李赞博士，南开大学统计与数据科学学院

The optimal planning of condition-based maintenance strategy based on the degradation process is of great concerned in the past few years. A lifetime delayed degradation process with heterogenous degradation rate is conducted in this study. Amounts of engineering data show that the initiation time is various from different component and is corrective with the degradation rate. Random effect is considered to describe the correlation between initiation time and degradation rate. As a result, a sequential maintenance policy is developed to minimize the expected cost, providing a dynamic inspection, suiting the heterogeneity between products. Value of information is involved for optimized the subsequence inspection and maintenance action.