Below is a model that I’ve put together to summarize the various pieces of each stage. I’ve intentionally not linked to any of the improvement and control elements. The improvement methodologies you seek will be largely determined by the steps before & the experiments you choose to use will also vary heavily. However, suffice to say, the key thing here is to validate the relationship between your inputs (X) and outputs (Y), Once validated, you can implement appropriate improvement actions. You then need to create a control and monitoring plan to ensure that the improvements you made continue to achieve the required performance levels and continue to deliver value to your business and customers.
Define & Measure
- Define Organization strategy for Six Sigma
- Problem statement, objective statement & key metric definition
- Define specification limits (VOC)
- Understand current process with As-Is models and value stream maps (VSM)
- Fishbone diagramming & FMEA
- Standard deviation & short term variance calculation
- Voice of the customer & X-Y Matrix
- Yield & Z-Score calculations
- Calculating the capability index
- Find the important X’s and sample data + Multi-Vari charts
- Calculate confidence scores
- Calculate the correlation coefficient (how tightly is your data correlated)
As mentioned above, the improve and control sections have not been included here. The results of your experiments will determine the actions you take & it is therefore too broad of a subject to cover through Netshock posts. Once you reach the control stage, you’ll require a method of measuring the X’s and you’ll need to implement a review process for those measures. If the variables start to slip beyond acceptable limits, we can put remedial actions in place to bring them back in-line with business and customer expectations.