Cost Optimization of Product Families using Analytic Cost Models

This paper presents a new method for analysing the cost structure of a mass customized product family. The method uses linear regression and backwards selection to reduce the complexity of a data set describing a number of historical product configurations and incurred costs. By reducing the data set, the configuration variables which best describe the variation in product costs are identified. The method is tested using data from a Danish manufacturing company and the results indicate that the method is able to identify the most critical configuration variables. The method can be applied in product family redesign projects focusing on cost reduction to identify which modules contribute the most to cost variation and should thus be optimized.