Sequential Bayesian Design for Accelerated Life Tests

To ensure the product can last long enough in the field, the accelerated life test is a useful experiment to obtain the product's lifetime information and then make prediction at the normal use condition. Due to the limits of budget, machines and time, it is important to design an efficient experiment to obtain enough lifetime information. By using the Bayesian technique, this study proposes efficient sequential Bayesian strategies based on multiple experimental machines when there are more than two machines in a lab, called the batch sequential design. This study also uses the simulation way to explore the performance of the proposed strategies and compare the performance to the existing sequential Bayesian design based on only one machine.