Robust Design creates products and processes that perform consistently despite noise, variation, wear, environment, and normal usage differences.
Definition
Robust Design is the practice of designing products, processes, and systems so performance remains stable under expected sources of variation. It focuses on reducing sensitivity to noise factors such as environment, material variation, operator differences, wear, customer use, and upstream variation.
The aim is not only to hit a target under ideal conditions, but to keep performance close to target in real conditions.
History
Robust Design is closely associated with Genichi Taguchi and quality engineering methods that emphasize variation reduction and loss to society. It also connects with DOE, tolerance design, reliability, and design for Six Sigma.
When to Use
Use Robust Design during product development, process development, parameter optimization, supplier qualification, and chronic variation reduction. It is valuable when performance is sensitive to environmental, material, setup, or usage variation.
Step-by-Step
- Define the ideal function, target, and critical responses.
- Identify control factors and noise factors.
- Choose experiments or simulations to test sensitivity.
- Find settings that reduce variation while meeting targets.
- Review tradeoffs, cost, manufacturability, and safety.
- Confirm performance under realistic noise conditions.
- Set tolerances and controls based on sensitivity.
- Transfer learning into design standards and control plans.
Examples
- Product: A seal design works across temperature, pressure, and supplier material variation.
- Process: Welding settings are chosen for strength consistency despite material lot variation.
- Service: A workflow is designed to maintain turnaround time despite daily demand variation.
Common Pitfalls
- Optimizing average performance while ignoring variation.
- Not identifying real noise factors.
- Testing only ideal lab conditions.
- Using tight tolerances instead of reducing sensitivity.
- No confirmation under realistic use.
- Ignoring cost and manufacturability tradeoffs.
