A NONPARAMETRIC APPROACH TO ESTIMATE SYSTEM BURN-IN TIME

System burn-in can get rid of more residual defects than component and subsystem burn-ins because incompatibility exists not only among comp onents, but also among different subsystems and at the system level. T here are two major disadvantages for performing the system burn-in: th e high burn-in cost and the complicated failure rate function. This pa per proposes a nonparametric approach to estimate the optimal system b urn-in time. The Anderson-Darling statistic is used to check the const ant failure rate (CFR), and the pool-adjacent-violator (PAV) algorithm is applied to ''unimodalize'' the failure rate curve. Given experimen tal data, the system burn-in time can be determined easily without goi ng through complex parameter estimation and curve fittings.