Workshop DoE Tuesday, June 30, 2009

Time		Hall Dublin
		DESIGN OF EXPERIMENTS (DoE) In the Research- and Development departments it is necessary to get reproducible results in a short time, because tests will cost time and money. Exactly planned and evaluated tests are the key to quick marketability and high quality. To minimize the number of tests the tool of Design of Experiments (DoE) can be used. Condition precedent for successful application is the understanding of the method and the correct interpretation of the results.
09.00 -09.30		INTRODUCTORY REMARKS MOTIVATION D. Schramm, Cooper Standard Automotive, Lindau (D)Examples: Mischen Injection Molding Extrusion Compounding
09.30 -10.30		BASIC PRINCIPLES / METHODS Th. Wember, P & P Informationstechnologie GmbH, Willich (D) Systematic Model Development Possibility to Plan Development Robust Statistic Models Minimization of Efforts Statistical Noise Experimental Room / Operating Area Math Model, Confidence Intervall Experimental Design (fractional factorial, d-optimal Designs)
10.30 -11.00		BREAK
11.00 -13.00		DISCUSSION OF EXAMPLES D. Schramm, Cooper Standard Automotive, Lindau (D); Th. Wember, P & P Informationstechnologie GmbH, Willich (D)
13.00 -14.30		BREAK
14.30 -16.30		INVESTIGATION OF HISTORICAL DATA Th. Wember, P & P Informationstechnologie GmbH, Willich (D) Compound Database Production Data Methods Regression Neuronal Nets Decision Trees

Time