Cpk, CPU, and CPL

Measures of potential process capability, calculated with data from the subgroups in your study. They measure the distance between the process average and the specification limits, compared to the process spread:

· CPL measures how close the process mean is running to the lower specification limit

· CPU measures how close the process mean is running to the upper specification limit

· Cpk equals the lesser of CPU and CPL.

If Cpk, CPU, and CPL are equal, the process is centered at the exact midpoint of the specification limits. Compare your values to benchmarks to determine whether to improve your process; many industries use benchmark values of 1.33. The graphs below illustrate various situations:

Cpk = CPL = CPU = 2.76
	This graph shows an ideal case. The value of all three indices is 2.76, which indicates that the distance between the process average and the each specification limit is 2.76 times the one-sided process spread.
Cpk = min {CPL = 0.96, CPU = 4.56} = 0.96
	When CPL < CPU, the process is not centered. It is more likely to produce defective units that violate the lower specification limit.
Cpk = min {CPL=4.56, CPU = 0.96} = 0.96
	When CPL > CPU, the process is not centered. It is more likely to produce defective units that violate the upper specification limit.

A disadvantage of the Cpk index is that it only represents one side of the process curve, and tells you nothing about the other extreme. For example, the two graphs below display processes with identical Cpk values. However, one violates both specification limits, and the other only violates the lower specification limit. To avoid mistakes, consider all indices along with a graphical summary to draw meaningful conclusions.

Cpk = CPL = CPU = 0.96		Cpk = min {CPL = 0.96, CPU = 4.56} = 0.96

Note:

These capability indices are analogous to Ppk, PPL, and PPU, which assess the same properties using overall process standard deviation instead of considering subgroups.