Yujia Luo, a Master of Science candidate in the Department of Mathematics and Statistics, will present their STAT 5P99 MRP titled Optimal Design of Progressive Stress Loading Accelerated Life Tests under Progressive Type II Censoring on Monday, July 14, 2025 at 2:30 PM online on Microsoft Teams.

The examination committee includes supervisor Dr. Xiaojian Xu and Supervisory Committee Member Dr. S. Ejaz Ahmed.

Students (both graduate and undergraduate) as well as other members of the Brock Community are invited to attend. If you are interested in the presentation, please contact Jesse Larone at [email protected] for a link to the team presentation. Please join with microphones and cameras turned off.

Keywords: Mean Lifetime Estimation, D-optimality, A-optimality, Exponential distribution, Weibull distribution, Robust design, Relative Efficiency

Abstract:
In traditional life testing experiments, it is often difficult to observe failures in highly reliable products due to their long lifespans, especially under normal operating conditions.  Accelerated life tests (ALT) is a widely used strategy to obtain sufficient failure data within a shorter period.

This project focuses on developing optimal and robust designs for ALT using progressive stress loading plans under progressive Type-II censoring.  The underlying lifetime distributions are assumed to be either exponential or Weibull.  Optimal designs are derived based on three different criteria: c*-optimality, D-optimality, and A-optimality.  The resulting designs demonstrate high efficiency.

However, these optimal designs are only locally optimal, as they depend on unknown parameters that must be estimated.  To address this uncertainty, we adopt a minimax approach to identify robust stress loading rates that minimize the maximum potential loss.  The resulting minimax designs are shown to be robust over a plausible range of parameter values.