Statistical Methods Applied to a Semiconductor Manufacturing Process

5 Summary

Statistical methods applied to semiconductor manufacturing process are normally viewed. An example with L₁₆(2¹⁵) orthogonal array is presented including split-unit design and estimation of one missing value. The result of this experiment was extraordinarily effective to improve the yield beyond expectation. Dynamic and statistical characteristics about consuming time in the manufacturing process are studied by computer simulation, in particular, under the condition that the working time is larger than arrival time interval with gamma random variables. The study will be useful to assure quality in manufacturing process for preventing quality deterioration in accordance with time.