Optimal test plans for accelerated life tests based on progressive type-I censoring with engineering applications

In reliability engineering studies, the accelerated life tests with optimal test design that balances cost constraints is particularly important since the lifetimes of electro-mechanical, electronic, and mechanical goods are typically modelled with this characteristic. Comparative optimal test plans for accelerated life tests are developed in a variety of applications, including engineering, biological, medical, and other related fields. In this study, we compare, under cost restriction with a specified budget, optimum multiple constant stress and step stress accelerated life tests based on progressive-type-I censoring. The goal is to assess the advantage provided by step-stress testing over constant-stress testing. In order to conduct an accelerated life test effectively, stress durations need to be carefully specified during the design stage due to limited resources in engineering practice. The estimated accuracy of the parameters of interest and the expense of the experiment are directly impacted by the stress durations. Under a prespecified budget and cost constraints, this article examines the optimal stress durations based on D-, E-, T-, C-, R-, and P-optimal criteria. These criteria are also contrasted together. We illustrate the application of the optimal durations using numerical and simulation examples. A study is conducted to determine how sensitive the optimal duration is to model parameter misspecification.