A new generalized burn-in procedure for items in stochastically evolving population

In practice, the reliability performance of items in operation strongly depends on the operational history or the working environment. Thus, most often, the population stochastically evolves depending on it rather than being static. For example, the failure and repair history strongly affects the future reliability performance of the item in operation. In this case, even though the population of items initially starts from a homogeneous one, it gradually evolves to a heterogeneous one. Until now, most studies on burn-in have been performed assuming that the population is static. In this paper, a new generalized burn-in procedure for items in stochastically evolving population is developed and its stochastic properties are analyzed. Based on the obtained theoretical results, the optimal burn-in procedure which balances two criteria is studied.