Hybrid Block Censoring in Bathtub‐Shaped Lifetime Models and Their Optimal Censoring Strategies

ABSTRACT In this research paper, we consider the lifetimes of engineering systems across different test conditions, using a two‐parameter bathtub‐shaped model to reflect varying risk levels. Both frequentist and Bayesian methods are proposed for estimating model parameters with an evaluation of their respective confidence intervals. These methods provide estimates for the component reliability, hazard rate, and median time to failure as measures of reliability performance. Additionally, criteria for selecting an optimal block hybrid censoring plan are discussed. The methods are validated through extensive simulations and data analyses, highlighting their effectiveness and applicability to real‐world survival analysis problems.