Statistical Process Control (SPC) is used for quality control to ensure that the process produces within customer requirements.
With SPC you can detect errors early and prevent problems, rather than correct problems as they occur.
SPC also helps you to respond properly to measurement result and to address stability and capability.
In addition to reducing errors and quality costs, SPC will increase productivity because it is less likely that the finished product will need to be reworked or scrapped.
An SPC chart is designed like a Run Chart with control limits that distinguish normal and special variation. It is important to distinguish normal and special variation because they require different approaches when reducing variation.
Here's an example: Let's say that scrap at a manufacturing company arise from special cause variation; you can look for a specific cause (perform a root cause analysis). If scrap is due to normal variation, it is not possible to identify one special cause; the variation is part of the natural variation of the process. Normal variation is the result of random / natural variation of all variables that affect the process.
All processes have variation!
If a machine produces scrap because the normal variation is too wide, you have several options:
- You can invest in a machine that is more precise.
- You can choose to keep the machine and include the expected scrap cost in the product calculation to get the right price for the product.
- You can change the process to make the normal variation of the machine acceptable.
It is important to determine normal variation of machines to prevent spending time reducing variation that can not be reduced. It doesn't help to tune and adjust a machine that has too high normal variation. All machines have natural variation depending on how it is designed and mounted; By tuning and adjusting a machine with undesired normal variation it's likely to worsen the situation. Here is a video about application of statistical process control:
Other examples of SPC applications are:
- Preventative Maintenance: Equipment wear will gradually increase variation. SPC will reveal increased variation of measurements and notify maintenance needs. An example is O-rings used for testing capillaries at a pharmaceutical company: When an O-ring starts to wear, it results in greater variation in the measurements. SPC can decide when it is time to change the O-ring. Dirty equipment can also cause greater variation of the measurements, indicating that it's time to clean! Operators with SPC knowledge can identify measurements that can predict maintenance need.
- Determining Capability: Is the process able to deliver on the customer's requirements?
- Assessing Stability: is the process predictable and stable over time?
- Compare machines and equipment for continuous improvement and to make the right decision when making new investments
- Evaluate the effect of improvement work: has the average and / or variation changed?