The Design of Open Engineering Systems Lab

University at Buffalo - The State University of New York

  • Increase font size
  • Default font size
  • Decrease font size

Evaluating the Impact of Performance Uncertainty on Vehicle Demand and

Research Area: Research Publication Year: 2008
Type of Publication: Technical Report
Authors: Naim, Aziz; Ferguson, Scott; Lewis, Kemper
Developing a design that is both highly desirable and technically feasible is a fundamental aspect of preliminary vehicle design. From a customer perspective, desirability can be mapped to the performance specifications of a design. Desirability drives the customer-perceived value of the vehicle, which when combined with price, serves to drive the vehicle’s market share. Such vehicle designs must also be technically feasible from an engineering perspective, ensuring and establishing confidence that they can be realized and manufactured. Traditional analyses of desirability, however, are conducted by presenting discrete performance specifications. Incorporating uncertainty into a preliminary design’s performance allows the consumer to be presented with performance specifications characterized by either a mean value or distribution, permitting investigation of variation from a customer’s perspective. The goal of this research is to take a set of non-dominated solutions and combine them with a random utility model, to explore how the variance in both models impacts the final solution. This paper explores the impact of uncertainty in vehicle design, with potential vehicles first being created and assessed for feasibility. A Technical Feasibility Model is utilized to narrow down the possible designs to those that are both feasible and Pareto optimal. Capturing customer preference ranges for each objective is made possible with the implementation of utility functions whose coefficients are dictated by a referenced market model. In addition to defining the customer preference ranges for this problem, uncertainty is applied to each performance objective considered within the vehicle model. The vehicle designs are then evaluated, combining the variances associates with uncertainty in vehicle performance and the aspects of a random utility model.
Multidisciplinary Analysis and Optimization