Optimal and heuristic solutions to the variable inspection policy problem

Abstract

This paper presents optimal and heuristic methods for finding the variable inspection policy that minimizes expected total cost per unit. These methods utilize a recursive function to compute the minimum remaining cost over all policies that begin with the given inspection sequence. The results of an exploratory experiment suggest that the best performance, in terms of reduction in computational effort relative to deviation from the optimum, is obtained with adaptive heuristics, where the number of inspection operations considered for addition to a given sequence varies with the position in the sequence. A sensitivity analysis indicates that the optimal and heuristic procedures are not sensitive to variations in the incoming fraction nonconforming.