TAAF Stopping Rules for Maximizing the Utility of One-Shot Systems

Test-analyze-and-fix (TAAF) is the mostcommonly recognized method of improvingsystem reliability. The work presented hereaddresses the question of when to stop testingduring TAAF programs involving one-shotsystems when the number of systems to beproduced is predetermined and theprobabilities of identifying and successfullycorrecting each failure mode are less than one.The goal here is to determine when to ceasetesting to maximize utility where utility isdefined as the number of systems expected toperform successfully in the field afterdeployment of the lot.Two TAAF stopping rules are presented.Simulation is used to model TAAF executionunder different reliability growth conditions.Four discrete reliability growth models(DRGMs) are used to generate real worldreliability growth and to estimate reliabilitygrowth using hypothetical observedsuccess/failure data. Ranges for the followingparameters are considered: starting reliability,growth rate, maximum achievable reliability,number of systems to be produced, probabilityof incorrectly identifying a failure mode, andprobability of an unsuccessful designmodification.Conclusions are drawn regarding stopping ruleperformance in terms of stopping rule signallocation, utility loss, achieved reliability, andfraction tested. Both rules perform well andare implementable from a practical standpoint.Specific recommendations for stopping ruleimplementation are given based on thecontrollable factors, estimation methodologyand lot size.