Make Valid Control Chart and Subgroup Assumptions

Control Charts Subgrouping At A QS 9000 Automobile Supplier
The second example occurred at a QS 9000 supplier to an automobile manufacturer. As per the documented procedures of the company, the organization was required to send detailed process capability reports with a particular machined component. The organization manufactured the machined components and then sent all the machined components in a bin to the final inspection department. The final inspection departments then took five as the subgroup and inspected the entire lot. At the end of the inspection cycle, they plotted an X Bar R chart and the process capability indexes. The automobile company receiving this material was satisfied that the supplier was complying to the requirements.

What was wrong with the application described above?

The organization had only learned the application of statistical process control (SPC) and control charts, without fully understanding the concepts and underlying assumptions. The key requirement for the application of SPC is the conservation of time sequence during the application of the charts. The subgroup assumption violated this requirement, as the subgroups were made at the end inspection process. Since the time sequence is not preserved, the assumption that the variability within the subgroup is only due to common causes cannot be validated.

The first mistake that can be discovered through common sense is that the organization is only attempting to comply with documentation requirements. They gained no knowledge of the underlying process variability and would have tremendous difficulty trying to improve it. It is application of statistical techniques such as these that dilute the essence of good standards like the QS 9000 and ISO 9000.

Next Page > Calculating Process Capability Without Control Charting