Visual Six Sigma: Making Data Analysis Lean

6.8. Utilizing Knowledge

In a formal DMAIC project, the utilization of knowledge begins in the Improve Phase and continues into the Control phase. As part of its Control Phase activities, the team prepares a comprehensive control plan for the anodize process. The plan includes specification of the optimum settings for the five Xs, as well as the new protocol for controlling the variation of these variables. The control plan also specifies the use of statistical process control to monitor the Xs, the four Ys, and the project KPI, process yield. Recall that the project goal was to improve the anodize process yield from 19 percent to a minimum of 90 percent, and to sustain that improvement.

About four months after the new process settings and controls are implemented, Sean and his team collect the associated data, including the final yield numbers. They add the yield values to a data table that contains yields for the initial 60-lot baseline period (BaselineYieldAll.jmp).

The team decides to continue to use an individual measurement chart to monitor process yield. Recall that although the yield measure is truly a proportion defective, Sean had reason to choose an individual measurement chart, rather than a p chart, to monitor this proportion. Sean makes BaselineYieldAll.jmp the active window. He selects Graph > Control Chart > IR and populates the launch dialog as shown in Exhibit 6.79. The column Phase is used to generate control limits for each portion of the data (Before and After). ...