Design of One-Sided Screening Specifications for Multi-characteristic Product

Screening procedure is one of the effective approaches for quality assurance. Most researches study in the situation of single characteristic. However, many products have more than one quality characteristics. This research examined a multi-characteristic screening procedure using multiple screening variables. An extended OMS procedure using a heuristic algorithm is employed to reach the required average outgoing quality (AOQ) or conforming rate. Then two optimal models, namely the individual cut and linear cut approaches, are proposed to determine the optimal cutoff points of screening variables by maximizing the selected rate under a prespecified AOQ or conforming rate. An example of a complete inspection process is presented to illustrate the use of the proposed models