Optimum design of generalized adaptive sampling plan for solid supplier-buyer purchasing partnership with yield-driven validation

Abstract

A flexible, effective scheme to operate successive lot disposition is critical for developing a solid supplier–buyer relationship. The consecutive lots' yields from in-control, automated manufacturing processes contribute substantial, valuable yield-record information. Feeding this information back into the inspecting compliances and decision criteria improves the sampling plans’ performance and supplier–buyer resources allocation. The existing multiple-lot dependent state plan (MDSP), however, engenders an adverse situation. It situates higher inspection costs and lower lot-sentencing standards when an increase of cumulative-lot yields considers in a lot-sentencing decision. This adversity significantly impacts a long-lasting supplier–buyer partnership. A recently developed adaptive MDSP (AMDSP) has resolved the adversity. But its sentencing rules administer a wide range of marginally satisfactory yields. To mitigate these, in this paper we propose a generalized AMDSP (GAMDSP) validated by the process yield-driven index. The process-yield-validated GAMDSP contains combinatorial mathematical treatment that activates its manufacturing traceability to historical lot-yield levels. When encompassing more preceding lot-yield histories in the decision, the GAMDSPs manifest the benefit of inspecting sample reduction. The GAMDSPs also increase the lot-acceptance yield standards and show lot sentencing with a narrow range of marginally satisfactory yields. Besides, they are sufficiently flexible. By manipulating their adaptive mechanisms, the GAMDSPs can convert into a specific sampling plan that is suitable for the different stages of supply chain management. Finally, a case study demonstrates the applicability of our proposed GAMDSPs.