Learn more about the Control Plan

So it is official then. The D, M, A and I stages have worked as per plan and you have moved into the Control Stage of a Six Sigma DMAIC cycle. By the end of the Improve stage, you would have validated your improvements and post identifying that the improvements did work, you now move into the Control stage trying to first establish if your process has reduced variability and if it is able to maintain a state of statistical control on the reduced variability. Once you can prove statistical control in revised conditions, you can then move on to the Control Plan document.

What is a Control Plan?
Now, a lot of people treat the Six Sigma Control Plan as a long term document, which allows transitioning of the information from one team to another, supporting the theory that Six Sigma projects ideally should be long term projects. Nothing wrong with that! I though prefer using the Control Plan as a documentary summary of my key process characteristics. It is in this excel spread-sheet that I will document what has been measured in the process and how much the measure really was.

Pre-conditions
One of the important pre-conditions for Control Plan is that the process must be operating in stable conditions. That means your Control Chart should show no special causes of variation. For example, a Control chart is plotted on sample data for Average Handle time

As you can see the process is working in stable conditions for sample data collected on subgroup size of 2 plotting a Mean-Range chart for 11 days. We can thus move ahead and collect further samples and work to plot our Control Plan.

The Control Plan Template
Let us first explain each of the columns in a typical Control Plan and then work to update a Control Plan with the help of a live example.
a)      Update the Critical to Quality Characteristic (CTQ).
b)      Under the third column titled Significant Characteristic Description, describe the CTQ characteristic you would measure.
c)       If you are measuring continuous data, specify either X bar – R, or X bar – S or I – MR chart, or if you are measuring discrete data, specify either p, np, u or c chart.
d)      Under Chart champion and chart location, write the name of the individual who claims ownership for the chart and which part of the office premises has the chart readings been taken from.
e)      The measurement method clearly explains how the measurements were recorded. You could document your stepwise procedure here.
f)       The reaction plan would document things or activities that would be done should an unusual cause be identified.
g)      The Sampling Plan section would contain more details on how the sampling activity would be done.
h)      Finally, under stability conditions write Yes if process is stable and No if process is not stable. Please note that the truth needs to be recorded here. If the process is unstable in reduced variability and it shows a high degree of capability, you can still afford writing a No here.
i)        Finally, once all is done, write the Cp and Cpk calculated in the Control Phase under reduced variability conditions. Good.

Now that you have understood the basics of a Control Plan, let us see how a sample control plan looks like. Please note that this Control Plan is updated with the Average Handle Time metric in mind.

Every entry in the Control Plan in a sense is self-explanatory.

Obvious – The Plan needs to be updated only for CTQ metrics, as you want to know how are the CTQ metrics performing as on date and going forward, how would the teams measure and report the performance of the CTQ metrics.
Again – The idea is to ensure that Six Sigma efforts remain sustained and succeeding teams manage to mirror exactly what was done by the team that brought the project to the Control Phase.

Caution
1)      Control Plan should always be updated only after you consider the stability of the process. You measure stability in the Measure Phase first. Here you just do the same process all over again, this time under reduced variability conditions.
2)      The type of data you would measure should ideally be continuous, but if your CTQ is discrete or attribute, don’t lose sleep over it. You do have a format for updating the Control Plan for Attribute data as well.
3)      Post the Control plan; make sure to re-validate the Measurement System. This needs to be done because you want to be sure of the granularity and the efficiency of your measurement system to report data for your process, which is now working at a reduced variability.

Summary
A Control Plan is often considered one of the mandatory activities and documents that need to be done in the Control Phase of a Six Sigma DMAIC project cycle. With any other activity in Six Sigma, there are certain protocols, rules and regulations one needs to follow while updating the Control Plan.

Related Reading:

• Variation, shift in mean and treatment of variation
• The Metrics story
• Six Sigma Green Belt Certification Training