Chapter 6. Improving the Quality of Anodized Parts

Even though defect reduction is often viewed as the overarching goal of a Six Sigma project, optimization is just as important. In this case study, we follow the efforts of a Six Sigma team working to both optimize an existing manufacturing process and reduce the number of defects it produces. The company is Components Inc., a manufacturer of aluminum components used in high-end audio equipment. The surfaces of these components are anodized and then dyed to produce a visually smooth, rich, black surface.

Unfortunately, Components Inc. currently has a significant problem with discoloration of their black components. Lot yields in manufacturing are extremely low, causing rework and compromising on-time delivery. Failure to fix this problem could cost Components Inc. its major customer and result in the loss of more than a million dollars.

Management assembles a Six Sigma project team, under the leadership of Sean Cargill, a black belt, whose charge it is to improve the yield of the anodizing process. This case study follows Sean and his team as they work through all of the steps of the Visual Six Sigma Data Analysis Process.

The team members make extensive use of dynamic visualization in achieving their goal. They identify four Ys and five Hot Xs that relate to yield. Measurement System Analysis (MSA) studies are conducted and the results are explored using variability charts. Specification limits for the four Ys are determined using EDA and visualization tools that include Distribution with dynamic linking, Graph Builder, Scatterplot Matrix, and Scatterplot 3D displays.

To understand how the yield can be increased, the team members design an experiment that has to satisfy various constraints. They fit models to the four Ys, simultaneously optimize the Xs using the prediction profiler, and then conduct simulations to estimate capability at the new optimal settings. This predicted capability is explored using a goal plot.

The new settings for the Xs are implemented in production and the project moves into its Control phase. A control chart of post-implementation data shows that the process is stable and highly capable, delivering predictable performance and high yields. The project is deemed a resounding success. The increased predictability of supply means that Components Inc. is able to retain its key customer, and the increased yield reduces annual scrap and rework costs by more than a million dollars.

The platforms and options used by Sean and his team are listed in Exhibit 6.1. Their data sets are available at http://support.sas.com/visualsixsigma. To share the excitement of this project team's journey, join Sean and his teammates as they tackle this ambitious project.

Figure 6.1. Platforms and Options Illustrated in This Case Study