Developing a cost-efficient dual sampling system for lot disposition by considering process yield and quality loss

Abstract

In the past few decades, various sampling strategies have been developed for different perspectives or occasions to benefit from cost reduction or time-saving when one performs lot disposition, particularly for those inspected items or expensive procedures. The quick switching sampling (QSS) system has been important in examining fewer samples under the same quality requirements and was designed with a dually flexible mechanism for varied qualities. However, most related studies considered only a fixed switching policy either critical value or sample-sized tightenings, which might neglect the collective effect of two types of QSS systems. In this study, we integrate two variables QSS (VQSS) systems in a generalized form, known as integrated VQSS (IVQSS), based on the Cpmk index, a non-scale process performance indicator considering centering, variation, and quality loss simultaneously to accommodate advantages of two individual systems. The operating characteristic function is derived using the Markov Chain theory. Furthermore, we developed a mathematical model to minimize the average sample number and limit two acceptable sampling risks under desirable submitted quality levels. The plan parameters are obtained for making a reliable judgment on submissions by solving this optimization problem. Finally, to illustrate the practicability of the proposed method, we demonstrated a numerical application obtained from a solar panel industry and provided the solved plan parameters under commonly used conditions for convenient use.