Optimal burn-in policy by using an integrated Wiener process

The conventional burn-in test, which only collects go/no go (or time-to-failure) data over short testing periods becomes rather ineffective for highly reliable products. This decision problem can be solved if there exists a suitable quality characteristic whose degradation over time can be related to the product's lifetime. An innovative approach to determine the burn-in policy by using a suitable degradation model has been suggested in the literature. However, a major disadvantage of this procedure is that it only uses information contained in current degradation data, and it ignores any information given by the entire sequence of observations. Thus, it becomes relatively insensitive to detecting weak items if the degradation path of the product degrades very slowly. To remedy this weakness, similar to the CUSUM control charting scheme in statistical process control, we propose an integrated Wiener process to model the cumulative degradation path of the product's quality characteristic. Then the optimal burn-in policy can be easily obtained. In addition, we also use an analytic approach to compare the efficiency of our proposed procedure with the method that is previously reported in the literature.