sequential experimental design

**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.