Follow one defective machined component from detection and containment through purchasing, equipment capability, corrective action and management system learning.

A defect becomes a system issue when its causes, affected outputs and preventive controls cross departmental boundaries. The practical challenge is to select a response supported by evidence across the operating interfaces: Quarantine suspect stock and establish the lot boundary.
This illustrative case is assessed through seven operational junctions, each identifying a real decision, a plausible failure route and records that could support management assurance. The recommended methods remain subject to the organization’s own risks, laws and contracts.
Illustrative scenario. A receiving inspector discovers that a machined coupling bore is outside the approved tolerance after several lots have already moved between production cells.
Scope and limitations. The example is a fictional manufacturing incident used to explain a repeatable traceability method; no actual product recall, customer complaint or measured loss is claimed.
Standards and applicability. Traceability is a relationship between product identity, process history, inspection and disposition. It is not an instruction to create a single massive spreadsheet. Where a machine, contractor or sensor system already retains reliable history, the improvement task is to make that history retrievable and controlled without losing the underlying timestamps.
The quality management requirements supply the governance frame, while quality plans help organize a product- or project specific response. The guide does not prescribe an eight discipline form or a fixed sampling frequency; the organization must justify its chosen controls against defect criticality and customer obligations. Consult ISO 9001:2026, quality management and ISO 10005:2018, guidance on quality plans for the official publication scope.
01Establish the defect and affected population
A bore measurement fails the drawing limit on an outgoing coupling. The quality technician records the actual reading, instrument identification, drawing revision, lot identifier and time of detection rather than simply marking a box “failed”.
The first decision is not whether the component can be reworked; it is whether other units made under the same settings might also be affected. A clean sample from the latest shift does not prove the earlier population acceptable.
Place identified lots on controlled hold, prevent inadvertent release, preserve a reference part and document the technical basis for any expanded or narrowed scope.
Evidence to examine. Hold transaction, lot genealogy, timestamped inspection record and approved containment decision. Compare source records with the implemented control, noting any disagreement that could change the decision.
02Connect the failure to process conditions
Production reports a tool change during the preceding shift. Review setup sheets, CNC revision, tool wear offsets, first off approval and maintenance activity in sequence, including what actually happened rather than what the work instruction says should happen.
A machine setting may look correct now because operators adjusted it after discovery. Historical electronic records, where available, are stronger evidence than retrospective statements made to defend the process.
Use a timeline to align machine state changes with the first known nonconforming product, and test whether the same condition exists on other machines or shifts.
Evidence to examine. Machine event trail, first piece approvals, setup revision history, maintenance record and comparison samples. Compare source records with the implemented control, noting any disagreement that could change the decision.
Decision point. Decide who has authority to accept the immediate risk and which objective evidence would justify restoring normal operations. Make any time limited approval visible to affected teams.
03Test measurement system confidence
Before attributing the fault to machining, inspect the gauging method: Calibration status, fixture condition, temperature assumptions, measurement uncertainty and whether different inspectors obtained consistent results.
Incorrect measurement can create both false rejection and false release. A calibration certificate by itself does not prove that the technique used at the station was fit for this product feature.
Repeat measurements using an authorized method and independent competent personnel; investigate any disagreement without overwriting original readings.
Evidence to examine. Gage identification, calibration trace, method instructions, repeatability evidence and documented disposition of conflicting results. Compare source records with the implemented control, noting any disagreement that could change the decision.
04Examine purchasing and outsourced processing
Trace the raw material batch and any contracted finishing operation. Confirm that purchase requirements, concession decisions, incoming verification and provider communications preserve the same part revision and critical dimensions.
External provision is not automatically the root cause merely because it appears earlier in the chain; responsibility for defining and verifying requirements remains with the manufacturer.
Use a supplier evidence request tied to the suspect characteristics, not a generic questionnaire or demand for unrelated certificates.
Evidence to examine. Approved supplier scope, order specifications, incoming inspection results, delivery records and relevant sub tier traceability. Compare source records with the implemented control, noting any disagreement that could change the decision.
Evidence check. The decisive verification at this junction concerns examine purchasing and outsourced processing. Check whether the approved response can be reconstructed using approved supplier scope, order specifications, incoming inspection results, delivery records and relevant sub tier traceability.

05Decide and authorize product disposition
Engineering evaluates technical suitability; quality controls nonconforming outputs; commercial teams manage customer communication where needed. Rework, concession, scrap and acceptance are different decisions and need different authorities.
Repairing one component is a correction. Eliminating a repeated setup escape requires investigation of why the control failed and how the organization will know it remains effective.
Document disposition criteria, decision rights, any customer/regulatory approval and a release gate before stock returns to normal flow.
Evidence to examine. Signed disposition, rework instructions where applicable, final verification record and trace to released serial or lot numbers. Compare source records with the implemented control, noting any disagreement that could change the decision.
06Find a cause that predicts recurrence
A cause statement such as “operator error” is inadequate until the team tests training, shift handover, software controls, setup verification and conflicting production incentives. Consider why the inspection barrier failed as well as why the machining drift occurred.
A preventive action is strongest when it changes a system condition: Validated setup parameters, a reliable first off hold, clearer escalation or robust in process sensing.
Select actions proportionate to escape risk and check that new steps do not create an uncontrolled bottleneck or informal workaround.
Evidence to examine. Cause analysis, challenged hypotheses, action owner, due date, revised control and verification plan. Compare source records with the implemented control, noting any disagreement that could change the decision.
Audit challenge. Sample at least one ordinary operating period and one high pressure or exceptional condition. Compare what staff do with the controlled process and investigate discrepancies without rewriting history.
07Verify effectiveness and feed management review
Use subsequent lots and shifts to challenge the new setup process; look for the specific failure mechanism and a sustained reduction in escapes rather than a one day compliance demonstration.
Management should see defect exposure, customer impact, recurrence trend and resource decisions in one storyline, while commercial confidentiality is protected.
Close the action only when the changed control has been observed in use and the agreed effectiveness period has produced credible evidence.
Evidence to examine. Post change sampling, process audit results, complaint trend, effectiveness sign off and management review input. Compare source records with the implemented control, noting any disagreement that could change the decision.
Implementation priorities. Quarantine suspect stock and establish the lot boundary; Reconstruct machine, operator, measuring device and supplier histories; Decide product disposition under approved authority; Verify the correction at the production process, not just in the record. These are practical checkpoints for this example; local procedures must define the actual sequence and decision owners.
Closing Perspective
A defect becomes a system issue when its causes, affected outputs and preventive controls cross departmental boundaries. In this case, closure depends on the ability to decide product disposition under approved authority while retaining the earlier decision history.
For this particular case, the next review should challenge verify the correction at the production process, not just in the record. The final operating stage provides a suitable evidence trail: Post change sampling, process audit results, complaint trend, effectiveness sign off and management review input.
From ISO Toolkits · implementation resources
Explore editable risk, audit, control and evidence templates that may be tailored to this sector and its management system scope.
Browse verified ISO Toolkit Library · Review sample templates
References & Further Reading
- ISO 9001:2026: Quality management
- ISO 10005:2018: Guidance on quality plans
- ISO 19011:2026: Management system audit guidance
- ISO 10015:2019: Competence management guidance
Source status reviewed 19 October 2026. The discussion describes management practices, not verbatim normative clauses.