units form the submitted lot and ensure the sample can be selected from the lot without supplier cherry-picking.
Evidence route
Lot identity, quantity, production period and sampling access
Acceptance sampling context
AQL inspection in China, explained without a quality guarantee.
AQL is used within attribute acceptance-sampling systems to design a lot-disposition plan around defined risks. It does not mean that a buyer agrees to receive a stated defect percentage, and it does not prove that every unit in an accepted lot conforms.
Decision premise
Sampling question: what lot, defect classes, plan, inspection level and acceptance criteria govern the disposition?
Control framework
Six inputs that make an AQL plan interpretable
AQL is only one input. Lot formation, sample size, inspection level, defect classification, acceptance numbers and switching rules determine what the plan actually does.
02
Defect classification
Agree critical, major and minor definitions against product use and customer criteria before inspectors classify observations.
Evidence route
Defect catalogue, visual limits and classification authority
03
Plan and inspection level
Name the applicable standard or customer plan, inspection level, normal/tightened/reduced status and single/double/multiple plan where relevant.
Evidence route
Plan reference, code letter and switching status
04
Sample and selection
Record required sample size and how units were selected across cartons, cavities, machines, shifts or production sequence when relevant.
Evidence route
Selection record, sample size and sample identifiers
05
Acceptance criteria
State acceptance and rejection numbers by defect class, plus any zero-tolerance critical condition defined by the customer.
Evidence route
Ac/Re criteria, observed counts and disposition result
06
Reaction and escalation
Define what happens after non-acceptance: containment, sorting, rework, resubmission, corrective action or customer decision.
Evidence route
Hold status, NCR, resubmission basis and authorised release
Decision evidence
Terms that should not be collapsed into ‘AQL passed’
The same AQL value can lead to different sample sizes and decision risks under different plans. The full plan context belongs in the report.
| Term | What it controls | What it does not mean |
|---|---|---|
| AQL | A design baseline used within an acceptance-sampling plan | Guaranteed outgoing defect percentage |
| Inspection level | Relationship between lot size and sample-size code | Product criticality by itself |
| Ac / Re | Observed-defect decision boundary for the defined sample | Permission to ship known nonconforming units |
| Accepted lot | The sample result met the plan's lot-disposition rule | Every unit was inspected or conforms |
| 100% inspection | Every in-scope unit or feature is checked as specified | Automatic elimination of measurement or human error |
Scope limits
When sampling needs stronger controls
Safety- or function-critical characteristics may require different or full verification.
Small or mixed lots can make a generic plan inappropriate.
Known process instability should not be managed only by final sampling.
The customer specification or purchase contract always governs the plan.
Standards context
Use external standards to define the decision—not to decorate a claim.
Independent technical context for sampling-plan choices and operating-characteristic trade-offs.
NIST sampling-plan guidance