Capability analysis, with its formulas and the confusion around Ppk, Cpk, Cpm, and other indices, is often perceived as far too difficult, complex, and challenging to consider utilizing. Good reasons for undertaking the analysis are sometimes not considered, in the light of this perceived complexity.
In fact, good reasons for using capability analysis not only exist, but provide compelling argument for utilizing this statistical tool. And it’s really not that hard.
Capability analysis offers a way to compare a process to a customer’s requirements, coming up with a score that facilitates communication with the customer to indicate how well a process is able to meet these requirements. Since the language and terminology surrounding capability analysis are consistent and generally agreed upon, discussions can be effective, and can initiate continued progress with meeting and surpassing the customer’s requirements.
In capability analysis, these customer requirements are expressed as specifications, not to be confused with control limits. Some examples of customer requirements (specifications):
- Each item should weigh at least 12 ounces and no more than 12.01 ounces
- Thickness should be 2 mm, plus or minus .1 mm
- We will accept no more than two visual defects per ten meters of fabric.
Specifications can take many forms, as these examples illustrate. To calculate a capability score, the measurement system must measure and track values related to customer specifications. In the examples, measurement systems need to track weight, thickness, and number of visual defects, respectively, in order to come up with a capability score. There is a small collection of statistics that help to describe a capability score. The most common ones are called Cpk, Cp, Ppk, and Pp. Each of these statistics, known as a capability index, is the result of simple math that compares data from the process with the customer specification.
For Cpk and Ppk, the higher the number, the better the process is at meeting customer requirements. Lower values indicate that the process might not be capable of producing most of its output within the customer requirements (specifications).
Two flavors of capability analysis represent actual and potential.
These statistics indicate potential capability:
Cpk, Cp, Cr.
They indicate the best that the system can potentially do, relative to customer requirements.
These statistics indicate actual capability:
Ppk, Pp, Pr.
They reflect real output from a process, compared to customer requirements.
Capability analysis may also indicate the percentage of output that falls outside customer specifications. Ideally, this will be a very low number, since anything outside specification limits represents extra time, cost, and other resources.
It’s really not that hard, as you can see. And here’s why it’s important.
- Capability analysis indicates how well a process can be expected to meet customer requirements.
- Capability analysis offers an easy way to assess quality, and is part of the panoply of quality management tools that support improvement efforts.
- Using software, capability analysis is fast and easy, without demanding knowledge of formulas or indices.
- Capability analysis offers a common voice to facilitate communication between customer and supplier.
Before undertaking capability analysis, a control chart of the measured values should show that the process is in control, or stable. If a process is not stable, capability analysis is premature, and the results will not be reliable. Another assumption about the data that must be made prior to pursuing capability analysis, is that it comes from a normal distribution. If this cannot be verified, adjustments to capability analysis must be made.
These statistics can be computed by hand, and you might want to walk through them…once. After that, the job is better left to SQCpack, so you can focus on process, thereby reducing variation and improving ways to meet customer requirements. SQCpack offers the easiest SPC solution for helping your organization utilize the power of data analysis to help reduce variability, enhance productivity, and improve profitability. It produces the control charts that are essential to capability analysis and provides a solution to the challenge of understanding your processes.