Sequential Experimental Design is a DOE strategy where experiments are planned in stages — each stage's design is informed by the results of previous stages, enabling efficient convergence toward optimal conditions by focusing experimental effort on the most promising regions.

Sequential DOE Workflow

• Stage 1 (Screening): Broad factorial or Plackett-Burman design to identify significant factors.
• Stage 2 (Augmentation): Add center points and axial runs to the significant factors for curvature estimation.
• Stage 3 (Optimization): Fine-tune around the predicted optimum with additional experiments.
• Adaptive: Each stage adapts based on what was learned — no wasted experiments in unimportant regions.

Why It Matters

• Efficiency: Uses 30-50% fewer experiments than a single large design covering all factors.
• Learning: Each stage builds knowledge that focuses subsequent experiments.
• Risk Reduction: Small initial experiments reduce the risk of running many wafers in an unproductive region.

Sequential DOE is learning as you experiment — using each batch of results to plan the next, most informative set of experiments.