Our goal is to select a robust vehicle portfolio mix and optimize its design attributes such that contribution margin is maximized while being regulation compliant under varying fuel prices. Compliance to regulation is measured in terms of the Corporate Average Fuel Economy or CAFE. We formulate this vehicle portfolio optimization problem as a mixed integer non-linear programming problem, both under static and varying fuel price scenarios. We demonstrate our approach using a case study in which an in-house market simulator is employed for incorporating consumer preferences in portfolio decisions. This market simulator uses real-time preferences from tens of thousands of shoppers and captures preference heterogeneity using different Logit coefficients for each shopper and hence is computationally expensive. Also, it does not explicitly model the influence of fuel price in predicting demand. To overcome these issues and to facilitate portfolio optimization we use meta-models of the market simulator. Our results show that while remaining regulation compliant it is also possible to achieve significant improvement in the portfolio’s contribution margin. In some scenarios, the improvements in contribution margin are more than 40% when compared to the traditional approach of using expert judgment to decide the portfolio mix.

This content is only available via PDF.
You do not currently have access to this content.