摘要
This study extends the conventional skip-lot sampling (SkSP) framework by proposing a two-level skip-lot sampling plan (SkSP-2L) based on variables inspection using the process capability index Cpk. While the traditional SkSP-2L, rooted in attributes inspection, has demonstrated efficiency in reducing the average sample number (ASN), it becomes increasingly impractical in modern high-quality manufacturing environments where defect rates are extremely low. In such contexts, attributes-based methods often require prohibitively large sample sizes to ensure statistical validity. To overcome this limitation, the proposed variables-based SkSP-2L integrates the widely adopted Cpk index to facilitate more efficient and informative lot evaluation, aligned with contemporary process performance assessment practices. The study outlines the plan's operating procedure, mathematical formulation, and sensitivity analysis. Comparative evaluations demonstrate the proposed method's superiority over traditional SkSP-2 and single sampling plans in reducing ASN. A real-world case study further validates its practical applicability in precision-focused manufacturing settings with stringent quality requirements.