In system reliability engineering, systems are composed of multiple components and can exhibit complex structures. Engineers and researchers are often interested in both system and component lifetime distributions. Frequently, only the lifetimes of n-component coherent systems are observed, not those of individual components. Recent developments have focused on parametric and nonparametric inference for component lifetimes based on system data. This talk discusses statistical inference of component reliability characteristics using system lifetime data with the same structure. We first address the problem of testing the homogeneity of component lifetime distributions using system data with known signatures. Several nonparametric test statistics based on the empirical likelihood method are proposed, applicable to both complete and Type-II censored data. Their performance is compared with existing methods. We then study the optimal design of constant-stress life-testing experiments using n-component systems. Proposed schemes aim to shorten experimental time while maintaining statistical accuracy. Performance comparisons of these schemes are conducted via mathematical analysis and Monte Carlo simulation. The merits of the proposed designs and directions for future research are discussed.

Hon Keung Tony Ng is a Professor with the Department of Mathematical Sciences, Bentley University, Waltham, MA, USA. He received a Ph.D. degree in mathematics from McMaster University, Hamilton, ON, Canada, in 2002. He is the co-editor of Communications in Statistics – Simulation and Computation, an associate editor of IEEE Transactions on Reliability, Journal of Statistical Computation and Simulation, Naval Research Logistics, Sequential Analysis, and Statistics & Probability Letters, and serves on the editorial board of International Journal of Reliability, Quality and Safety Engineering and Statistical Methods & Applications. His research interests include reliability, censoring methodology, degradation modeling, ordered data analysis, non-parametric methods, and statistical inference. He has published over 180 research papers in peer-reviewed journals. He is the co-author of the books Precedence-Type Tests and Applications (Wiley, 2006) and Fiber Bundles: Statistical Models and Applications (Springer, 2023), and co-editor of Ordered Data Analysis, Modeling, and Health Research Methods; Statistical Modeling for Degradation Data; Statistical Quality Technologies: Theory and Practice; Bayesian Inference and Computation in Reliability and Survival Analysis; and Recent Advances on Sampling Methods and Educational Statistics. Professor Ng is an elected senior member of IEEE (2008), an elected member of the International Statistical Institute (2008), an elected fellow of the American Statistical Association (2016), and a senior member of the American Society for Quality (2025).