Inference of the Life Information from Accelerated Degradation Data

In the applications of reliability engineering, life information of components is important to improve the quality of products from manufacturers. However, engineers lack of confidence to verify the mean time to failure of high-reliability products from aging or degradation tests. Because the maximum likelihood estimate for degradation data often overestimates the mean time to failure of components, a bias-corrected maximum likelihood estimation method is provided to infer the life information of high-reliability components. Moreover, three nonparametric bootstrap estimation methods with algorithms are provided to assign measures of accuracy to statistical estimates. An example of degradation data of GaN-based power light emitting diodes is used to demonstrate the use of the proposed method.