This article describes the repair-replacement process of a product under warranty as a stochastic process and develops a model to estimate the costs due to warranty. The product is assumed to be made up of several components for which failure rates are available. Markovian states are defined dependent on the number of failures of each component. A replacement policy is superimposed, and an approach to obtaining the transition probabilities is illustrated when the failures are solely due to chance. The use of the model to estimate the warranty costs during an accounting period, when there are warranties outstanding for the products sold during prior periods in addition to the products sold during the current period, is discussed.

In this paper, a mathematical modeling approach to the audit staff allocation problem is proposed. The solution procedure is interactive and recognizes the multiple-criterion nature of the problem. The interactive procedure operates within a fixed number of iterations and presents a fixed number of candidate solutions to the decision-maker at each iteration. The decision-maker is required only to select his most preferred candidate from each of the groups of solutions presented to him. Criterion preferences in terms of a weighting vector are not elicited. The procedure's final solution provides two types of recommendations' (a) how the projected workload is to be allocated among the audit staff, and (b) whether or not the audit staff should be increased or decreased at its different levels.