Six Sigma Overview

The Classical View of Performance

• Six-Sigma is a philosophy:
• Why isn’t “99% acceptable” good enough...??
– 20,000 lost articles of mail every hour.
– 15 minutes each day of unsafe drinking water.
– 5,000 incorrect surgical procedures per week.
– 4 or more accidents per day at major airports.
– 200,000 wrong drug prescriptions each year.
– 7 hours each month without electricity.

Long-Term Yield

3 Sigma Process Capability 93.32% Historical Standard

4 Sigma Process Capability 99.38% Current Standard

6 Sigma Process Capability 99.99966% World-Class

 History of 6 Sigma
• 6 Sigma manufacturing philosophy came from Motorola
They recognised that sufficient process improvement would not
occur using a conventional approach to quality. It was developed
to help them reduce variation within a process by focusing effort
on improving inputs to a process rather than reacting to outputs.

• The process was failing the customer expectations

– Traditionally, processes aimed for process capability of 3 to 4
sigma (Cpk=1.0 to 1.33 or 93% to 99.3% acceptable)
– The customer received 6200 defective product per million at best
– Processes now aim for 6 sigma (Cpk=2)
– The customer would receive 3.4 defective product per million

On target, minimum process variation

 6 sigma Process Capability
“What is it (CPK)”
• 3 Sigma ( Process capability of 1 CPK )
– if the process (lorry) slightly varies then the scrap or damage will
occur

• 6 Sigma ( Process capability of 2 CPK )

– if the process (lorry) varies, there will be no scrap or damage

Curbs
= required
process tolerances

CPK of 1 CPK of 2
(3 sigma) (6 sigma)
 Understanding Variability
 Variation exists in everything. Even the
best machine cannot make every unit
exactly the same.
 Improved capability, becomes a necessity,
due to the need of :
• improved designs

• lower costs

• better performance

 All of this leads to the need of tighter

tolerances
 This means that the ability to operate to a
tight tolerance, without producing defects
becomes a major advantage
 Improvement methodology

KPIV
Key
Controllable Inputs Process
X1 X2 X3
Quality Input

Characteristics: Variables
Inputs:
Outputs
Raw materials, The Process
components, etc. Y1, Y2, etc.

N1 N2 N3
Uncontrollable Inputs
On target,
minimum process variation
Improvement methodology
DMAIC

Define

Measure

Analyze

Improve

Control
 Improvement methodology
Define

• Define terms of reference (Charter the

•To develop a team charter.
•To define the customers
and their requirements
(CTQ Critical to Quality).
•To map the business
process to be improved
Characteristics
project)
– Team / customer / project charter
– Brain storming
– Mind maps
– Affinity diagrams
– High level Process Maps
– Systematic diagrams / Fault tree
– Business Process Mapping
Product / customers

• Define customer requirements (Voice of

the customer)
– QFD Quality Function Deployment
Importance out of 10
 Define
• Define terms of reference (charting a project)
– What you can deliver to the customer and the support you need
from the customer to facilitate a successful improvement (contract
of engagement)

• Brain storming, Mind maps, Affinity diagrams, High level Process

Maps, Systematic diagrams / Fault tree, Business Process Mapping
– Tools to explore a problem, project or current thinking.
– Tools to group those ideas logically.
– Then define a route map to improvement, the risk involved and
how to mitigate that risk.

• Define customer requirements (Voice of the customer)

– QFD Quality Function Deployment, is a method of defining what
the customer needs, what is critical to there business success &
prioritise objectives to meet the customer need.
 Improvement methodology
• Voice of the process
– Data Collection - 7 quality tools
– Tally charts
Measure – Bar charts
– Pareto
•To measure and – Run charts
understand baseline – Control charts
performance for the – Cause & effect
current process
– Check sheets

– Evaluate measurement systems

• Gauge R&R

• Select measures of performance

– Quality Function Deployment
 Measure
• Voice of the process (7 quality tools)
– Tally charts, Bar charts, Pareto, Run charts, Control charts, Cause
& effect, Check sheets.

• Evaluate measurement systems Gauge R&R

– Every process has variation and measurement system, tools &
cmm are no exception.
– Typical your measurement process needs to be ACCURATE,
REPEATABLE & REPRODUCIBLE to less than 10% of the
tolerance you are trying to measure to & proven to be so.

• Select measures of performance

– QFD Quality Function Deployment is a method of defining what
the customer needs and what is critical to there business success
and prioritising performance measures to support the customers
need.
 Improvement methodology
• Investigate source of variation
(Special cause / Common causes)

– Stratification of data to get information

– Cause & effect
– CP & CPK
Analyze
– Fault tree
– Contingence analysis
Seek to:- – FMEA (Failure Mode Effect Analysis)
Prioritise
Understand – Design of experiments (DOE)
Clues – Detailed process maps
Causes
Monitor improvements
Look for signals
 FMEA
(failure mode effect analysis)
Why Battles are Lost

Lost Lost Lost Lost Lost

Nail Shoe Horse Soldier Battle

Cause Failure Effect

Mode
FMEA
•Identifies the ways in which a product or process can fail
•Estimates the risk of specific causes with regard to these failures
•Prioritizes the actions that should be taken to reduce the chance of failure
 DOE - (design of experiments)
will help us identify...
A1 A2
í factors which shift the average

B1
B2 í factors which affect variation

C2
C1 í factors which shift the average and affect
variation

D1=D2
í factors which have no effect
 DOE - (design of experiments)
Measure the Process
Controllable Inputs Establish the
performance
X1 X2 X3 baseline
Quality
Characteristics:
Outputs
LSL USL
Inputs: Y1, Y2, etc.
Raw
Materials,
The Process
components,
etc.
S O D R
Process Actions
Potential Failure Mode Potential Failure Effects E Potential Causes C Current Controls E P
Step/Input Recommended
V C T N

Load DMF/DMF Fool proof this process

Operator Certification/ Process
Load Accuracy Mischarge of DMF Viscosity out of spec 7 SOP not Followed 5 5 175 using input from TQL
Audit
Team
Steam to Include Daily sign-off of
DICY/Scale Scale Not Zeroed Mischarge DMF 3 Faulty Scale 2 None 9 54 Scale funtion in Shift
Accuracy set-up verification.
Load DMF/DMF
Maintenance Procedure (SOP
Load Accuracy Mischarge of DMF Viscosity out of spec 7 Equipment Failure 2 3 42
5821)/Visual Check
Steam to

N1 N2 N3
DICY/Scale Scale > 0 Low DMF Charge 3 Water in Jacket 2 Visual Check of Jacket (SOP 5681) 4 24
Accuracy
Steam to
DICY/Scale Scale Inaccurate High DMF Charge 3 Tank Hanging Up 2 Visual Check (SOP 5681) 4 24
Accuracy

Uncontrollable Inputs
 DOE - (design of experiments) Analyse the
Key Outputs:

1
2
Variable
Process
How Measured When Measured

Noise Variables: Variable How Measured When Measured

1
2
3
4
5
Controllable Inputs
Controllable Inputs Variable How Measured When Measured

1
2
3
4
X1 X2 X3
5

Overall Sampling Plan:

Quality
Characteristics:
Outputs
LSL USL
Inputs:
Y1, Y2, etc.
Raw
Materials,
The Process
components,
etc.
Run Temperature Pressure
1 Hi Hi
2 Hi Hi
C a p a b ility u s in g P o o le d S ta n d a rd D e v ia tio n

X b a r a n d R Ch a rt Ca p a b ility Histo g ra m
3 Lo Hi
3 .0

4 Lo Hi
Means

2 .5 U C L =2 . 5 6 8
M U =2 .3 7 6
L C L =2 .18 3
2 .0

1. 5
1.5 2 .5 3 .5
S ubgr

0 .9
1 2 3 4

U C L =0 . 9 6 2 1
No rm a l P ro b P lo t 5 Hi Lo

N1 N2 N3
Ranges

0 .6

0 .3
R =0 . 5 16 2

6 Hi Lo
L C L =0 .0 7 0 2 7
0 .0
1.5 2 .5 3 .5

L a st 4 S u b g ro u p s Ca p a b ility P lo t
P ro c e ss To le ra n c e
7 Lo Lo
3 .0 1.8 3 17 5 2 .9 19 5 8

Uncontrollable Inputs
Values

2 .5 C p : 2 .7 6 I I I

2 .0

1. 5
C P U : 2 .9 9
C P L : 2 .5 3
C p k : 2 .5 3
1
I I

Sp e c if ic a t io n s
4
I

8 Lo Lo
1 2 3 4 St D e v : 0 .18 13 0 6
Su b g ro u p N u m b e r
 DOE - (design of experiments) Improve the
Three Factor Design
Process
PrimWdth
Controllable Inputs
X1 X2 X3
Nip FPM

ScrewRPM

Quality

X Characteristics:
Outputs
LSL USL
Inputs:

X
Y1, Y2, etc.
Raw
Materials,
The Process
components,
etc.

X LSL USL

N1 N2 N3
Uncontrollable Inputs
 DOE - (design of experiments) Control the
Process
Error Controllable Inputs

Proofing X1 X2 X3
Quality
Characteristics:
Outputs LSL USL
Inputs:
Raw Y1, Y2, etc.
Materials,
components, The Process
etc.

Work
Instructions Check

5 C’s Lists
N1 N2 N3
Uncontrollable Inputs
 Analyze
• Investigate source of variation (Special cause / Common causes)

– Special cause variation are the one off, occasional and obvious
cause of a process / quality problems.

– Common cause variation are the day in day out causes of process
problems, because the process is not stable enough, they are hidden
(these form 80% of process problems)

– Conventional non-conformance management systems seek to solve

special cause variation (e.g. concessions) - but these only represent
15 - 20% of the total variation.

– 6 Sigma addresses all variation.

 Improvement methodology
• Prioritise improvements
– Impact Vs Effort
– Brainstorming
– Affinity diagrams
– Solution selection matrix

• Tactical implementation plans

– Deliver improvements (reduce variation
Improve systematically)

Customer protection
Get control
Improve process
 Improve
• Prioritise improvements
– Tool commonly in uses are, Impact Vs Effort,
Brainstorming, Affinity diagrams, Solution selection
matrix.
– These tools help define the best method to meet the
customer need (as defined in the QFD)

• Tactical implementation plans

– Deliver improvements to reduce variation
systematically i.e. make a change, note the
improvement and make the next improvement.
– Critical we need to establish that any change is a
change for the good.
Improvement methodology
• Control the process
– Recover
• Control plans
• Escalation process
– Prevent
• Poke yoke (mistake/ error proof)
– Monitor
• Control charts
• Checksheets
• Documentation and Standardisation

Control
 Control
• Control the process

– Recover, Control plans, Escalation process.

– Prevent by Poke yoke (fool proof the process) to fundamentally

remove the rood causes of process variation.

– Monitor, Control charts, Checksheets, Documentation and

Standardisation, to ensure that stable process is maintained and that
the process does not degrade.

• The objective is to remove the root causes of process variation,

management are only left with a few critical input variables in the
process that need controlling and not all inputs as before.
 Where does 6 Sigma fit with Lean
• Lean and 6 Sigma both seek to
deliver business improvement

• They are different in the

6 Sigma methods used and tools
improvements employed
– Lean typically address the
total manufacturing
environment
– 6 sigma typical address the
root cause of process
Lean
variation
improvements

• There is significant benefit from

using the most appropriate tools
and improvement methodology
to meet the customer
requirements