Darrell Teegarden
Mentor Graphics

A methodology using modeling and simulation, in conjunction with Design for Six Sigma (DFSS), provides an incubator for innovation where a broad range of new ideas can be evaluated, optimized for cost, and quickly put into production. Two complementary technologies make this possible: DFSS methodologies, which provide techniques for designing experiments and assessing quality; and model-driven development techniques, which allow the design to be captured using mathematical models—or virtual prototypes—and exercised using a computer.

Many organizations have turned to DFSS methodologies to systematically build in quality at each step of the process. Choosing these methodologies, however, requires a commitment to systematic design experimentation that can become prohibitively expensive, cumbersome, and time consuming if it is to be accomplished using physical prototyping. This paper describes how DFSS methodologies combined with model-driven development can produce order-of-magnitude improvements in both productivity and quality when virtual prototyping, automated data collection, and statistical analyses are used to guide the model-driven development process.