Adjustable Threshold Design for Efficient Lot Assessment under Variables Inspection

Authors:

Abstract
In today’s high-tech manufacturing industries, where quality levels are consistently high and inspection costs remain a concern, reducing the number of samples required for lot acceptance is both practical and necessary. While the repetitive group sampling plan (RGSP) is known for its efficiency, a major drawback lies in its reliance on fixed decision thresholds, which can result in prolonged or even theoretically infinite resampling, especially when the lot quality falls within a moderate range. This study introduces a modified RGSP under variables inspection, designed with an adaptive mechanism that adjusts the rejection threshold dynamically as sampling progresses. By narrowing the threshold region, the plan effectively avoids the issue of endless resampling and accelerates decision-making. In addition, process capability indices are incorporated into the plan’s structure to better reflect the needs of high-quality environments, where defect-based criteria may no longer be practical. The proposed approach maintains the flexibility of RGSP while improving its efficiency and relevance to modern industrial applications. Computation results demonstrate that the method reduces sampling effort without compromising decision reliability, making it a promising option for industries aiming to balance quality control with operational inspection cost.

This publication has 32 references indexed in Scilit: