| Abstract: | Statistical results of all the 1621 failure analyses, which cover a four‐year effort, are presented. The failure analyses were part of an elaborate test and ESS process of a high‐reliability system. Of the failure analyses, 97.5% identified the root cause. The classification of the failures by 11 root cause classes, including the test and ESS process itself is provided. The distribution of failure root causes detected during the various phases of the process is shown. Each test phase was analyzed according to the share of failures found and the fraction of failures that occurred during that phase. Applying a linear formula that summarizes the relative efficacy of each test phase compares the performances of the various phases. The term efficacy is used to describe the ability of a test phase to reveal failures without generating new problems. Copyright © 2002 John Wiley & Sons, Ltd. |
