There is a new paper on our website! Phil has put together a very useful presentation that details how a team overhauled a poorly performing anodizing process (yield 40% at best), and turned it into one that produced virtually no defects!

The team accomplished this dramatic turn-around with heavy use of JMP's dynamic visualization capabilities (distribution, scatterplot matrix, 3D scatterplot) and JMP's Custom Profiler. Custom Profiler was used to design an experiment under fairly stringent trial constraints. The Profiler was used to find optimal settings and simulation was used to predict process capability using knowledge of variability in the factors. The paper also describes the interpretation of sensitivity indicators. The new process had a yield of 99%!

Phil is presenting this talk at the Midwest SAS User's Group meeting in Indianapolis on October 13. Be sure to attend if you can!!!

Anodize Parts Talk - MWSUG

What folks had to say about the Spring 2008 DOX Course:

“Using a project related to our company was a very practical way to learn the materials, since you are comfortable using the data because it relates to your “widgets”. The method forces you to use data to validate assumptions, and you may discover that the assumptions are not accurate.” - Charles Ackerman, Staff Engineer and Manager of Compliance Standards at FCI USA, Inc., Electrical Division in Manchester, NH

"The program has been great to allow me to use practical applications to things previously learned. I have found the program to be very beneficial and enlightening. I would recommend it to anyone who works in this field. It's very professionally done." - Richard Helminski, a quality assurance manager

We are excited to be offering a 6.5 day public training program on the design of experiments this fall (2008). The training will be held in Manchester, NH at the UNH Manchester campus. There will be full day training sessions on October 2, 3, November 6, 7, December 11, 12, and there will be a half-day symposium on February 13, 2009.

Participants should come to the training with a company project that would benefit from experimental design. In fact, we encourage that small teams from a company attend the training. This approach leverages the knowledge obtained and helps build a small core of DOE experts within a company.

The analysis tool used during the training is a wonderful statistical package called JMP. Each participant will obtain a one-year license for this software.

The instructors are Phil Ramsey and Mia Stephens, two of the NHG partners. Both have extensive experience in the field of DOE. They have helped develop a large number of green belts, black belts, and master black belts. And they both have great enthusiasm for the use of statistical methods, and design of experiments in particular, in the area of process and product design and improvement. (For bios, see our About Us page ).

For more information on the DOE training, see the listing under our Public Workshops tab.