A simulation model for warranty analysis

A computer simulation model is developed to yield simulated warranty costs for various combinations of warranty provisions and product life distributions. The inputs to the model are the form of the life distribution (gamma, Weibull, or truncated normal) and its parameters, the lengths of the warranty period and of the life cycle of the item, whether or not the warranty renews upon failure of an item, the form of the rebate function, the seller’s and buyer’s costs and claim validation cost, the probabilities of a claim being made, validated, and of a new purchase being made, the form and rate of the discounting function, and the number of replications of the simulation desired. The output of the model are the mean and standard deviation of the n simulated costs generated for a given set of inputs. The model is validated by comparing its output against the true expected costs generated by a mathematical model in a few special situations where this is possible.