Concise Process Improvement Methods

Abstract

This thesis reviews two methodologies for process improvement; Six Sigma and the Shainin System. A strengthened methodology is developed following the 12-step Six Sigma DMAIC cycle with an added Shainin loop in the Analyse phase to narrow down sources of variation. This Hybrid Six Sigma framework is used to develop a sampling strategy known as the Process Variation Diagnostic Tool (PVDT).
The PVDT allows a Gage R&R and a Provisional Process Capability study to be carried out with just 20 samples. It also allows for an IsoplotSM and a Shainin Multi-Vari study. The method was then reviewed in three different industrial situations to demonstrate its effectiveness. Applying the PVDT allowed the project teams involved to quickly produce Gage R&R and Provisional Process Capability Studies. It reduced samples required from the combined 110 measurements from 60 products typically taken in industry to 60 measurements from 20 products. A significant advantage was the ability to extract a Shainin Multi-Vari Study from measurements taken for the PVDT. This technique allowed the project team the ability to categorise the most significant families of variation. From these case studies it can be seen that at the border of the Measure/Analyse phase in Six Sigma the proposed PVDT offers an efficient method of collecting Six Sigma metrics and steering the course of an improvement project.
A teaching vehicle known as the PIM game is introduced to demonstrate and facilitate the teaching of a number Process improvement Method. These methods are directly related to Six Sigma and Shainin methods developed in this thesis. The historical development and need for a teaching game are discussed.
Finally the thesis proposes a new method of destructive measurement system analysis (MSA). An industrial problem is used to benchmark the method against a traditional approach to destructive MSA. The project highlights when there is a second non-destructive test a conservative estimate of Gage R&R can be determined for destructive test equipment.