Gamma Degradation Models: Inference and Optimal Design

During the past two decades, degradation analysis has been widely used to assess the lifetime information of highly reliable products. Usually, random effect models and/or Wiener processes are well suited for modelling stochastic degradation. But in many situations, such as materials that lead to fatigue, it is more appropriate to model the degradation data by a gamma degradation process which exhibits a monotone increasing pattern. This article surveys the theoretical aspects as well as the application of gamma processes in degradation analysis. Some statistical properties of degradation models based on gamma processes under different tests are also given. Furthermore, the corresponding optimal designs for conducting the degradation experiments efficiently are reviewed. Finally, some extensions and their applications of gamma-process degradation model are presented.