Comparison of type-I and type-II accelerated life tests for selecting the most reliable product

In product research and development, one might wish to select the most reliable of several highly-reliable competing products. Both type-I and type-II accelerated life tests (ALT) are well-known to estimate and compare the life of highly reliable (nonrepairable) products within a reasonable testing time. A common decision problem is to choose the better plan for achieving the above goal. This paper proposes an intuitive selection rule for each ALT plan, and then derives the respective optimal ALT sampling plan when the acceleration factor (AF) is known and unknown. Finally, some criteria of Mackay are used to compare the two plans. From a practical point of view, the type-II ALT sampling plan is more convenient, and needs smaller sample size than the type-I ALT sampling plan. Nevertheless, the latter has shorter mean duration time and life testing time than the former.<
>