Optimal design of a data-driven inspection scheme with a decision support system for improving lot acceptance determination and supplier compliance
摘要
Acceptance sampling systems have attracted increasing attention in recent years because they allow decision rules to be dynamically adjusted based on historical inspection results, thereby enhancing the efficiency of sampling inspection. Among acceptance sampling systems, the two-plan sampling system (TSS) exhibits greater flexibility, allowing for specifying the number of previous lot quality records to be considered. The term “two-plan” denotes the existence of two decision rules: tightened and normal inspections. However, existing TSS methods are limited in that they can only adjust one parameter within either the tightened or normal inspection regime, i.e., the sample size or the acceptance criterion. In this paper, we propose a novel method that enables simultaneous adjustment of both the sample size and acceptance criterion in both the tightened and normal inspection stages of TSS. This method, referred to as the generalized two-plan sampling system (GTSS), is designed to incorporate both existing TSS frameworks. Comparative analysis with existing methods demonstrates that the proposed GTSS further improves the efficiency of sampling inspection, including enhanced cost-effectiveness and increased discriminatory power in assessing lot quality. Moreover, we developed a cloud-based decision support system for the proposed method to assist practitioners in implementing it more effectively in industrial applications. Finally, we illustrate the practical applicability of the proposed method through a real-world case study.