AUGMENTING A FACTORIAL EXPERIMENT WHEN ONE FACTOR IS AN UNCONTROLLABLE RANDOM VARIABLE - A CASE-STUDY

Process engineers at an automobile assembly plant wanted to develop a model for on-line process control in an assembly line. During an exper iment to estimate the model, an uncontrollable variable was discovered to have a significant effect, but its values in the experiment were n ot representative of values in actual production. Through this case st udy, we present design criteria and assess sampling strategies for aug menting a factorial experiment to incorporate an additional factor tha t was found to be an uncontrollable independent normally distributed r andom variable. Our results demonstrate the potential improvement in o n-line process control when experiments are augmented using these desi gn criteria.