Douglas Mader

This paper provides a number of key ideas pertaining to the successful deployment of various models for improvement including Lean, Lean Six Sigma and traditional Six Sigma. Training curricula are contrasted for maximum results.

This paper provides a number of key ideas pertaining to the successful identification, selection and launch of Lean Six Sigma projects. A structured process to decompose high-level organizational goals so that Lean Six Sigma projects become well-aligned and supported is presented.

This paper provides a number of key ideas pertaining to the successful deployment of Design for Six Sigma, how to successfully execute DFSS training, some common pitfalls to avoid in DFSS deployment, and the standard DFSS training body of knowledge.

This paper provides an overview of Axiomatic Design, and it provides the linkage of Axiomatic Design to Design for Six Sigma and transfer function modeling. A practical approach is presented so that non-engineers can successfully utilize Axiomatic Design to examine and optimize coupling for the design of a new product, process or service. An overview of requirements decomposition is also presented.

This paper provides an overview of how to implement successful Design for Six Sigma (DFSS) training for manufacturing, R&D, and service organizations. The paper also describes how service organizations can benefit from DFSS and how Service DFSS training needs to be customized according to the design environment.

This paper provides an overview of how to implement successful Design for Six Sigma (DFSS) training for manufacturing, R&D, and service organizations. The paper also describes how service organizations can benefit from DFSS and how Service DFSS training needs to be customized according to the design environment.

This paper provides an overview of how to successfully deploy Design for Six Sigma (DFSS), a description of the ICOV (Identify-Characterize-Optimize-Validate) DFSS methodology, and a description of how to integrate the ICOV DFSS methodology with an existing new product development (NPD) process.

Interpretation of the results of a gauge repeatability and reproducibility (R&R) study relies on two key metrics -- the combined gauge R&R as a percent of specification and the discrimination ratio. These metrics are useful in terms of the rules of thumb that have been developed over the years, but they fall short of effectively quantifying how well the measurement system will perform in a production environment. For any measurement system, the alpha (or producer’s) risk is defined to be the… Show more

Interpretation of the results of a gauge repeatability and reproducibility (R&R) study relies on two key metrics -- the combined gauge R&R as a percent of specification and the discrimination ratio. These metrics are useful in terms of the rules of thumb that have been developed over the years, but they fall short of effectively quantifying how well the measurement system will perform in a production environment. For any measurement system, the alpha (or producer’s) risk is defined to be the probability that the test will fail a unit of product that conforms to specification. The beta (or consumer’s) risk is defined to be the probability that the test will pass a unit that does not conform to specification. When the alpha and beta risks have been estimated, the economic impact of the performance of the measurement system can be evaluated. A method for estimation of the alpha and beta risks has existed for some time in the literature, but it is not commonly practiced in the analysis of measurement systems performance. The authors argue that risk analysis should be a fundamental component of any R&R study. A case study is provided to provide insight into the usefulness and applicability of this technique. Show less

Quality control and process control are based on data that are sequentially collected. The collection is displayed and analyzed, either in "real time", which means that a very fast program works on the data as soon as they are generated, or later after perhaps some cleaning up. It is a fact of life that most data are naturally multivariate. The classical Shewhart approach, dating back to 1924, tracked one variable by detecting shifts in the mean or variance with the assistance of control… Show more

Quality control and process control are based on data that are sequentially collected. The collection is displayed and analyzed, either in "real time", which means that a very fast program works on the data as soon as they are generated, or later after perhaps some cleaning up. It is a fact of life that most data are naturally multivariate. The classical Shewhart approach, dating back to 1924, tracked one variable by detecting shifts in the mean or variance with the assistance of control limits. Later the Western Electric Company (WECO) introduced a set of rules to interpret the manner by which out of control situations occurred. Over time a variety of additional charts have been developed and used as auxiliary tools in the arena of the SPC, SQC and TQM efforts in industry. These include Moving Range, CUSUM, EWMA, median, runsum, etc. A detailed description of these univariate control charts is found in Montgomery(1991) Hotelling in 1947 introduced a statistic which uniquely lends itself to plotting of multiple observations. This statistic, appropriately named "Hotelling T^2" is a scalar, that combines information from the dispersion and mean of several variables. Due to the fact that computations are plentiful and slightly complex and require some knowledge of matrix algebra, acceptance of multivariate control charts by industry was slow and hesitant. However, modern computers in general and the PC in particular changed all that and during the last decade, multivariate control charts started to arrive. Show less

This paper describes an application in real-time multivariate statistical process control for the automated power calibration of a family of high-energy power supplies. Subsequent to a variance-stabilizing transformation, data for the actual power produced by each unit under test is regressed on the specified input power over the operating range. The slope and intercept from the simple linear regression are then compared to the bivariate normal distribution for slope-intercept pairs from… Show more

This paper describes an application in real-time multivariate statistical process control for the automated power calibration of a family of high-energy power supplies. Subsequent to a variance-stabilizing transformation, data for the actual power produced by each unit under test is regressed on the specified input power over the operating range. The slope and intercept from the simple linear regression are then compared to the bivariate normal distribution for slope-intercept pairs from previous units using a prediction approach based on multivariate multiple linear regression. The approach has been effective in identifying quality problems on-line, and the efficiency of the multivariate SPC procedure has decreased total test cycle time at the automated test. Process capability results for the power calibration test are presented for the first 632 production units for the Lightning power supply family, and the results indicate strong evidence that a six sigma level of quality has been achieved for the power calibration process. Show less

This work describes the application of several common process control techniques.