Acceptance sampling plans have been widely utilized in manufacturing industry for inspection raw materials or final produced products. In most of plans, the decision criteria is only made based on current lot information. Unfortunately such plans need large sample size to make a decision criteria on sentencing the lot. Therefore, these sampling models cannot be an appropriate selection to apply in costly or destructive testing. In order to solve the mentioned problem, conditional sampling methods developed in which making the decision rule is based on the sample information taken from neighboring lots. Multiple deferred state (MDS) sampling plan, which belongs to the conditional sampling family, uses current lot information as well as future lots information. One of the usual assumptions to design MDS plan is that inspection is perfect and type I and type II inspection errors not occurred in the inspection process. But in practice, it is far from reality assuming the ideal inspection in each plan. In this paper, we develop MDS plan to the state where the inspection is not perfect. Also a specific case of MDS plan be more focused with some suitable examples to discuss the influence of inspection errors on the performance of it. The obtained results show that precision and performance of developed plan is under misclassification errors influence.

Keywords