Optimal warranty policy with inspection for heterogeneous, stochastically degrading items

A new renewable warranty policy is suggested that increases probability of its success and can decrease warranty costs. An item from a heterogeneous population is inspected at some intermediate time during a warranty period and, if the observed level of degradation/wear exceeds some optimally predetermined value, it is screened out and replaced by the new one. Deterioration in homogeneous subpopulations of items is modeled by the inverse-Gaussian (IG) process, whereas heterogeneous populations are described by the mixed IG process. Probabilistic and cost analyses of the model are performed and the detailed illustrative example is presented and discussed. (C) 2020 Elsevier B.V. All rights reserved.