The basic Engineering Design Optimization (EDO) problem involves the selection of process control variable settings that produce the best process performance. Since many engineers use computationally intensive computer models to evaluate performance, their inability to efficiently evaluate a large number of design alternatives becomes a key roadblock to completing a design optimization study. The Robust Engineering Design problem is an extension of the basic EDO problem. Here, the engineer wishes to select control variable settings that minimize the impact of input variation on process performance. In addition to a need for efficient estimates of performance variation, the engineer also needs estimates of input variation. This talk will survey a collection of tools that address the issues described above and thereby enable a variety of EDO studies that were previously impractical.

Kurt Palmer is an Assistant Professor in the Daniel J. Epstein Department of Industrial and Systems Engineering at the University of Southern California. His research focus is on data collection planning and empirical model building methods. He is currently developing methods intended to improve the efficiency of engineering design studies conducted on design simulators. Prior to joining the faculty at USC, Dr. Palmer was employed for six years as a Manufacturing Engineer at the Eastman Kodak Company; and for four years, he was the Owner of Quality Management Consulting in Rochester, NY. His Ph.D. in Industrial and Systems Engineering is from the Georgia Institute of Technology.