An improved method for evaluating equivalent acceleration factors of repairable systems in a fleet

One important issue in reliability modeling of repairable systems is to quantitatively evaluate the heterogeneity of systems in a fleet. Recently, a fleet heterogeneity evaluation method has been proposed in the literature. It first estimates the mean times between failure (MTBFs) of individual systems, then defines the equivalent acceleration factor (EAF) of a specific system as the ratio between the MTBFs of this system and a reference system, and finally a heterogeneity measure is defined as the ratio of the sample average and standard deviation of EAFs of all the systems. The smaller this ratio is, the larger the heterogeneity of the fleet is. It is well-known that the estimated MTBF is inaccurate when the failure number is relatively small. Thus, a main drawback of the above-mentioned method is that the MTBFs of different systems are estimated at different time points. This paper aims to address this issue by proposing an improved method for estimating the MTBFs and EAFs of systems. The proposed method uses an "average process" as the reference process and estimates the MTBFs at a common time point. Two datasets are analyzed to illustrate the proposed method and its appropriateness.