ntroductiontodoe(诺基亚公司doe设计实验培训教材).pptVIP

FN: Graphical Design of ExperimentsIntroduction and Overview The Focus of the Improve PhaseY = f(X) At this point in the Six Sigma process, we should have: An understanding of the magnitude and scope of the problem, including defects per unit (DPU) and Cost of Poor Quality (COPQ) Concrete goals and objectives, including projected DPU reductions and cost savings. A Y variable with a reliable measurement system A capability analysis of Y with an understanding of the nature of process deficiencies—is there a problem with mean, variance, distribution shape, or some combination of the three? A short list of potential X variables obtained through analysis of process demographics, multi-vari analysis, and other graphical and statistical techniques Experiments An “experiment” is... Any testing in which the inputs are either controlled or directly manipulated according to a plan A “designed experiment”... Adds structure to the plan and a specific methodology to the analysis of the results Experimentation Strategy Define the problem. Establish the experimental objective. Select the output (response) variables. Select the input (independent) variables. Choose the levels for the input variables. Select the experimental design. Execute the experiment and collect data. Analyze the data. Draw statistical conclusions. Replicate or validate the experimental results. Draw practical solutions. Implement the solutions. Hierarchy of Experimentation One factor at a time Screening designs (Fractional-Factorials) Characterization designs (Full-Factorials) Optimization designs (response surface methods) Process Design ProblemHow to Launch a Ball a Specific Distance Executing the Experiment Strategy Evaluate the ResponseMeasurement System There are a number of potential measurement systems for distance. Break into groups and do a quick measurement system evaluation. Establish a standard measurement methodology to be used in future experimentation. A suggested measurement method is