Improved reliability-prediction and field-reliability-data analysis for field-replaceable units

This paper presents a method for analyzing field-replaceable unit (FRU) field-data to obtain accurate field-reliability estimates. The estimates are used to improve reliability prediction and to prioritize the FRU that potentially need design improvement. Application of the method to a large telecommunication project reveals important differences in the age-dependent reliability estimates and the predicted constant estimate. The observed difference is used to measure the statistical bias due to special causes, and a correction term is added to obtain the improved reliability prediction model. The results of applying the priority scheme to FRU data are encouraging and show the importance of including quality-of-information in decision-making.