Maximum Likelihood Estimation and Conservative Confidence Interval Procedures in Reliability Growth and Debugging Problems

An examination of two problems. The first deals with estimating the reliability of a system undergoing developmental testing to increase its probability of successful operation, increase its time-to-failure, or decrease its failure rate. If one or more of these occur, reliability growth is taking place. Three models are formulated, and appropriate maximum-likelihood estimates and conservative confidence-bound procedures are derived for them. The second problem deals with the debugging of a new, complex system during the initial period of its life. At this time, failures and errors are corrected as they occur, with improvement in subsequent performance of the system. In practice, the debugging phase is considered completed when the failure rate reaches an equilibrium or constant value. Models are formulated for this phenomenon. Maximum-likelihood estimates are obtained for relevant failure rate functions and for the end of the debugging period.