Article ID: | iaor20072830 |
Country: | South Korea |
Volume: | 33 |
Issue: | 1 |
Start Page Number: | 76 |
End Page Number: | 85 |
Publication Date: | Jan 2007 |
Journal: | Journal of the Korean Institute of Industrial Engineers |
Authors: | Yum Bong-Jin, Koo Jung-Seo, Kim Min |
A reliability acceptance sampling plan (RASP) consists of a set of life test procedures and rules for either accepting or rejecting a collection of items based on the sampled lifetime data. Most of the existing RASPs are concerned with the case where test items are available at the same time. However, as in the early stage of product development, it may be difficult to secure test items at the same time. In such a case, it is inevitable to conduct a life test using sequentially supplied samples. In this paper, it is assumed that test items are sequentially supplied, the lifetimes of test items follow an exponential distribution, failures are monitored continuously, arrival times of test items are known, and the number of test items at each arrival time is given. Under these assumptions, RASPs are developed by determining the test completion time and the critical value for the maximum likelihood estimator of the mean lifetime such that the producer and consumer risks are satisfied. Then, the developed plans are compared to the traditional Type-I censored RASPs in terms of the test completion time. Computational results indicate that the test completion time of the developed RASP is shorter than that of the traditional Type-I censored plan in most cases considered. It is also found that the superiority of the developed RASP becomes more prominent as the inter-arrival times of test items increase and/or the total number of test items gets larger.