Lean Six Sigma Green Belt

Certification Course
Process and Performance Capability
 Learning Objectives

By the end of this lesson, you will be able to:

Identify a stable or unstable process

Verify process stability and normality in a process

Interpret the results of a graph on natural process limits and

specification limits

Calculate the Cpk and Cp and Cpm and Ppm of a process

Interpret capability indices

 Introduction

Sales Chart

How can we
summarize the current
performance of our
process to determine We need capability
if the product output is analysis to determine this.
meeting customers’ Then, we can decide what
needs? steps to take next.
Process and Performance Capability
 Activities in the Measure Phase

Test for stability

• MSA
• Collection of data
• Statistical calculations • Changes should not be
• Check for accuracy and made to an unstable
validity process

Determine if the process is

stable or not
 Activities in the Measure Phase

?
Why does a process become
!
unstable?
A process becomes unstable due to special
cause of variation.

Multiple special causes of variation lead to

instability.

A single special cause leads to an out-of-

control condition.
 Process Stability Studies

Variation can be due to two types of causes:

Common Cause Variation (CCV)

Special Cause Variation

(SCV)
 Process Stability Studies

Special Cause Variation

Common Cause Variation (CCV)
(SCV)

• Include factors external to and not

always acting on the process
• Include many sources of variation
• Sporadic in nature
within a process or inherent to it
• Contribute to instability to a process
• Have a stable and repeatable
output, which makes the output
distribution over time
unpredictable
• Contribute to a state of statistical
• May result in defects and have to be
control where the output is
eliminated
predictable within a range
• If identified, they point to the need
for root cause analysis
 CCV and SCV

Example of CCV

Difference in temperature inside

the oven due to oven design

400⁰F
 CCV and SCV

Example of SCV

Difference in temperature inside the oven

due to external factors
 Control Charts

If all the data points are within the UCL and LCL (red lines), the process could be stable.
 Run Charts

If p-values for any of the last 4 values provided in the chart are less than 0.05, the process has
special causes of variation, and the chances of the process being unstable are high.
 Verifying Process Stability and Normality

SCV present Output unstable Process not under control

CCV present Output stable Process under control

 Normality

Normality is the condition of a process that follows a normal distribution.

If the P value is greater than 0.05, the data set could be described as normally distributed.
 Process Capability Analysis

Customer needs
Process
and expectations

Process capability is how well the process is meeting customer needs and
expectations.

It is a metric that reflects only the common cause of variation.

Interpretation of Data
 Natural Process Limits vs. Specification Limits

Natural Process Limits Specification Limits

Indicators of process variation Targets set for the process

Voice of the process Voice of the customer

Based on past performance Based on customer requirements

Real-time values Intended result

Derived from process data Defined by the customer

Consists of Upper Control Limit (UCL) and Consist of Upper Specification Limit (USL)
Lower Control Limit (LCL) and Lower Specification Limit (LSL)
 Natural Process Limits and Specification Limits

LCL UCL

If the control limits lie within the specification limits, the process could be considered capable. Conversely, if
the specification limits lie within the control limits, the process will not meet customer requirements.
 Actions after Comparing Natural Limits and Spec Limits

Condition Action

Process limits fall within

No action required
specification limits

Adjust the process

Compare natural Process spread and
centering to bring the
process limits with specification spread are
batch within specification
specification limits approximately the same
limits

Reduce variability by
Process limits fall outside
partitioning and targeting
specification limits
the largest offender
 Process Capability Studies

Process capability is the actual variation in the process specification.

Plot and analyze the

Plan for data collection Collect data
results

The same data used for stability and normality analysis can also be used for capability.
Process Capability Index

Measures performance

Is the inherent variability of a

characteristic of a process or a
product

Is the indicator of the capability

of a process
 Process Capability Index

Upper specification limit − Lower specification limit

Process capability (CP) =
6σ

OR

USL − LSL
Process capability CP =
6σ
 Process Capability Index

The formula assumes the process is perfectly centered between the upper and lower specification limits.
It does not reflect how the process is actually performing in relation to the specification limits.
 Process Capability Index

Process Capability (Cpk) was developed to objectively measure the degree to which a
process meets or does not meet customer requirements.

𝐶𝑝𝑘 = 𝑀𝐼𝑁(𝐶𝑝𝑘𝑙 , 𝐶𝑝𝑘𝑢 )

𝑋ത − LSL
𝐶𝑝𝑙 =
3σ

USL − 𝑋ത
𝐶𝑝𝑢 =
3σ
 Comparison of Process Capability Indices

Cpk < Cp
Cpk < Cp Cpk = Cp
Cpk is negative
Mean is not centered between Mean is centered between
Mean exceeds specification limits
specification limits specification limits

Cpk cannot be larger than Cp because Cp represents the maximum capability with the current process
variation.
 Interpretation of Process Capability Indices

Sigma Level Process Capability

Not capable 2 0.67

Marginally capable 3 1.00

4 1.33

Capable 5 1.67

6 2.00
 Interpretation of Process Capability Indices

Process
Sigma Level
Capability
2 0.67

3 1.00

4 1.33

5 1.66

6 2.00
 Process Capability: Example

?
• Mean diameter = 3.02 centimeters
• Standard deviation = 0.02 centimeters
• Lower engineering specification limits = 2.95 inches
• Upper engineering specification limits = 3.05 inches
• Process is stable
• Diameters are normally distributed

Determine the Cp and Cpk

 Process Capability: Example

!
𝑈𝑆𝐿 − 𝐿𝑆𝐿 3.05 − 2.95 0.10
𝐶𝑃 = = = = 0.833
6𝜎 6(0.02) 0.12

USL − 𝑋ത 3.05 − 3.02 0.03

𝑋ത − LSL 3.02 − 2.95 0.7 𝐶𝑝𝑘𝑢 = = = = 0.5
𝐶𝑝𝑘𝑙 = = = = 1.17 3σ 3(0.02) 0.06
3σ 3(0.02) 0.06

𝐶𝑝𝑘 = 𝑀𝐼𝑁(𝐶𝑝𝑘𝑙 , 𝐶𝑝𝑘𝑢 ) = 𝟎. 𝟓

 Interpretation of the Example

The process is not The process is not be

Cpk = 0.5 Cp =0.833
capable in regards to capable even if mean
true performance was centered

The process is not centered

between the lower and upper Cp (0.833) > Cpk (0.5)
specification limits

The team must focus on reducing variation and work with

Remedial steps
operations to center the average output.
 Interpretation of the Example

The process is not The process is not be

Cpk = 0.5 Cp =0.833
capable in regards to capable even if mean
true performance was centered

The process is not centered

between the lower and upper Cp (0.833) > Cpk (0.5)
specification limits

If the Cp index shows process is capable but Cpk indicates a lack a capability, then it would mean that the
process variation is small enough to allow for a capable process.
 Identifying Characteristics

Criteria for a characteristic for a process capability study:

✓ Should indicate a key factor in the quality of the product or process

✓ Should influence the value of the characteristic through process adjustments
✓ Operating conditions that affect the characteristic should be defined and controlled
✓ Determined by customer requirements or industry standards

For a Six Sigma project, the characteristic is typically the “Y” or output variable defined in the
charter.
 Identifying Specifications or Tolerances

Process specification or tolerances are defined by:

OR
Industry standards based on Organization in consultation
the customer requirements with the customer
 Process Performance Indices

Process Performance Index verifies if the sample generated from the process is
capable of meeting the customer requirements.

It is only used when process control cannot be evaluated.

Process Performance Indices

Process Performance (Pp)

Process Performance Index

(Ppk)

Process Capability Index

(Ppm/Cpm)
 Process Performance Indices vs. Process Capability

Process Performance Indices Process Capability

Applies to a specific batch of output Applies to all batches of output

Based on sample standard deviation Based on population standard deviation

Long-term capability Short-term capability

 Process Performance Indices (Pp)

USL −LSL
Pp =
6s

Where,
USL = upper specification limit
LSL = lower specification limit
s = natural process variation
 Process Performance Indices (Pp)

Ppk = Min(Ppu, PpL)

Where,
USL−x
Ppu (upper process capability index) =
3s

x−LSL
PpL (lower process capability index) =
3s

Where,
x = process average (തx)
s = sample standard deviation
 Cpm and Ppm

𝑈𝑆𝐿 −𝐿𝑆𝐿 𝑈𝑆𝐿 −𝐿𝑆𝐿

Cpm = 2 2
Ppm =
6 ∗ 𝜎 + 𝜇 −𝑇 6 ∗ 𝑠2+ 𝑋ത −𝑇 2

Where,
USL = upper specification limit
LSL = lower specification limit
μ = process average
σ = process standard deviation
T = target value (typically, the center of tolerance)
ത sample average
𝑋=
 Cpm and Ppm: Interpretation

Improved
Improved
position
Cpm or Ppm

Improved Improved
variation Cpm or Ppm

• As the process variation reduces and/or as the process average reaches the target value, the Cpm or
Ppm index improves or becomes larger.
• The ideal value is 1.33 or greater.
 Process Mean Shift

Process shift ● Reflects how well a process is controlled

(ZST – ZLT) ● Usually a factor of 1.5

● Capability or the potential performance of the process in

Short-term capability
control at any point of time
(ZST)
● Based on the sample collected in the short term

Long-term performance
● Actual performance of the process over time
(ZLT)
 Process Mean Shift: Example

Sample at Time 3
Sample at Time 2
Sample at Time 1

Observations

• A typical process will shift by approximately 1.5 standard deviations.

• Long-term variation is more than the short-term variation.
• The difference between the short and long-term shifts is the Sigma Shift.
• A shift may be due to different people, raw material, wear and tear, time, and so on.
 Process Mean Shift: Implications

Sigma Multiple No Mean Shift Mean Shift ±1σ Mean Shift ±1.5σ
±3σ 2700 DPMO 22800 DPMO 66810 DPMO
99.73% Good 97.72% Good 93.32% Good
±4σ 63 DPM) 1350 DPMO 6210 DPMO
99.9937% Good 99.865% Good 99.379% Good
±5σ 0.57 DPMO 32 DPMO 223 DPMO
99.999943% Good 99.9968% Good 99.9767% Good
±6σ 0.002 DPMO 0.29 DPMO 3.4 DPMO
99.9999998% Good 99.999971% Good 99.99966% Good

Summary of the defect levels at different Sigma multiple values and mean shifts

Observation

The effects of a mean shift become more negligible as the process capability increases. A Six Sigma
process level of defects is not affected much by long-term variation.
 Conversion From Long-term to Short-term Capability

TO KNOW

Short-term Capability Long-term Capability

Zst Short-term data P Subtract 1.5

With data collected

Zlt Long-term data Add 1.5 P
 Process Variations

Short-term Variation Long-term Variation

Attributed to common cause variation Attributed to both common and special

cause variations

Variance inherent in the process Shows variation within and between sub-
groups

Shows variation within subgroups Increases due to special causes

Small number of samples collected at short Special causes have to be identified and
intervals corrected for improvement

Captures common cause variation

Common causes are difficult to identify and

correct and may require process redesign
Long-term vs. Short-term Process Variation

Complaint Resolution Time (Hours)

Week 1 Week 2 Week 3
48 50 49
49 48 48
48 36 39
53 50 49
58 50 34
50 62 33
46 45 57
50 47 48
49 51 47
47 44 39

Subgroup size = nweek = 10

 Long-term vs. Short-term Process Variation

1 Average of each week = Total complaints resolved per week/subgroup size 8

St. Dev. Short Term = SST =

2 Grand average = Sum of per week average/3 𝑇𝑜𝑡𝑎𝑙 𝑣𝑎𝑟𝑖𝑎𝑡𝑖𝑜𝑛 𝑤𝑖𝑡ℎ𝑖𝑛 𝑠𝑢𝑏𝑔𝑟𝑜𝑢𝑝𝑠
(𝑇𝑜𝑡𝑎𝑙 𝑠𝑎𝑚𝑝𝑙𝑒 𝑠𝑖𝑧𝑒 − 𝑛𝑢𝑚𝑏𝑒𝑟 𝑜𝑓𝑠𝑢𝑏𝑔𝑟𝑜𝑢𝑝𝑠)

Variation within subgroups for a week = Σ(Week average – x)2, where x is

3 each reading in the week

9 St. Dev. Long Term = SLT =

Variation between subgroups for a week = (Grand average – week average)2
4 ∗ nweek
𝑂𝑣𝑒𝑟𝑎𝑙𝑙 𝑉𝑎𝑟𝑖𝑎𝑡𝑖𝑜𝑛
(𝑇𝑜𝑡𝑎𝑙 𝑠𝑎𝑚𝑝𝑙𝑒 𝑠𝑖𝑧𝑒 −1)

Total variation between subgroups = Σ(variation between subgroups for

5 each week)

Total variation between subgroups = Σ(variation between subgroups for

6 each week)

Overall variation = Total variation within subgroups + Total variation

7 between subgroups
 Long-term vs. Short-term Process Variation

Measure Value
Grand average 47.5
Total variation within subgroups 1023.8
Total variation between subgroups 161.67
Overall variation 1185.5
Standard Deviation short term 6.2
Standard Deviation long term 6.4
 Process Performance

Guideline summary of process performance

Process Less CCV, No SCV Less CCV, High CCV, High CCV,
SCV present No SCV SCV present
Variations Only CCV CCV and SCV CCV CCV and SCV
Variability Less High Moderate to High Very High
Capability High Less Very Less Poor
Possibility of defects Less High Very High Defective Process

In Control, Out of Control, Incapable, Incapable,

Process Condition
Capable Incapable In Control Out of Control
 Attribute Data and Capability

DPU DPMO

DPMO = Defects per opportunity ∗

Defects
DPU =
Total no.of units 1,000,000
 Process Performance: Example

The Quality Control department checks the quality of finished goods by sampling a
batch of 10 items from the produced lot every hour.

It collects the following data over 24 hours.

1 1 1 1 1 1 1 1 1 1 2 2 2 2 2
Hour 1 2 3 4 5 6 7 8 9
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4

Defectives 2 1 0 0 2 3 1 4 5 1 2 0 0 1 2 1 1 1 4 0 0 0 1 2
 Process Performance: Example

𝑝ҧ or DPU is used to calculate process capability.

Total no. of defectives = 34

Subgroup size = 10
Total no. of units = 10 ∗ 24 = 240
34
DPU = = 0.0142
240

DPMO = DPU ∗ 1,000,000 = 141666.66

Observation

The process is currently working at 2.6 σ, which is 86.4% yield.

 Process Capability for Attribute Data

Defectives Defects

σ= 𝑝ഥ (1 − 𝑝)ҧ Τ𝑛 σ=
ഥ
𝑢
; 𝑐ҧ
𝑎

• 𝑝ഥ is used to check process capability • 𝑐ҧ is used when the sample size is

for constant and variable sample constant.
sizes. ഥ is used when the sample size is
• 𝑢
variable.
 Key Takeaways

Variations can be due to Common Cause Variation (CCV) and

Special Cause Variation (SCV).

Process normality is verified using a normal probability plot.

Process capability refers to how well the process meets

customer needs and expectations.

Process shift reflects how well a process is controlled. It is the

difference between short-term and long-term performance.

For attribute data, process capability is determined by the

mean rate of non-conformity.
Knowledge Check
Knowledge
Check
In the Measure Phase, which of these options should be determined first?
1

A. Process Capability

B. Process Stability

C. Normality

D. Sigma Shift
 Knowledge
Check
In the Measure Phase, which of these options should be determined first?
1

A. Process Capability

B. Process Stability

C. Normality

D. Sigma Shift

The correct answer is B

Ensuring that the process only has common cause variation or is stable is the first activity in the Measure Phase.
Knowledge
Check
If Cpk = 1.1 and Cp = 1.33, what can we conclude about the process ?
2

A. Process is neither capable nor centered.

B. Process is capable but is not centered.

C. Process is capable and centered.

D. Process is not capable but is centered.

 Knowledge
Check
If Cpk = 1.1 and Cp = 1.33, what can we conclude about the process ?
2

A. Process is neither capable nor centered.

B. Process is capable but is not centered.

C. Process is capable and centered.

D. Process is not capable but is centered.

The correct answer is B

Since Cp and Cpk are greater than 1, we can say the process is currently marginally capable and because Cpk is not
equal to Cp the process is not centered.
Knowledge
Check
If Cpk = 0.7 and Cp = 1.33, what actions should be taken?
3

A. Process needs to be centered.

B. Variation needs to be reduced.

C. Stability needs to be rechecked.

D. Process performance index needs to be calculated.

 Knowledge
Check
If Cpk = 0.7 and Cp = 1.33, what actions should be taken?
3

A. Process needs to be centered.

B. Variation needs to be reduced.

C. Stability needs to be rechecked.

D. Process performance index needs to be calculated.

The correct answer is A

The process needs to be centered since a Cp value of 1.33 indicates good capability and with Cpk value less than 1
indicates centering is the issue.
Knowledge
Check
If short-term variation is calculated at 3.5σ, what is the expected long-term variation?
4

A. 3.5σ

B. 2.0σ

C. 5.0σ

D. 1.0σ
 Knowledge
Check
If short-term variation is calculated at 3.5σ, what is the expected long-term variation?
4

A. 3.5σ

B. 2.0σ

C. 5.0σ

D. 1.0σ

The correct answer is B

The question requires the use of the 1.5 Sigma shift. Moving short-term to long- term Sigma will mean subtracting 1.5
from 3.5, which is 2.0.
 Lean Six Sigma Activities and Tools: Measure
Activities

❑ Review Project Charter ❑ Process Map Flow Identify Root Causes ❑ Develop SOP’s, Training Plan
❑ ❑ Develop Potential Solutions
❑ Validate High-Level Value ❑ Identify Key Input, Reduce List of Potential and Process Controls
❑ ❑ Evaluate, Select, and
Stream Map and Scope Process and Output Root Causes ❑ Implement Solution and
Optimize Best Solutions
❑ Validate Voice of the Metrics Confirm Root Cause to Ongoing Process
❑ ❑ Develop ‘To-Be’ Process
Customer and Voice of the ❑ Develop Data Collection Output Relationship Measurements
Maps
Business Plan Estimate Impact of Root ❑ Confirm Attainment of
❑ ❑ Develop and Implement
❑ Validate Problem Statement ❑ Validate Measurement Causes on Key Outputs Project Goals
Pilot Solution
and Goals System Prioritize Root Causes ❑ Identify Project Replication
❑ ❑ Implement 5s Program
❑ Validate Financial Benefits ❑ Collect Baseline Data Statistical Analysis Opportunities
❑ ❑ Develop Full Scale
❑ Create Communication Plan ❑ Determine Process Complete Analyze Tollgate ❑ Training
❑ Implementation Plan
❑ Select and Launch Team Capability ❑ Complete Control Tollgate
❑ Cost/Benefit Analysis
❑ Develop Project Schedule ❑ Complete Measure ❑ Transition Project to Process
❑ Complete Improve Tollgate
❑ Complete Define Tollgate Tollgate Owner

Define Measure Analyze Improve Control

❑ Project Charter ❑ Process Mapping ❑ Cause & Effect Matrix ❑ Process Flow ❑ Mistake-Proofing
❑ Voice of the Customer ❑ Data Collection Plan ❑ FMEA Improvement ❑ Standard Operating Procedures
❑ SIPOC Map ❑ Statistical Sampling ❑ Hypothesis Tests ❑ Design of Experiments (SOP’s)
❑ Project Valuation (ROI) ❑ Measurement System ❑ Simple & Multiple (DOE) ❑ Process Control Plans
❑ Stakeholder Analysis Analysis (MSA) Regression ❑ Solution Selection ❑ Visual Process Control Tools
❑ Communication Plan ❑ Gage R&R ❑ ANOVA Matrix ❑ Statistical Process Controls
❑ Effective Meeting Tools ❑ Control Charts ❑ Components of Variation ❑ Piloting (SPC)
❑ Timelines, Milestones, ❑ Histograms ❑ Pugh Matrix ❑ Visual Workplace
and Gantt Charting ❑ Normality Test ❑ Pull System ❑ Total Productive Maintenance
❑ Pareto Analysis ❑ Process Capability ❑ Metrics
Analysis ❑ Team Feedback Session

Tools
 Measure Tollgate Questions

✓ Has a more detailed Process Map been completed to better understand the process and problem, and where in the
process the root causes might reside?
✓ Has the team conducted a process analysis, identifying areas where time and resources are devoted to tasks not critical
to the customer?
✓ Has the team identified the specific input (x), process (x), and output (y) measures needing to be collected for both
effectiveness and efficiency categories (i.e. Quality, Speed, and Cost measures)?
✓ Has an appropriate sample size and sampling frequency been established to ensure valid representation of the process
we’re measuring?
✓ Has the measurement system been checked for repeatability and reproducibility, potentially including training of data
collectors?
✓ Has the team developed and tested data collection forms or check sheets which are easy to use and provide consistent,
complete data?
✓ Has baseline performance and process capability been established? How large is the gap between current performance
and the customer (or project) requirements?
✓ Have key learning(s) to-date required any modification of the Project Charter? If so, have these changes been approved
by the Project Sponsor and the Key Stakeholders?
✓ Have any new risks to project success been identified, added to the Risk Mitigation Plan, and a mitigation strategy put in
place?

Note: With answers to these questions you are now ready to move to the Measure Phase.