Design of Experiments


Design of Experiments

This chapter uses the repair shop example in Chapter 2: Overview of SAS Simulation Studio, to demonstrate how you can use JMP software to generate experimental designs for a Simulation Studio model. One of the goals in that example is to ease the bottleneck at the quality control station. Suppose you have the option of adding additional workers at the service desk, repair desk, and quality control station so that each station can have one, two, or three workers. These are the factors of your experiment. The responses you could monitor are the average utilizations at all three stations (to make sure workers are not idle or overworked), the number of fixed parts, and the average waiting time at each of the three stations. Since you now have three factors, each defined at three levels, you might want to generate an experimental design (such as a full or fractional factorial design) to guide your simulation runs. This is a more efficient way to explore the effects of different parameters on your model responses than just randomly selecting combinations of your factor values to try. You can do this with the Simulation Studio Experiment window and JMP software.