Interactive quality improvement of a process subject to complete inspection

Abstract

An ongoing production process produces products with quality characteristics following a known probability distribution. Two process states, in-control and out-of-control, are assumed. The process is subject to complete inspection. Whenever the quality characteristic of a product produced exceeds a predetermined action limit, remedial action is taken to restore the process to the in-control state. In addition, the decision maker has a learning opportunity to improve the process by investing resources to identify and eliminate the causes of non-conformance to the target characteristic value. A learning action taken would reduce the probability of shifting from the in-control state to the out-of-control state. A cost model is developed in this paper to determine the optimal number of learning actions to be taken and the optimal action limit. The model gives insight into the tradeoff of cost of quality and cost of prevention.