A product cost estimation for the early design of sedans using neural networks

Publication Type:

Journal Article


Ju, B.; Xi, L.-F.


International Journal of Automotive Technology and Management, Volume 8, Number 3, p.331-349 (2008)




Reducing the product cost at the early design stage is rarely implemented among Chinese automakers due to the lack of an effective cost estimation tool. In this paper, a Back Propagation (BP) neural networks-based cost estimation method is proposed for designers and managers. The product cost differences that are caused by design changes could be evaluated and the market position of a new sedan could be targeted. To avoid the inaccessibility of the confidential cost data, the product costs are calculated backward from retail prices with product cost ratios. Instead of introducing pilot data, real-world data are adopted in this study. The prices and specifications of sedans are retrieved through an authentic automobile website, while the product cost ratio is estimated by a panel of experts with the Delphi method. The neural networks are trained with the data to estimate the product cost. The application of this method for new sedan development in a Chinese automotive company is also presented. Copyright © 2008 Inderscience Enterprises Ltd.

  GIS Gerpisa / gerpisa.org
  4 Avenue des Sciences, 91190 Gif-sur-Yvette

Copyright© Gerpisa
Concéption Tommaso Pardi
Administration Juan Sebastian Carbonell, Lorenza MonacoGéry Deffontaines

Powered by Drupal, an open source content management system