A control chart is one of many process improvement techniques that is used to monitor a process variable over time and if it is in control. The variable being looked at can be from any company or industry like manufacturing, service, healthcare etc.
It was invented by Walter Shehart in 1924 for manufacturing, and later extended by W. Edward Deming to be used for quality improvements in all departments of an organisation.
When used for Continuous Improvement, there are always additional process improvement tools that should be used along with it. The key to using the Control Chart efficiently is to understand the variation you are dealing with.
Types of Variation in Control Chart
Variation comes from two sources; it is either common or special causes. When using driving to work as an example, then it might take around 30 min on average. Some days it will take a little extra time and some days less so the time is usually between 25-35 minutes. The 10 variation is due to common causes like amount of green and red lights on the way, traffic ect. These things are predictable and always there as part of the process, but as long as you plan with at least 35 min to drive to work, you will get there in time.
Now let’s imagine you get a flat tire on the way, this will definitely increase the time above 35 minutes maximum you have with common variation. It might increase the trip to an hour and is seen as a special cause of variation. Something happened outside the normal process, which was not supposed to happen. Special causes are unpredictable events.
Control Chart for Process Improvement
It is important to know the type of variation in your process, as the actions for improving the process is taken based on that.
To eliminate special causes, an 8D Problem Solving approach could be used to identify the root cause and prevent it from ever happening again. If on the other hand only common causes are present in the process, the process itself needs to be changed for improvements to be achieved.
How to Generate a Control Chart
If looking at the example used earlier with the process of getting to work, then it could look something like this.
The process variable is the time to get to work seen on the left axis, an on the horizontal axis is the days driving to work.
After enough point has been plotted the process average can be calculated as well as the upper and lower control limit (UCL and LCL). These values are determined by the process and sampling used to calculate them, also the way they are calculated depends on the type of control chart used.
If a process is in control, only common causes are present, and all data points are expected to fall between the UCL and LCL.
If you are interested in Control Charts and implementing them in your organisation, then our Statistical Process Control (SPC) Training might be interesting for you.
Alternatively you can continue to our quality training page for an overview of the training we provide.Go to Quality Training