The team’s objective is to concentrate on the process map and narrow the process input variables down to the vital few variables (VFVs), because these VFVs have the greatest impact on the early failure rate. However, the data for the historical failures was limited and incomplete, which prevented the team from understanding the true root-cause contributing to early life failures. To cope with this issue, the team used in-house testing beds and performed controlled experiments (e.g. DOE) to determine which ones are the actual VFVs that are responsible for the improvement of the overall reliability performance. Two hypotheses are created as the objective for the controlled experiment.