Simplified Taguchi/DOE Methods
I.D. # 96017 Duration 2 Days

Companies realize that they need to do more with less which means we need to use the most efficient and effective methods. This seminar blends the philosophy of Taguchi with the simple graphical methods of Box, Hunter, & Hunter to give a powerful set of DOE tools.

Wide use of Design of Experiments or DOE methods has been hindered by complications in planning a DOE to handle interactions and by analysis complexity of ANOVA. A Preferred Columns Method simplifies planning so engineers can assign factors to an array in minutes. Graphical methods allow quality professionals to distinguish large (active) factors from small terms and portray these findings to broad audiences. By simplifying DOE's, road blocks are removed so that more people can begin using these highly productive methods.
Learning Objectives
By attending this seminar, you will be able to:
  • Explain Taguchi's Quality Philosophy of Closeness to Target
  • Plan efficient factorial experiments using the Preferred Columns Method
  • Compare data sets with dual or multiple stem and leaf plots
  • Analyze the data using simple graphical methods such as dot plots and probability plots
Who Should Attend
This seminar will be most beneficial to engineers and quality professionals who work in product design, manufacturing, testing, reliability, quality, process, or development.
Prerequisites
Participants should have an undergraduate engineering degree or experience as a quality professional.
Seminar Content
  • Taguchi Loss Function
  • Comparison of Two Means
    • Dual stem & leaf plots
    • Exceedances test
    • Graphical T-test for differences
  • One-Factor Experiments
    • Multiple stem & leaf plots
    • Graphical F-test for multiple comparisons
  • Two-Factor Experiments
    • Response plots to see effects and interactions
    • Graphical F-tests to compare averages
  • Full Factorials
    • Math model behind factorial experiments
    • Independent factors vs. interactions
    • Randomizing the runs
  • Planning Fractional Factorials
    • Measurement system that uses continuous data
    • How to select factors for a DOE
    • Confounding patterns and resolution
    • Preferred Columns Method to set up a fractional factorial
    • Data sheets
    • Customizing for a 4-level factor
  • Analyzing Fractional Factorials
    • Column effects table
    • Quantifying the noise of repeats
    • Dot plots with error distributions
    • Interaction plots
    • Prediction model
    • Cube plot
    • Probability plot of column effects
Instructor(s): Jerry L. Roslund
Dr. Jerry Roslund is an independent consultant specializing in the Design of Experiments (DOE) and Reliability Methods. Prior to retiring from GM, he provided leadership as a GM Technical Fellow for Statistical Methods and Reliability and as a Validation Technical Integration Engineer at GM North America Vehicle Engineering. Specifically, Dr. Roslund conducted seminars on the Key Aspects of Quality, Reliability and Durability (QRD), Weibull Analysis, and Design of Experiments (DOE) for leaders and engineers within both GM and GM Suppliers.

Dr. Roslund joined General Motors at Detroit Diesel Allison, transferred to Saturn, and retired from GM in 2007. Dr. Roslund spent 24 years at General Motors solving problems using statistical methods and providing in-house consulting on a daily basis. He also developed numerous course textbooks and conducted over 200 seminars providing a wealth of experience and a vast number of case studies.

Dr. Roslund is a member of SAE, ASQ, and the Society of Reliability Engineers. He received his B.S. degree in Mechanical Engineering from the University of Nebraska, M.S.M.E. from Cleveland State University, and Ph.D. in Systems Engineering from Oakland University.
Testimonial
"Very practical seminar that teaches the use of a powerful tool, the DOE. Excellent instructor."
Paul Adamovicz
Senior Applications Engineer
Robert Bosch Corporation

"This seminar had my gears turning, thinking of applications to use the content at work. It was exciting knowing that I fully understood the topic and could easily use it at work with tremendous results."
Dorothy Brown
Quality Supervisor
CME Corporation
Fees: $1285 SAE Members*: $1028 - $1157
* The appropriate SAE Member discount will be applied through the Registration process.  Discounts vary according to level of membership: Elite Member 20%; Premium Member 15%; Classic Member 10%
CEU 1.3