DESIGN OF EXPERIMENT In cost effective development of a process, an in-depth analysis is required. The traditional multi-variable approach to obtaining the data for analysis is a full factorial experimental design. Full factorial is not a viable solution path in these days of simultaneous engineering to bring products and processes to market faster and cheaper. For 4 controllable input factors a full factorial experimental design would require 256 experiments. A 2-level factorial design would require 16 experiments for four variables. Depending on required sample sizes the costs and time get out of control very quickly with a full factorial approach.

In addition to managing the business side of experimental design Automation Association utilizes professional, highly experienced, statisticians who will develop the protocol for your experimental design and do your data analysis. It is amazing the insight these professionals have in understanding a process through the lens of statistics. The other thing they bring to the table is true independence. I can’t tell you the number of times I felt we had the inputs and outputs figured out and I get the call that we have nothing more than a random number generator running in the back room! On the other hand, I can tell you exactly our success rate of launching a new process when the statistician says we are stable and the control handles are understood. 100%!

Site Design by Leverage5 Privacy Policy