Process Capability analysis and Baselining

In the Measure phase, we are learning about Process Stability, Process is Capable or not, and finally Baselining.

Baselining: Discrete data:

Unit:  It is an item being processed. for example, a drawing, application, Request for Service.

Opportunity: is any event/characteristic/attribute that is essential and measured from a customer's perspective. It provides a chance of not meeting a customer's requirement.

for example each of the 10 attributes on the trade draft form.

Defect: is an event of non-conformity to the customer requirements. for example, any of the 10 attributes incorrectly drafted.

Defective: is a Unit with one or more defects. for example, a Trade draft with one or more of the 10 key attributes incorrectly drafted. 20% defectives if 20 out of 100 Trades had at least one defect (at least 1 key attribute incorrectly drafted)

example 1

example 2

DPU = 18/1000 = 0.018 (1.8%)

DPO = 18/(1000 x 5) = 0.0036

DPMO = 0.0036 x 10,00,000 =  3600

Sigma level = 4.15

example 3

Say, 500 opportunities, # of Defect oppurtunies per unit= 2, Defetcs = 20.

DPU = 20/500= 0.04 (4%)

DPO = 20/(500 x 2) = 0.02

DPMO = 0.02 x 10,00,000 =  20,000

Sigma level = 3.55

------------------------------------------------------------------------------------------------------------------------

Baselining: Continuous data:

Voice of Customer defines Specification limits (LSL, USL)

Voice of Process defines Control limits (LCL, UCL)

Process Capability (Cp) is a statistical measurement of a process’s ability to produce parts to the tolerance range. 

Reducing variation in the process is the most desirable way of improving Process Capability.

Cp= Tolerance range / Process Spread

Cp = (USL - LSL) / 6 Std.Dev.

Cp = (33 - 27) / 6 = 1.0

If the data is Stable and Normal. We do Capability analysis.

Process Capability (Cpk) is a statistical measurement of a process’s ability to meet the target (center), or in other words, it describes how well the process mean is close to the specification target.

Cpk = Minimum (CpU , CpL)

example1

CpU = (USL - mean)/ 3 Std. dev.

CpU = (33-33) / 3 = 0

CpL = (mean - LSL)/ 3 Std. Dev.

CpL = (33-27)/3 = 2

Cpk = Minimum (0,2) 

Cpk =  0

example2

CpU = (USL - mean)/ 3 Std. dev.

CpU = (33-33) / 6 = 0

CpL = (mean - LSL)/ 3 Std. Dev.

CpL = (33-27)/6 = 1

Cpk = Minimum (0,1) 

Cpk =  0

So, Finally

Ppk is used to assess the long-term, overall variability, whereas

Cpk is the capability index for short-term, potential variability.

The higher the Cp and Cpk values are, the better the process

Let's Connect on Linkedin https://www.linkedin.com/in/negi-deepak/