Utilizing the information theory of entropy to solve an off-line inspection problem

This study presents an off-line inspection problem for a batch produced from a process subject to random failures and exhibiting manufacturing variations. The objective of this paper is to develop an inspection policy in which units should be inspected in a particular order to find the transition unit in the batch under a required confidence level. This study develops an algorithm to compute the expected number of inspections. This approach uses the information theory of entropy to select an un-inspected unit to be inspected, and effectively minimizes the uncertainty of the transition unit in the production batch. A numerical example illustrates the proposed off-line inspection policy, and the effects of model parameters on the expected inspection number are investigated. The numerical example in this study indicates that full inspection is required when the required confidence level is one or the process has larger manufacturing variations.