Most experimental design projects tend to stop after screening, characterizing, or optimizing. However, once a process is optimized, it may need to be implemented in manufacturing. To make the process manufacturable, the process noise variables (e.g. raw material batches, various machines, various operators, etc.) need to be considered. Robust design provides a method to leverage controllable factors to mitigate the transmission of variation from noise variables to process outputs. A dual response model is considered to keep the process mean on target, while reducing noise as much as possible.
In addition, before release of a process to manufacturing, the effects of noise can be quantified via a ruggedness test to ensure that noise variables are not drivers of the process. Both robustness and ruggedness are practical methods to help to make processes more manufacturable.
Who Should Attend
This course is designed for engineers, quality professionals, researchers, and managers who need to implement improvements in manufacturing and want to learn methods to reduce the effect of noise variables or show that noise variables are not drivers of the process.
Through training, participants will:
- Understand Taguchi ideas of robustness, classification of variables, and loss
- Know how to improve the practical side of manufacturing to limit the effect of noise
- Know how to plan, conduct, and analyze designs for robustness improvement
- Know how to plan, conduct, and analyze designs for ruggedness testing
Important Taguchi Ideas
- Robustness & Ruggedness
- Classification of Variables
- Signal to Noise Responses
- Orthogonal Arrays
- Controllable × Noise Variable Interaction
- Dual Response Model for Location and Dispersion
Experimental Designs for Robustness and Simultaneous Optimization and Robustness
- Analysis and Interpretation
- Ruggedness Testing
- Res III Designs
- Replication Noise
Basic DOE and Advanced DOE or the equivalent
Instructor-led class training, with opportunities to practice learned skills using prepared data
Minitab or JMP Statistical Software along with a customized Excel Solution for a Dual Response Model