Military strategy: the way to better autos, aircraft
UB professor receives prestigious NSF award to apply game theory to design process

By ELLEN GOLDBAUM
News Services Editor

Most people don't think of the design of an aircraft or automotive part as a particularly contentious process, but a UB professor has received a prestigious National Science Foundation grant to apply game theory-the same theory military strategists use-to improving the process.

Kemper Lewis, assistant professor of mechanical and aerospace engineering, recently was awarded an NSF Faculty Early Career Development grant that can pay up to $400,000 over four years with industry support. He is working with Alison Rolls Royce Engine Company; Praxair, Inc.; Milliken Research Associates, a company that consults in the design of race cars, and NASA-Langley Research Center.

The awards recognize young faculty members who have demonstrated outstanding potential as science and engineering investigators and educators.

It's not, Lewis explained, that designers are at war with one another.

It's just that the whole design process for aircraft, cars, buildings, copiers or any large-scale system is so complex, and so many variables must be considered all at once, that arriving at an optimum process that will result in the best product is extremely difficult.

"In war, there's the issue of noncooperation in terms of the enemy," said Lewis. "In design, the parallel is isolation. Design teams are isolated from the information they need, either from other designers or design teams. They may have to assume a worst-case scenario and that can lead to suboptimal design."

The goal is to create tools and methods that will help designers make the best trade-offs, situations where a change in one design parameter significantly affects one or more design objectives, constraints or disciplines, such as structures or aerodynamics.

While the basis for the research is to simulate design processes mathematically, Lewis noted that there are undoubtedly some politics involved as well.

"If we can get design teams to recognize the important trade-offs inherent in their design processes," he said, "they can start to exercise these trade-offs using rational decision-making, which will lead to better cooperation. Then we will be able to break through some barriers, which in practical terms means breaking through some of the politics."

At the same time, the project aims to provide designers with tools that will help them deal with uncertainty and to figure out which aspects of a project they can compromise on when they don't have all the necessary information.

"This is all about process and product optimization," said Lewis. We're trying to get the right product but in order to do that, we must determine which process gives us the product we want."

Lewis conducts research on game-theoretic and design-based approaches to complex systems design, multidisciplinary optimization, mixed discrete/continuous design optimization and product and process trade-offs.

He earned two bachelor's degrees from Duke University and master's and doctoral degrees from Georgia Institute of Technology.

Lewis is a recipient of a Sloan Foundation New Faculty Fellowship, a United Engineering Foundation Fellowship, the Xerox Engineering Excellence in Design award and the UB Riefler Award