Design of experiment is the best way to optimize a process at minimal cost

The authors deal with offline statistical control techniques which optimize a process and highlight the use of statistically designed experiments. The authors also describe the advantage of offline quality control over an online quality control mechanism. They define a systematic road map to carry out a designed experiment. The design of experiment (DOE) method not only defines the optimum window of operation but also the optimum parameters in those windows. In DOE the variables affecting the process are identified through a brain-storming session. Once the variables are identified, different levels of each variable are selected. The objective is to find the best possible way to optimize the manufacturing process, which is only possible if all variables with different levels are tested with each other in a controlled fashion. The inclusion of DOE in a corporate culture is discussed. A case study performed on a wire bond process where this process has been successfully implemented is described