Abstract:

Appendix A in GJB/Z77 describes a method to use AMSAA model to estimate reliability growth for multi systems development, in which the trend test, parameters′ estimation and goodness of fit test for a numerical example are given. For the case that more than one system of the same type is put into reliability growth test, once a Type B failure mode seen during test, corrective action will be taken to all systems. In this paper, it is shown that there is something wrong with this method. According to GJB/Z77, the maximum likelihood estimation of MTBF for multi systems reliability growth test is much larger than that according to AMSAA model for a single system; the more systems put into test,the larger the estimation of MTBF.

Key words: r eliability gr owth, multi-systems, AMSAA model, GJB/ Z77