Quality control and product servicing: A decision framework (Q750071)

This paper considers the joint design of inspection and service capacity by recognizing the cost interdependence of quality control and servicing. It provides a model based on an economic framework that considers the trade-offs between costs of quality control and servicing under some simplifying assumptions. In doing so, an acceptance sampling scheme is used to represent the quality control efforts, and the level of service provision is determined by the servicing capacity provided by the firm. The cost of provision is defined to be the fixed and variable costs involved in providing repair and servicing for defective products that have been returned by the unsatisfied customers. For the sake of simplicity, the paper ignores the product reliability and dynamic aspects of product servicing and considers only the long run costs of repairing a defective product (returned by a customer) as the cost of servicing. An acceptance sampling plan is used for purposes of quality control. Defective items that are identified in the sample are assumed to be reworked and corrected and the probability of product failures after rework is assumed to be negligible. It is assumed that rejection of a lot is followed by 100\% inspection to correct any defects. Concerning acceptance sampling two types of costs are considered: the cost of quality control and cost of rework of items after inspection. Two cases are considered for modeling uncertainty concerning the proportion of defective items and in both cases the problem of how quality control and servicing capacity may be jointly designed is analyzed. In one case, uncertainty regarding proportion of defectives is modeled using a non-degenerate probability distribution. In the other case, a degenerate probability distribution is used and the results from the two cases are compared. It is shown that for the case of degenerate probability distribution, the quality control scheme reduces to inspection screening, that is, the inspection of items serves as a screening effort for defective items. For some values of the proportion defective, it is shown that partial inspection is more desirable than the 0 or 100\% inspection. For non-degenerate probability distributions, sufficient conditions are derived under which the optimal quality control policy and optimum servicing capacity can be obtained.