Six Sigma inspired sampling plan design to minimise sample size for inspection

摘要

One of the important property of the c = 0 plan is that it provides a perception that defective product will not be tolerated, ideally a zero defect environment requirement. But this plan fails in discriminating the lots with non-conformities even with larger sample size. In this paper, a method based on Design for Six Sigma (DFSS) road map Plan, Identify, Design, Optimise and Validate (PIDOV) is proposed for problem solving. This procedure retains the simplicity of single sampling plan by attributes. The procedure uses a recursive form of hypergeometric probability formula for sampling plan design to meet the stipulated consumer's risk and producer's risk. New levels of AQL and LTPD are chosen and corresponding changes in the given specifications are made yielding to smaller sample size. The new tightened specification allows the probability of occurrence of the defective in a smaller size of the sample. The parameter Average Total Inspected (ATI) calculated shows considerable improvements from the existing sampling plans. Thus the proposed procedure shifts the application of acceptance sampling plan to acceptance control plan to ensure minimum deviations from the desired specification mean, an obvious objective of a Six Sigma program.