Skip to content 
Traditional life Analysis of data involves examining time-to-failure outcomes (for a system, component, or product) collected according to usual conditions to evaluate life specifications and provide predictions regarding product performance. In many cases and for a variety of reasons, such specific data can be considered complicated, or completely impossible. Accelerated life testing is designed to experience product failures in order to investigate their failure models and to comprehend their life qualities in only a short period of time, allowing us to save time and money. The most frequent type of stress loading is constant-stress accelerated life testing, which is simple to perform and has many advantages when applied at a greater than usual stress levels. In constant stress accelerated life testing, each item in the test is put under with a predetermined level of stress. This study introduces a maximum likelihood estimation method of providing point and interval estimations of the parameters, acceleration factor, reliability function, and hazard rate function according to Type II censoring data applying constant stress accelerated life testing. Every stress level’s failure times are considered to follow the Gull alpha power Fréchet distribution. For future order statistics, a two-sample prediction method is also being studied. In addition, a numerical study and a real data set are provided to demonstrate the variety of outcomes obtained.