Quality at the source and segregating nonconforming product

Why it matters

When a nonconforming product is caught at final inspection, it can be prevented from reaching the customer. But up to that moment it may have consumed handling, assembly, machining and inspection resources. Quality at the source aims to link the defect to the operation where it arose and to prevent it from being carried to the next process. Segregating all nonconforming product at final inspection is a valuable protection; still, it does not change the fact that the defect was produced at the source and moved forward inside. In this lesson it is stated clearly across which boundary the escape passed. An escape to the next station and an escape to the customer are not the same indicator. For both, the starting set, the inspection point and the known states must be defined separately.

Segregation is keeping a nonconforming or suspect product physically and in the records away from usable product. Sticking a label may not be enough on its own; if the product can be put back on the line by mistake, the protection is incomplete. The product ID, quantity, location, reason for segregation and decision status are recorded together. A suspect product whose conformity is not yet known is also segregated, separately from a verified nonconforming product. Keeping the suspect scope wide is not a claim that every product is defective. It is a temporary decision that prevents wrong use until the uncertainty is closed. When and on what evidence this scope will be narrowed must be set in advance.

Adding an inspection at the source is not the same as removing the cause of the defect. While an operation continues to produce nonconforming product, segregation can be done right beside it. Such an arrangement protects the next process, but it does not count as having permanently reduced defect formation. To sustain quality at the source, the acceptance condition, the reaction to an abnormality and the verification after correction must be set up together. Calling on the worker to be more careful is not an assurance on its own. A visible status mark is also not evidence that the inspection was really done. Temporary segregation and permanent prevention are tracked as separate goals; one is not closed because the other is complete.

Numerical monitoring shows the success of catching and the limit of protection. But to calculate the escape rate, it must be known how many nonconforming products there really are. Automatically counting products that passed inspection but were not examined as good lowers the ratio artificially. In the example in this lesson, the real status of all products was verified afterward; so the quantities caught at the source and those that escaped can be reconciled. In a real application such complete information may not exist. Then the known numbers, the unknowns and the audit scope are written separately. An escape of zero observed, especially under limited inspection, does not mean the probability of escape is zero. The decision must be made together with this limitation.

Formula