Extracting Practical Information from Quality-by-Design Models

Presenter:  Rob Lievense

Using Quality by Design

See how to:

  • Understand the value of QbD
    • Developmental studies lead to enhanced knowledge of a process (design space)
    • Higher levels of quality, faster development and approval of new products, and reduced cost, regardless of industry
  • Understand prerequisites
    • Assure risk analyses with cause and effect diagrams is completed
    • Solicit knowledge of first principals and prior experience to identify  process inputs that could likely affect results
  • Explore design space
    • Reduce models to include significant factors
    • Explore models dynamically using Prediction Profiler
    • Use Profiler to optimize responses
  • ·Create control space within design space
    • Create Contour Plot grid to visualize design space
    • Define control space for a pair of inputs with acceptable operation ranges
    • Identify realistic control space that is well within design space
  • ·Simulate using Prediction Profiler using Process Variability
    • Visualize variability of actual process input measures
    • Summarize behavior of  process inputs using Distributions
    • Include input variability with simulation to estimate likely defect rate
  • ·Estimate robustness of control space
    • Create simulated data for each corner of control space
    • Combine into one JMP data table
    • Run capability study to determine worst-case robustness of control space

Note: Q&A included at times 14:01, 14:48, 16:04, 17:58, 18:28, 32:19, 34:25, 35:45, 37:09, 39:11, 40:32 and 42:38.

Back to Top