home > research > initiatives > parameter design

Parameter Design

Devising Robust Strategies for Experimentation

This research establishes guidelines for the selection of experimental strategies in the parameter design of engineering systems.

We are currently exploring the performance of “one-factor-at-a-time” experimentation plans in comparison with that of orthogonal arrays.

Focusing on degree of improvement, we use computer simulations to determine those conditions for which adaptive one-at-a-time strategies outperform orthogonal arrays.

Initiative Leader:
Daniel Frey
Assistant Professor of Mechanical Engineering and Engineering Systems
danfrey@mit.eduon
Key Research Paper:

A Role for "One-Factor-at-a-Time" Experimentation In Parameter Design


Professor Frey’s research focuses on the robust design of engineering systems—ensuring their proper functioning regardless of variations due to manufacture, wear, deterioration, and environment. He also explores the complementary roles of experimentation and simulations in systems engineering.