Retest workflow after rohs reach testing failure: part 2
The correct response to a failed RoHS REACH result is not a fast retest but a disciplined sequence: quarantine the affected goods, trace the failure to a material or process cause, and rewrite the sampling plan so the next result answers the right question. Buyers who follow this order spend less on laboratory work and reach defensible shipping decisions faster than those who simply test again and hope. This guide lays out that sequence step by step.
Key takeaways
- Containment comes before investigation and investigation comes before any retest, with each step documented as it happens. [2]
- Acceptance logic treats a failure as a trigger for defined follow-up on the affected lot, not as an invitation to test repeatedly until a pass appears. [1]
- Tracing the failed component to a supplier batch or process step determines the true scope of the problem.
- The re-sampling plan must be rewritten to reflect the investigation instead of repeating the plan that missed the issue.
- A factory self-retest without independent sampling describes different goods and does not clear the original lot.
A failed restricted-substance result triggers urgency, and urgency produces the classic mistake: ordering a retest before understanding what failed. This second part of the retest series slows the process down at exactly the right point, covering containment, investigation, and re-sampling discipline. The retest itself, and how to read it, follows in later parts.
Two public bodies of knowledge anchor the workflow. Factory acceptance practice explains why verification happens against agreed criteria before goods change hands, with failed items triggering defined follow-up rather than informal retry. Quality management practice explains how nonconforming product is identified, segregated, and dispositioned through documented steps instead of corridor conversations. Together they describe a retest as the endpoint of an investigation, not as a second lottery ticket.
Readers coming from part 1 already know how to classify failure types. Here the focus is operational: who quarantines what, how the failed material is traced to a supplier batch, how the re-sampling plan is rewritten to reflect what was learned, and what the file must contain before anyone authorizes the next laboratory run.
Contain first, then think
Containment is a physical act before it is a paperwork act. The affected lot gets segregated, labeled, and removed from the shippable pool so that no parallel process can load it while the investigation runs. Quality management practice treats control of nonconforming outputs as a core discipline: identify the suspect goods, prevent their unintended use, and record who authorized each subsequent disposition. Rushing past this step is how failed goods quietly ship while everyone assumes someone else stopped them. [2]
Notification runs alongside segregation. The factory, the forwarder, and your own merchandising team each need the same short message: which lot is held, which test line failed, and that nothing ships from that lot until the hold is formally lifted. Many buyers coordinate this containment through UTS RoHS REACH testing field support so the hold is verified on site rather than assumed from email replies. A hold confirmed by an independent visit carries weight that a message thread does not.
Resist the pressure to retest during the containment window. Until the failed component is traced to a material or batch cause, a new laboratory run cannot distinguish between a genuinely corrected lot and a luckier draw. The money spent on a premature retest usually buys a second confusing document instead of an answer.
Trace the failure to its source
From line item to homogeneous material
Reports state failures at the line-item level, but causes live one level deeper, in the homogeneous material or the process that applied it. A failed solder joint points to a solder alloy or flux batch; a failed coating points to a paint or plating supplier; a failed plastic housing points to a resin or colorant lot. Work with the factory to map the failed line to the exact material input, then pull receiving records for that input across the production window.
From material to production scope
Once the suspect input is named, the scope question answers itself by following that input through production records. Which purchase orders received the same supplier batch, which production dates consumed it, and which finished models contain the affected component. Acceptance practice frames this as verifying against criteria lot by lot rather than assuming uniformity across everything the factory ever made. Test the lots that share the suspect input and leave unrelated lots to their own evidence. [1]
Rewriting the sampling plan with what you learned
| Plan element | First-round weakness | Rewritten version | Why it matters |
|---|---|---|---|
| Lot definition | Whole order treated as uniform | Split by suspect input batch and production date | Retest results map to genuinely uniform goods |
| Stratification | One variant overrepresented | Quotas per variant tied to carton counts | No hiding place for affected units |
| Selection method | Convenience selection near the door | Systematic draw across all stored cartons | Selection represents the full hold area |
| Grouping | Many parts composited per line | Separate lines for suspect components | A pass on clean parts cannot mask dirty ones |
| Documentation | Sampling notes kept separately | Unit IDs shared with lab upfront | Report and field record reconcile exactly |
A containment-to-retest sequence
- Segregate and notify the same daySegregate the affected lot physically, label it held, and notify the factory, forwarder, and internal team with the same written message.
- Trace the material through recordsMap the failed line item to the homogeneous material and pull receiving and production records for that input.
- Set the investigation scopeDecide which lots share the suspect input and hold those as well until their own evidence exists.
- Rewrite the sampling ruleRewrite the sampling rule to fix the weaknesses the failure exposed, and record the new rule before anyone visits.
- Sample fresh and test once readyDraw fresh samples under the new rule with independent selection, then commission the laboratory run against reconciled unit IDs.
Documented control of suspect goods is what separates a professional response from a scramble. Identification, segregation, recorded disposition, and re-verification are the backbone of credible nonconforming-product handling in any managed quality system. [2]
Sources
- Wikipedia: Factory acceptance test —— Overview of factory acceptance testing performed before shipment or installation.
- Wikipedia: ISO 9000 —— Overview of the ISO 9000 family of quality management standards.
Frequently asked questions
A failure just came in. What are the first three moves?
Quarantine first, investigate second, retest third. Physically segregate the affected lot so it cannot ship, confirm exactly which component and material failed, and trace that material to a supplier batch or process step. Only when the re-sampling plan reflects that understanding should anyone pay for another laboratory run.
The factory retested on its own and sent a pass. Does that clear the shipment?
Usually it means the new result describes only the newly sampled goods, not the original lot. Retests use fresh samples drawn under a stated rule, and unless the investigation shows the original lot and the new goods share the same homogeneous materials, the earlier failure still stands against the earlier goods. Keep the two populations separate in your records.
One component failed in one model. Must I assume sibling models are affected?
Treat it as one data point about one stratum, then widen the evidence deliberately. Check whether the failed component appears in other models or production dates, review incoming material records for shared supplier batches, and extend sampling to those areas under the same documented rule. A single failure triggers a bounded investigation, not an automatic condemnation of everything.
Which documents should the final file contain?
The sampling sheet with unit IDs and carton numbers, the laboratory report with method notes, the material trace showing which supplier batch the failed part came from, the corrective note from the factory, and the re-sampling record. Filed together, these show what failed, why it plausibly failed, what changed, and what the new result actually covers.