Extrapolation in Robust Design Problems

• Activity Number: 70
• Type: Invited
• Date/Time: Monday, August 4, 2003 : 8:30 AM to 10:20 AM
• Sponsor: Section on Physical and Engineering Sciences
• Author(s): Michael J. Luvalle*+ and Ana Ivelisse Aviles
• Companies: Lucent Technologies and NIST
• Keywords: extrapolation ; robust design ; scale-up

Abstract

Robust design experiments can be used in a number of problem areas. This paper focuses on the problem of efficiently designing a scale up from laboratory to factory production. We search for a design point for our process that minimizes variability while getting the operating characteristics on target. In a laboratory setup, the purpose is to show that a product can be created by highly skilled and educated workers to meet an approximation of the customers initial requirements. In the factory the goal is to have a process for creating the product which provides customer satisfaction in an efficient and economical manner. Thus it is often true that the laboratory processes are nowhere near optimized for the manufacturing environment. It may often be the case that the optimal, (or even the economic break even point) for reduced variability with the product on target is outside of the regime where experiments are reasonably conducted on the current laboratory setup. We study experiment design in this situation. We examine the problem of reducing the risk associated with scale up by adding exploratory experiments in the laboratory accessible regime.