A paper cup quality inspection sampling plan should explain which lot is being evaluated, how samples are selected, what is checked and what happens after a failure. A list of tests without decision rules is incomplete. Equally, several passing cups do not demonstrate that every piece in a shipment conforms.
Separate lot acceptance from process monitoring
The NIST engineering statistics handbook explains acceptance sampling as a decision about whether to accept or reject a lot based on a sample. It distinguishes that purpose from estimating the lot's quality. Read the NIST explanation.
Use that distinction when designing the factory's records. Incoming-material acceptance, startup verification, in-process monitoring and final product release can have different boundaries and reactions. Do not use one convenient sample size as an unexplained rule for all four tasks.
Define the lot and approved methods
Identify the product, material association, production boundary and quantity included in the decision. Specify the applicable drawings, customer requirements and method revisions. Explain how samples represent the defined lot rather than only the easiest cups to reach.
Use appropriate selection methods under the approved plan. Identify sample locations and timing, and preserve each sample's link to the lot. Avoid preferentially selecting attractive cups, or replacing a failed sample with a passing one without recording the original result.
| Plan element | Required definition | Failure to avoid |
|---|---|---|
| Lot boundary | Product, quantity and material/production identity | Unknown decision scope |
| Characteristic | Requirement and defect definition | Inconsistent pass/fail interpretation |
| Method | Procedure, conditions and equipment | Results that cannot be compared |
| Selection | Approved sampling method and locations | Convenience sampling presented as representative |
| Decision rule | Applicable sample size and acceptance/rejection basis | Limits invented after testing |
| Reaction | Hold, investigation and authorized disposition | Failed evidence without action |
| Record | Sample identity, result, owner and revision | An untraceable release signature |
Sampling sizes, severity classifications and acceptance rules must come from the agreed quality requirements and applicable plan. No universal AQL, inspection level or rejection count is prescribed here.
Match the method to the characteristic
Visual defects, dimensions, leakage and other characteristics require appropriate methods. Keep their conditions and counting rules distinct. A destructive test can consume the selected samples, so record their original identity before testing and define how their quantity is treated.
The dimensional measurement guide and leak-test validation guide support method-specific review. Those references do not replace the buyer's accepted criteria for the particular cup application.
Check the capability of the measurement or inspection method before interpreting a close pass/fail result. If inspectors disagree, resolve the method and reference issue rather than averaging incompatible judgments into an apparent acceptance.
Build the reaction into the plan
State who holds affected output and who decides its disposition. Use the defect classification and containment guide to preserve the scope and evidence of a failure. Reinspection or sorting needs an authorized method and record; it should not erase the initial result.
Review the plan after material, tooling, process or customer requirements change. The production change control guide connects such revisions to a verified baseline. Retain historical plan versions so previous release decisions remain interpretable.
Keep a failed sample in the decision record
Suppose a selected cup fails the defined leakage check. Testing a second attractive cup and replacing the first result does not preserve the original sampling decision. Retain the failed unit’s identity, test conditions and result, then apply the approved reaction and any authorized follow-up plan.
If the test itself was invalid, record the specific invalid condition and the pre-established rule that permits a repeat. A spill or lost identity may require a different disposition from a genuine product failure. Do not invent an invalid-test reason after seeing an unfavorable result, or combine additional samples with the original count without an applicable decision rule.
Separate corrective action from retrospective acceptance
A machine adjustment can justify a new identified production verification, but it does not change the evidence for cups already made. Define the affected production boundary and obtain an authorized decision for held stock. Keep any sorting, reinspection and final release linked to the original finding.
When planning an equipment trial, agree which results concern the demonstrated machine application and which release a production lot. The same sample record can inform both discussions, but neither decision should silently substitute for the other.
Frequently asked questions
Can one fixed sample quantity be used for every lot?
Only if the applicable approved plan supports that choice. Lot boundaries, characteristics and decision requirements need explicit review.
Does a passing sample prove zero defects in the batch?
No. Report the decision made under the sampling plan and its scope. Do not describe sampled acceptance as inspection of every cup.
Can only accessible top-of-carton cups be sampled?
Use the approved selection method so the sample represents the defined decision boundary. Convenience alone is not justification.
What happens when an inspection fails?
Apply the documented reaction, preserve identity and result, and obtain an authorized disposition before product release.
Discuss your cup application and inspection interfaces with HANNAI when planning production verification.