Using reliability and warranty data to determine the optimal number of parts for a lifetime buy

This paper presents a novel approach to predicting quantities of parts needed for a lifetime buy (LTB) prompted by technology obsolescence or production discontinuance in order to support a long term service contract. LTB is a critical business decision with serious consequences to a company's finances and its reputation, therefore accurate future demand modeling is very important. The forecasting method presented in this paper is based on a combination of factors including reliability and warranty data, demand expectations, penalty for unavailability of parts, aftermarket competition and a few others. The paper will introduce two new terms, demand attrition and demand retention, that describe the decrease in parts demand in the post-warranty period due to aftermarket and non-authentic parts alternatives. The application of the attrition functions (reduction in product population and product demand) has proven to be an important tool to provide accurate estimates for the LTB quantity predictions. Product shipping history can be a source of data to derive the demand retention function, which will depend on the type of the product, warranty terms, and availability of supply alternatives. This paper also presents a case study of an automotive electronics radio after a supplier's notice of discontinuance for the main microprocessor. The implementation of this method at Delphi Electronics & Safety has saved thousands of dollars on the LTB parts inventory by improving the accuracy of forecasting.