Abstract
The repetitive sampling plan (RSP) is a statistical method utilized to assess the quality of a lot by inspecting a certain number of samples and resampling them when necessary. RSP's flexibility helps maintain positive relationships with business partners. Compared to the traditional single sampling plan (SSP), RSP reduces inspection costs by examining fewer samples while providing equivalent protection to both suppliers and customers. However, the conventional RSP applies identical criteria for lot sentencing during resampling, resulting in potential inefficiencies. To address this issue, this study proposes an adaptive mechanism that modifies RSP during resampling to narrow down the critical region. An optimization model is created to minimize the average sample number (ASN) while ensuring the required quality and tolerable risk levels are maintained. Results show that the proposed method, compared with SSP and conventional RSP, leads to a substantial reduction in ASN and improved discriminatory power. Additionally, the study includes tables of solved plan parameters for commonly used quality and risk requirements to facilitate practical applications.