Verify a PCB Supplier Corrective Action with a Defined Effectiveness Check

A supplier can complete an action list without demonstrating that the problem has been resolved. Training records, revised instructions, and a new fixture describe changes made. An effectiveness check must show whether those changes address the original failure under relevant production conditions.

Bosch’s problem-solving booklet distinguishes implementing corrective actions from verifying their effectiveness. Use that distinction when reviewing a supplier response, while applying the actual issue-specific requirements and evidence agreed for the build.

Connect the action to the demonstrated cause

State the observed failure, the proposed cause, and the evidence linking them. Then explain how the action changes that causal path. An instruction to “take more care” is difficult to verify because it does not identify a controlled process condition or a measurable outcome.

Separate occurrence prevention from escape detection. A screening step may keep defective boards from shipment while the process continues to create the defect. That can be a useful temporary control, but it answers a different question from removing the cause.

Define success before reviewing results

Record whether verification used ordinary production or a specially supervised trial. A carefully selected trial can demonstrate a mechanism, but it may not show that the action remains effective across shifts, material variation, and routine setup changes.

Review evidence rather than a completion date

1. Confirm the action was implemented where intended.
2. Check the verification population and method.
3. Compare results with the predefined success criteria.
4. Investigate remaining events and unintended effects.
5. Document the reason for closure or further work.

Absence of observed recurrence needs context. A short run with little opportunity for the failure to occur gives limited evidence, even if its defect count is zero. Report the relevant production exposure and any restrictions instead of presenting a calendar interval alone.

Keep the original issue definition stable. If the supplier changes the defect code or measurement method during the action, reconcile the old and new definitions before interpreting an apparent improvement. Otherwise the trend may reflect recording changes rather than product changes.

The closure record should show what changed, why it should work, and what observations support that expectation in practice. This makes the action transferable to related products without assuming that one successful trial proves universal effectiveness.