2005 Volume 13 Issue 1-2 Pages 3-23
The maintenance and repair cost of durable goods has traditionally been hidden from consumers and yet has been non--negligible part of Life Cycle Cost (LCC) computation. Predicting the maintenance and repair cost is difficult because many of these durable goods do not have constant failure rates. For some durable products such as automobiles, it is often the case that we have at least a rough idea as to their reliability. In this study we propose and illustrate a method to convert automobile reliability data in the U.S. to their monetary maintenance and repair cost. In our method, we first estimate a statistical model from the widely available reliability data. Then we predict the reliability from the model. Finally we convert the predicted reliability to cost figures. The proposed statistical model takes care of the possible bias introduced by partially missing reliability data. Conversion to cost figure is done on the twenty-six 1996-model-year vehicles popular in the U.S. during 1992--1999.
