untitled 6σ
(Quality Assurance: A historical perspective)
(Quality: What have been done and what are needed) 6σ (GE) (Design
for Six Sigma: GE as an example) 6σ
(Does 6σ work in Taiwan?)
Bradley Efron (2004)
Statistics in Nature Assumption: Nature has no will and run by
rules that makes no exception—no magic, no miracles, no answered
prayers, no appeals to higher authority. Mechanics astronomy
physics chemistry Biology medicine
Statistics in humanity (human will actually play a major
role)
Economics and psychology Life itself might be exempt from common
physical rules.
Theory and Reality
Theorem 1 Don’t put all eggs into one basket.
Theorem 2 If you put all eggs into one basket, Watch the
basket!
1988
Quality Management
Not to teach them how to improve, But teach them how to speed up
the improvement.
How long a minute is depends upon Which side of the bathroom door
you’re on.
Quality Revolution
What is Quality Control All About?
What is Quality?
Balance & Profit
S. Thomas Foster
What is Quality?
? ? ? Index Management
How to Determine Quality Characteristics?
Mission Vision Action Items
BSC Six Sigma Mission/Vision Mission
To develop and implement the tools to help foster a culture of
continuous improvement on existing products
and processes, and a “Do it Right The First Time” philosophy on new
generations of products and
processes. Through appropriate use of these tools, BSC will benefit
from significant advances in quality and
Customer satisfaction. Vision
We will drive significant business improvement utilizing the Six
Sigma processes. To ensure success, we will create a sustainable
system of training to facilitate the usage of the Six Sigma tools
as a core competency at
BSC.
QUALITY ASSURANCE: HISTORICAL HIGHLIGHTS
1920/30’s: Walter Shewhart introduces statistical process control
(SPC) at AT&T 1940’s: War effort accelerates applications;
development of MIL STD 105, MIL STD 414 and other sampling
acceptance plans (Dodge & Romig) 1946: American Society for
Quality Control (ASQC) established (and start of numerous
professional conferences on quality control) Late 1940’s: Box et al
adapt design of experiments to industrial applications 1950’s:
Deming hailed for bringing total quality management (TQM) to Japan;
Deming Award established 1970’s: Taguchi introduces “off-line
quality control” and robust designs Early 1980’s: Deming, Juran,
Crosby, Shanian, Golomski et al bring their versions of TQM to U.S.
Mid-1980’s: Six Sigma introduced at Motorola 1986: Malcom Baldridge
Award (for performance excellence) launched; Motorola Six Sigma
program among first winners 1987: ISO 9000 international standards
established 1995: Six Sigma comes out of the closet with adoption
at GE 1997: ASQC becomes American Society for Quality (ASQ)
Some Key players
W. Edwards Deming Joseph M. Juran Kaoru Ishikawa Armand Feigenbaum
Philip Crosby Genichi Taguchi Others
Shewhart Control Charts
73.995
74.005
74.015
Two Main Issues
Western Sensitizing Rules
1. Outside 3-sigma limits 2. 8 points in a row in zone C 3. 6
points in a row increasing or decreasing 4. 14 points in a row
alternating up or down 5. 2 out of 3 points in a row in Zone A 6. 4
out of 5 points in a row in Zone B 7. 15 points in a row in Zone C
8. 8 points in a row not in Zone C
see also Balkin and Lin (2001)
Deming (1986)
The biggest problem that most any company in the Western world
faces is not its competitors, nor the Japan. The biggest problems
are self-inflicted.
Everyone doing his best is not the answer. Everyone is doing his
best! It is necessary that people understand the reason for the
changes that are necessary.
Total Quality Management
DEMING’S 14 POINTS
1. Constant of purpose 2. Adopt the new philosophy 3. Build in
quality 4. Awarding business on value, not price 5. Continually
eliminate problems 6. Training on the job 7. Adopt and institute
leadership 8. Drive out fear 9. Not Departmentalism but team work
10. Not slogans, but how to 11. No quotas 12. Allow pride in work
13. Education and retraining 14. Top Management Push
DEMING’S 14 POINTS (From Out of the Crisis)
1. Create constancy of purpose for improvement of product and
service 2. Adopt the new philosophy 3. Cease dependency on
inspection to achieve quality 4. End practice of awarding business
on basis of price tag alone. Instead,
minimize total cost by working with a single supplier 5. Improve
constantly and forever every process for planning, production
and service 6. Institute training on the job 7. Adopt and institute
leadership 8. Drive out fear 9. Break down barriers between staff
areas 10. Eliminate slogans, exhortations, and targets for the work
force 11. Eliminate numerical quotas for the work force and
numerical goals for
management 12. Remove barriers that rob people of pride of
workmanship. Eliminate the
annual rating or merit system 13. Institute a vigorous program of
education and self-improvement for
everyone. 14. Put everybody in the company to work to accomplish
the transformation
Goal
S/N Ratio ()
Quality Loss Function
Robust Design (Taguchi)
What do the customers want? What don’t the customers want?
Variation Deduction
Output
Input
Clay tiles fired in kiln Problem: Size variation in tiles
(Possibly due to temperature gradient in kiln)
Possible Remedies: 1. Buy a new kiln with precise controllability
of
temperature and temperature gradient 2. Seek a more “robust” recipe
for tile clay
Design of Experiment Solution: Add 5% Lime
Size variation reduced No additional cost
— (Taguchi Three-Stage Design)
(System Design)
(Parameter Design)
(Tolerance Design)
Quality Creation (Kano)
Konica Example (1970s)
What is the major complaint from the customer?
Picture failure
Konica Example
Under exposure Out of focus Blank film
Konica Example
Under exposure ==> Built in flash bulb Out of focus ==> Auto
focus Blank film ==> Auto loading and winding
Standard & Award
Malcom Baldridge Award
Similarities: Bladrige & Six Sigma
6σ focuses on training specialist vs. broad based training in
organization
EmphasizeEmployee Education, Training & Development
Work Systems
Information Management
6σ address through project improvement & process monitoring;
Bladrige encourage more organizational-wise approaches
Emphasize Customer Relationship & Satisfaction
Specified at the project level in 6σCenter piece for
improvement
Customer & Market Knowledge
Not specified in 6σStrategy Development
Not considered in 6σPublic Responsibility & Citizenship
Visionary LeadershipOrganization Leadership
Cumulative Financial SavingOrganizational Effectiveness
Project LevelFinancial & Market Results
Support Processes
Business Support Processes
Product & Service Processes
6σ does not explicitly considerEffective use of team should promote
employee satisfaction
Employee Well-Being & Satisfaction
ISO 9000
Theoretical Model Uderlying the Deming method
Theoretical Framework for
Ishikawa Seven Tools:
Check Sheets Pareto Chart Cause and Effect Diagram Histogram
Scatter Plot Regression/Correlation Control Chart
How to construct it? How to read it?
Check List
Pareto Chart
Cause-and-Effect Diagram
73.995
74.005
74.015
Too Little Information vs Too Many Information
Ishikawa Seven Steps Problem:
Recognition of the features of the problem Analysis:
Finding out the main causes Action:
Action to eliminate the causes Check:
Confirmation of the effectiveness of the action
Standardization:
Permanent elimination of the causes Conclusion:
Review of the activities and planning for future work
Tools of Quality Plan-Do-Check-Act Cycle Ishikawa Seven Tools
Histogram, Pareto Charts, Cause-and-Effect Diagram, Check Sheets,
Scatter Diagrams, Flowcharts, Control Charts.
Seven New Tools The Affinity Diagram, The Interrelationship
Digraph, Tree Diagram, Prioritization Grid, Matrix Diagram, Process
Decision Program Chart, Activity Network Diagram
Five S’s Seiri (organize); Seiton (neatness); Seiso (cleaning);
Seiketsu (standardization); Shitsuke (discipline).
Mt. Rushmore, front site
Six Sigma History
Late 1700’s Carl Frederick Gauss introduced the concept of the
Normal Curve 1920’s Walter Shewhart set 3 sigma deviations from the
mean as the correction point for a process 1979 an exasperated
Motorola executive named Art Sundry said, at a meeting, "The real
problem at Motorola is that our quality stinks!" and in 1985
Motorola Engineer / Executive, Bill Smith coined the term “Six
Sigma” Mid 1980’s, Motorola Chairman Bob Galvin set a new standard
of 6-sigma saving $2.2 billion in the process. 1988 Motorola won
the Malcolm Baldrige National Quality Award
Six Sigma History 1992 Allied Signal's Larry Bossidy adopted Six
Sigma and developed an entire system of leadership and support
systems began to form around the statistical problem solving tools
developed by Motorola. 1995, Jack Welch, CEO of GE, began to study
Six Sigma and eventually dove in. When GE does something, it does
it all the way. GE reported to have saved billions of dollars with
Six Sigma. GE’s success prompted a number of other companies to
adopt Six Sigma during the final years of the 1990's such as
Honeywell, Kodak. Six Sigma is well established in US industry and
is currently sweeping across Europe
WHAT IS SIX SIGMA?
Six Sigma is a quality metric, a highly disciplined improvement
process, a culture Brief history
Developed at Motorola (mid ‘80’s—publicized in mid ’90’s) Embraced
by Allied Signal, Eastman Kodak, ABB, IBM, TI, etc. (‘90’s) Adopted
as a key company initiative world-wide by GE (Sep ‘95)
Top-down management commitment: Be “lunatic” about quality Welch
(GE ex-CEO): Change from fixing products to fixing processes for
perfection Formal Goal: Reduce defects to 3.4 per million
opportunities (6 sigma quality) Implementation via hierarchy of
trained
Green belts: Member of team responsible for implementing Six Sigma
Black belts: Leader of team responsible for implementing Six Sigma
Master black belts: Teacher and mentor of black belts Champions:
Senior management person responsible for obtaining resources and
generating (and maintaining) enthusiasm for Six Sigma
WHAT IS SIX SIGMA? (continued) Built upon
Disciplined approach with overall road-map and focus on improvement
Stepwise improvement process: Measure, Analyze, Improve, Control
(MAIC)
Measure: Identify customer needs (CTQs: critical to quality
characteristics) and quantify existing defect levels Analyze:
Determine root causes of defects Improve: Permanently eliminate
root causes of defects Control: Maintain gains over time
Define (problem and goals up-front) added (DMAIC) Highly
data-oriented (emphasis on classical statistical methods: gage
R&R, SPC, DOE, regression, etc. ) and various
quality/management tools (e.g., QFD, FMEA) Emphasis on common
language for metrics, tools, goals, etc.
Initially focused on “fixing the factory,” e.g., reduce rework and
scrap reduction
Evolved to include design, processes, customer focus, and
“everything we do”
The Goal
3.4 defects per million opportunities for all processes critical to
customer satisfaction
Companies Implementing 6-Sigma
General Electric ABB Vetco CitiGroup DuPont Bombardier Ceridian
Polaroid Carlson Co. Sony IBM
Motorola American Express Eastman Kodak Amazon.com Honeywell Dow
Chemical Ford Motor Cargill Unisys Iomega Pacific
AlliedSignal Seagate Technology 3M Toshiba Navistar Raytheon
Lockheed Martin US Bank Eaton Texas Instruments LG
Electronics
-6σ -5σ -4σ -3σ -2σ -1σ y +1σ +2σ +3σ +4σ +5σ +6σ
Lower Specification
±2σ 95.45 45,500
±1σ 68.27 317,300
±6σ 99.9999998 0.002 ±5σ 99.999943 0.57
±4σ 99.9937 63 ±3σ 99.73 2,700
-6σ -5σ -4σ -3σ -2σ -1σ y +1σ +2σ +3σ +4σ +5σ +6σ
Lower Specification
The Six Sigma Metric ±1.5σ Shift
The 1.5 Sigma Shift is expected to occur over the very long-term
(typically years)
The Six Sigma Metric Specification
Limit Percent within Specification
Defects per million (Centered)
Defects per million (1.5σ shift)
± 1σ 68.27 30.23 317,300 697,700
± 2σ 95.45 69.13 45,500 308,700
± 3σ 99.73 93.32 2,700 66,810 ± 4σ 99.9937 99.3790 63 6,210 ± 5σ
99.999943 99.97670 0.57 233 ± 6σ 99.9999998 99.999660 0.002
3.4
A common definition of a 6 Sigma process is one which achieves 3.4
defects per million or less.
SIX SIGMA METRICS (for normally distributed property) ““Short Term
CenteredShort Term Centered”” versus versus ““Long Term ShiftedLong
Term Shifted””
Six Sigma CenteredSix Sigma Centered
Lower Spec LimitLower Spec Limit Upper Spec LimitUpper Spec Limit
TargetTarget
6 6 Sigma Process CapabilitySigma Process Capability (without mean
shift)(without mean shift)
SHORT TERM
.001 .001 parts per millionparts per million .001 .001 parts per
millionparts per million
+6σ
LONG TERM
Lower Spec LimitLower Spec Limit Upper Spec LimitUpper Spec Limit
TargetTarget 3.4 3.4 parts per millionparts per million
Mean Shifted 1.5Mean Shifted 1.5σσ
6 6 Sigma Process CapabilitySigma Process Capability (with mean
shift)(with mean shift)
GOAL: REDUCE VARIABILITY (SIGMA) AND CENTER PROCESS ON TARGETGOAL:
REDUCE VARIABILITY (SIGMA) AND CENTER PROCESS ON TARGET
-6σ
4.5σ1.5σ
20,000 lost articles of mail per hour Unsafe drinking water almost
15 minutes per day 5,000 incorrect surgical operations per week 2
short or long landings at most major airports
each day 200,000 wrong drug prescriptions each year No electricity
for almost 7 hours each month
3σ Capability 93.32% Historical Standard
4σ Capability 99.379% Recent Standard
6σ Capability 99.99966% Industry Vision
*
*
*
Room for much improvement perceived quality levels
• U.S. Internal Revenue Service phone tax advice: 2 to 2.5
sigma.
• Restaurant bills: 3 to 3.5 sigma.
• Prescriptions: 3.5 to 4 sigma.
• Airline baggage handling: 3.75 to 4.25 sigma
• World-class quality companies: 4.5 to 5 sigma
• US airline fatality rate: Over 6 sigma
Sigma: a measure of quality
99% Good (3.8 Sigma) 20,000 lost articles of mall per hour Unsafe
drinking water for almost 15 minutes each day 5,000 incorrect
surgical operations per week Two short or long landings at most
major airports each day 200,000 wrong drug prescriptions each
year
99.99966% Good (6 Sigma) Seven articles lost per hour
One unsafe minute every seven months 1.7 incorrect operations per
week One short or long landing every five years
68 wrong prescriptions per year
Why Six Sigma?
Strategy for Running a Business Tool to Eliminate Variation Vision
of Product & Service Excellence Value to Our Customers Metric
of World Class Companies Goal for Competitive Strength
WHY NOW? • Accessibility to
– Automatic monitoring
– Large/immense databases
• General availability of statistical software to “tame” the data
(“democratization” of statistics)
• Improved communications: E-mail, video-conferencing,
Internet
• (Maybe) Impact of Deming, Juran, Taguchi et al
• (Definitely) Emphasis on bottom-line & gaining competitive
edge
• Room for much improvement (low-lying fruit); perceived quality
levels:
• U.S. Internal Revenue Service phone tax advice: 2 to 2.5
sigma
• Restaurant bills: 3 to 3.5 sigma
• Prescriptions: 3.5 to 4 sigma
• Airline baggage handling: 3.75 to 4.25 sigma
• World-class quality companies: 4.5 to 5 sigma
• US airline fatality rate: Over 6 sigma
Motorola Best Practice More than Manufacturing ... All Functions
Customer and Supplier Linkages Critical Top Level Commitment
Critical ... Not a “Program” Detailed, Extensive Training - 40
Hours/Employee/Year Grew Shares in Highly Competitive Global
Markets Shared What Worked and What Didn't
Six Sigma Journey
For Each Product or Process CTQ → Measure, Analyze, Improve,&
Control 1. Identify/Define CTQ’s: What are the customer needs &
key processes? (survey/interview/inquiries) 2. Measure: What is the
frequency of defects? (measurement system/process mapping/sigma
rating) 3. Analyze: When and where do defects occur?
(Statistics/Pareto/benchmarking/etc.) 4. Improve: How can we fix
the process? (design of experiments/expert brainstorming/etc.) 5.
Control: How can we ensure the process remains fixed? (measurement
feedback control/procedural/etc.)
Changing focus from Y to X
Y Dependent Output Effect Symptom Monitor
X1, …, Xn Independent Input-Process Cause Problem Control
Identifying and fixing root causes will help us obtain the desired
output
Premises underlying Six Sigma
All Processes Have Variability Variability Has Definable Causes
Typically Only a Few Causes Are Significant If Causes Can Be
Identified and Understood ... They Can Be Controlled Designs Must
Be Robust to the Effects of the Remaining Process Variations
Meco, Inc.: A Case Study
Meco, Inc. Located in Greenville, TN More than 600 Full Time
Employees Furniture Line:
Banquet tables Card tables Folding chairs Step ladders Stepping
stools
Grill Line: Charcoal grills Electric grills Portable Grills
Rotisserie Grills Smokers
Goals: Determine settings for paint guns so that a procedure can be
written for painting Elimination of reinforcement (manual) paint
position. Gain process knowledge.
Plan: Form team to carry out plan and goals. Study current process
(control charts). Design and run experiment.
Results: Procedure written for painting balck grill bowls. Reduced
average paint thickness approximately 42.5% Reduced variation in
paint thickness approximately 74.9%
Process Flowchart Hang product on line
Pre-treatment
Historical Data
Paint Thickness by Location and Hook, Current Data: Routine Paint
Process Settings Bowls Hoods Loc. Top X Bot X Top R Bot R Top X Top
X Top R Bot R 1 1.72 1.33 0.9 0.8 1.0 1.01 0.7 1.5/0.9 2 1.80 1.39
0.5 1.0/0.4 1.27 0.98 0.7 0.7 3 1.43 1.36 0.6 0.6 0.82 1.03 0.3 0.9
4 1.63 1.36 1.3/0.7 0.7 0.86 0.91 0.5 0.4 5 2.42 2.19 1.1 1.8/0.9
2.10 2.12 0.9 0.8 6 2.31 2.39 1.4 1.4 2.44 2.20 0.8 0.8 7 1.95 2.14
0.8 0.9 1.67 1.92 1.0 0.9 8 2.18 2.26 1.1 1.1 1.79 2.12 1.5/0.9 0.7
9 2.13 2.55 0.7 1.4 2.04 2.26 0.8 1.2/0.8 10 2.24 2.44 1.3 1.4 2.15
2.49 1.2/0.8 1.6/1.2
Paint Thickness by Location and Hook, Current Data Bowls
Special Powder Pressure Setting
Bowls Routine Setting Without Manual Reinforcement
Loc. Top X Bot X Top R Bot R Top X Top X Top R Bot R 1 1.59 1.2 0.7
0.6 1.35 1.3/1.03 0.4 1/0.1 2 1.575 1.12 0.7 0.5 1.45 1.075 0.8/0.4
0.7 3 1.14 1.26 0.6 0.6 0.925 0.9 0.3 0.4 4 1.21 1.28 1.0 0.9 1.00
1.05 0.2 0.3 5 1.86 2.08 0.6 0.6 2.95 2.425 0.4 0.7 6 1.89 2.19 0.8
0.9 2.35 2.175 0.4 0.4 7 1.59 2.1 0.6 1.1 1.925/1.7 2.3 1.2/0.8 0.4
8 1.56 2.32 0.6 0.6 1.55 2.625 0.6 0.5 9 1.99 2.15 0.8 1.4 2.525
2.8 0.4 0.3 10 1.73 2.375 0.7 0.8 2.05 2.975 0.6 1.3
Uncontrollable Variables: Line speed Ambient temperature Humidity
ground Forward/Dilution air pressure Fluidizing
Experimental Variables: Gun 1 position Gun 8 position Gun 9
position Powder pressure Gun 7 position Gun 11 position
Controlled variables: Experimental Levels Variable Coded-level (-1)
Coded level (0) Coded level (+1) Gun 1 position 58.3/74 59.3/77 Gun
8 position 48.3/81 49.3/84 Gun 9 position 37.5/80 38.5/83 Powder
pressure 25 lbs 35 lbs 45 lbs Gun 7 position 64.5/80 65.5/83
66.5/86 Gun 11 position 52/79 53/82 54/85
Experimental Responses Hook Location Response Top Hook
Inside (1-4) Side (5,6) Outside (7-10)
Average and range of paint thickness Average and range of paint
thickness Average and range of paint thickness
Bottom Hook Inside (1-4) Side (5,6) Outside (7-10)
Average and range of paint thickness Average and range of paint
thickness Average and range of paint thickness
Weight Top and bottom hook
Weight of paint covering parts
How would you run the experiment?
Experimental Design
Exp. No.
Run Order
Gun 1 Gun 8 Gun 9 Pressure Gun 7 Gun 11
1 20 -1 -1 1 -1 -1 0 2 19 1 -1 -1 -1 -1 0 3 17 -1 1 -1 -1 -1 0 4 36
1 1 1 -1 -1 0 5 34 -1 -1 1 0 -1 1 6 2 1 -1 -1 0 -1 1 7 23 -1 1 -1 0
-1 1 8 6 1 1 1 0 -1 1 9 22 -1 -1 1 1 -1 -1 10 26 1 -1 -1 1 -1 -1 11
28 -1 1 -1 1 -1 -1 12 10 1 1 1 1 -1 -1 13 27 -1 -1 1 -1 0 1 14 12 1
-1 -1 -1 0 1 15 31 -1 1 -1 -1 0 1 16 8 1 1 1 -1 0 1 17 9 -1 -1 1 0
0 -1 18 33 1 -1 -1 0 0 -1 19 14 -1 1 -1 0 0 -1 20 1 1 1 1 0 0 -1 21
18 -1 -1 1 1 0 0 22 21 1 -1 -1 1 0 0 23 24 -1 1 -1 1 0 0 24 35 1 1
1 1 0 0 25 29 -1 -1 1 -1 1 -1 26 15 1 -1 -1 -1 1 -1 27 32 -1 1 -1
-1 1 -1 28 16 1 1 1 -1 1 -1 29 13 -1 -1 1 0 1 0 30 30 1 -1 -1 0 1 0
31 3 -1 1 -1 0 1 0 32 11 1 1 1 0 1 0 33 5 -1 -1 1 1 1 1 34 25 1 -1
-1 1 1 1 35 7 -1 1 -1 1 1 1 36 4 1 1 1 1 1 1
Significant Variables for Each Response Variable
Top Hook Response 1% Significance 5% Significance 10% Significance
Inside Avg
Range P 9*7 7
8*11 9*P 1*7 1*7 8*11 8*7 1*P
Avg Range
8*7 1*7 9 7
Avg Range
P 11
Bottom Hook Response 1% Significance 5% Significance 10%
Significance Avg
Range P
1*11 9*11 9 9*P 8*P 9 1*7
Avg Range
P 9
8*7 1*11 8*P 1*P 9*11
Weight Response 1% Significance 5% Significance 10% Significance
Weight P 1 9*P 1*7 7 9*7
How would you analyze the data?
P is significant, set at level 0. Gun 8 is significant, set at
level -1. Gun 7 is significant, set at -1 or 0. Both 1x7 and 9x11
interactions are significant.
Gun 1 and gun 7 have a negative coefficient, thus must have
different levels. Gun 9 and gun 11 also have a negative
coefficient.
How would you run the confirmation runs?
Confirmation Run Order and Settings Variable Run 1 Run 2 Run 3 Run
4 Run 5 Run 6 Run 7 Run 8 Run 9 Gun 1 -1 -1 +1 +1 -1 -1 +1 +1 -1
Gun 8 -1 -1 -1 -1 -1 -1 -1 -1 -1 Gun 9 +1 -1 +1 -1 -1 +1 -1 -1 -1
Pressure 0 0 0 0 0 0 0 0 0 Gun 7 -1 0 -1 -1 0 0 -1 -1 0 Gun 11 +1
-1 0 -1 +1 0 0 +1 0
Paint Thickness by Location and Hook, Current and New Paint Process
Settings Current
Process New
Process Loc. Top X Bot X Top R Bot R Top X Top X Top R Bot R 1 1.72
1.33 0.9 0.8 0.95 0.8 0.6/0.3 0.2 2 1.80 1.39 0.5 1.0/0.4 0.9/0.7
0.9 0.6/0 0.2 3 1.43 1.36 0.6 0.6 0.83 0.6 0.2 0.2 4 1.63 1.36
1.3/0.7 0.7 0.60 0.87 0.5 0.1 5 2.42 2.19 1.1 1.8/0.9 1.67 1.23 0.1
0.4 6 2.31 2.39 1.4 1.4 1.43 1.1 0.2 0.2 7 1.95 2.14 0.8 0.9 1.43
1.47 0.2 0.2 8 2.18 2.26 1.1 1.1 1.13 1.63 0.1 0.3 9 2.13 2.55 0.7
1.4 1.33 1.07 0.4 0.3 10 2.24 2.44 1.3 1.4 1.47 1.43 0.3 0.3
The New Process Will ...
Reduce the black paint cost by $7,681 to $15,361 per year Decrease
the amount of recycled paint in the paint booth Eliminate the cost
of the touch-up paint process Decrease the time and cost in setting
up the paint process Decrease the number of bowls reworked
(stripping and repainting)
Procedure
1) Define your goal as clear as possible 2) Use SPC technique
(historical data) to understand
the process 3) Identify controllable / uncontrollable factors 4)
Use DOE technique to find the “optimal” setting
for potential improvements 5) Run confirmation experiments
Define Measure Analysis Improve Control
Six Sigma Improvement Process
Analyze
Define
How should I Bound the Problem?
Understand the Customer's Critical to Quality (CTQ) Factors Link
the CTQs to Key Internal Processes Prioritize Key Processes on
Which to Focus
Phase I: Measure
How Good Is the Process Today?
Determine How to Measure the Capability of the Process Validate
That the Measuring Approach is Reliable (Gage R&R) Gather the
Correct Sampling of Data
Process capability
Over time, a “typical” process will shift and drift by
approximately 1.5σ
Phase II: Analyze
What Causes Defects in the Process Today? Use the Sample Data to
Distinguish Current Performance vs. “Entitlement” Hypothesize What
the Underlying Factors Are That Are Causing the Difference Between
Current Performance and “Entitlement” Design Experiments to Confirm
the Importance of the Underlying Factors
Baselining & Benchmarking an existing process
Baselining = current process / benchmarking = ultimate goal
Phase III: Improve
How do I Fix the Underlying Causes of Variation? Conduct
Experiments to Quantify the Impact of the Underlying Factors on the
Capability of the Process Design Fixes to Control the Underlying
Factors Within Acceptable Bounds Confirm That the Economics of the
Proposed Fixes are Positive
Design of Experiments (DOE) SCREENING
For Experiments involving a Large Number of Factors Useful in
Isolating the “Vital Few” from the Trivial Many
CHARACTERIZATION For Experiments Involving a Relatively Small
Number of Factors Useful When Studying Relatively Uncomplicated
Effects & Interactions
OPTIMIZATION For Experiments Involving Only few Factors Useful When
Studying Highly Complicated Effects & Relationships
Phase IV: Control
How do I Control the Variation in the Underlying Factors?
Deploy Measurement Tools to Confirm Process Is in Control (SPC,
Cheek Sheets, etc.) Establish Feedback Mechanisms/Alarms to Confirm
Process Is in Control
Some basic six-sigma tools
Some basic six-sigma tools
Some basic six-sigma tools
Seven Basic Tools Analysis of Variance Variance Components
Regression Analysis Fault tree Analysis Failure Modes & Effects
Analysis Mistake Proofing Acceptance Sampling Reliability/Survival
Analysis
Black Belt
Belt Definitions People with a Yellow Belt (YB): participate in Six
Sigma Problem Solving Projects People with a Green Belt (GB): lead
less complex Six Sigma Projects, participate in BB Teams &
support YB Teams People with a Black Belt (BB): lead complex Six
Sigma Projects, mentor, support and develop GB’s & YB’s; Part
& Full Time People with a Master Black Belt (MBB): are the
experts in application & training
Yellow Belt
Green Belt
A 125-year old, $126B High-Tech Growth Company
Technology is Key to Growth
Founded by Thomas Edison
A diversified technology, manufacturing and services company with a
commitment to achieving
world leadership in each of its key businesses
Aircraft EnginesAircraft Engines Capital ServicesCapital
Services
Consumer ProductsConsumer Products Industrial SystemsIndustrial
Systems
Information ServicesInformation Services Medical SystemsMedical
Systems
PlasticsPlastics Power SystemsPower Systems
NBCNBC
General Electric Overview
20 KEY LESSONS LEARNED (Source: Hahn, G., Six Sigma Forum Magazine,
May 2002)
The time is right: Intense competitive pressures, consumer demand,
computers The enthusiastic commitment of top management is
essential (as shown by GE)
Be “lunatic” about quality Make Six Sigma a major factor in
promotion, incentive compensation, etc. Set up structure: Champion,
master black belts, black belts, etc. Provide the needed resources
and priorities Ensure everybody is paying attention Take overall
responsibility Follow up
Develop structure—not extra-curricular activity: Organization,
budget, objectives, responsibilities, measurement of results Commit
top people (champions, MBBs, BBs) Invest in relevant hands-on
training
Knowledgeable trainers who are outstanding communicators Customize
to needs of specific businesses Ensure common vocabulary
Incorporate hands-on involvement Consider engaging external gurus
to expedite Six Sigma introduction
Select initial projects to build credibility rapidly: Importance
evident, viable and doable in short time, readily quantifiable
(reap “low-lying fruit”—often in manufacturing)
SIX SIGMA: 20 KEY LESSONS LEARNED (Continued)
Make it all pervasive, and involve everybody (“the way we work”)
Emphasize Design for Six Sigma (DFSS) And don’t forget Design for
Reliability Focus on the entire system: Consider all CTQ’s
simultaneously Emphasize customer CTQ’s Include commercial quality
improvement Recognize all savings and costs—including savings from
averting problems Customize to meet business needs, e.g.,
Basic concepts of design of experiments universally applicable
Mixture experiments especially relevant for chemical processing
Planning of consumer surveys especially relevant for marketing and
servicing operations
Consider variability as well as the mean (“variability is evil”)
Plan to get the right data
In God we trust, all others bring data Advance planning is critical
Data quality generally more important than data quantity
SIX SIGMA: 20 KEY LESSONS LEARNED (Continued)
Beware of dogmatism: Goal is to gain quality improvement—not to use
any specific tool Avoid nonessential bureaucracy—a criticism of Six
Sigma Keep the toolbox vital; with experience
Some tools have been found more important than others Some not
originally included have proven their value
Expect Six Sigma to become a more silent partner over time
No more headline news when it is “the way we work” Retain momentum
by continued successes
THE FUN HAS JUST BEGUN!
A LOOK INTO THE FUTURE OF SIX SIGMA
At GE Will continue to evolve Will be de-emphasized per se--but
accepted as “way we work” Will be meshed with other initiatives:
e.g., e-Business Impact of training will remain; but some
practitioners might become rusty on tools
Rest of world Some companies will jump on Six Sigma wagon
Actively being promoted by ASQ, etc. Opportunity for academia
Similar cycle as GE seems likely
Others might not adopt Six Sigma per se--but will face similar
challenges Not limited to businesses; also applicable to banks,
social agencies, government, schools, etc.
Summary Define
Measure Identify CTQ & CTP (Critical to Process) Variables Do
Process Mapping Develop and Validate Measurement Systems
Analyze Benchmark and Baseline Processes Calculate Yield and Sigma
Target Opportunities and Establish Improvement Goals Use of Pareto
Chart & Fishbone Diagrams
Key components of breakthrough strategy: A mix of concepts and
tools
Improve Use Design of Experiments Isolate the“Vital Few”from
the“Trivial Many” Sources of Variation Test for Improvement in
Centering Use of Brainstorming and Action Workouts
Control Set up Control Mechanisms Monitor Process Variation
Maintain in Control Processes Use of Control Charts and
Procedures
QC … TQM … 6σ …
--
:
Appendix
More Readings
STILL QUESTION?
Send $500 to Dennis Lin Department of Management Science &
Information Systems Penn State University
+1 814 865-0377 (phone)
+1 814 863-7076 (fax)
Summary of Last Four Lectures
Reality
What’s Statistics All About?
Making Sense Out of Numbers
Math Notations x x, y, z, u, v, w x1, x2,…,xn y1, y2,…,yn; z1,
z2,…,zn; w1, w2,…,wn … α, β, γ, ... αi, βj, γk, ... αijk,
βj.k.pqr,
,
E.F. Schumacher When the Lord created the world and people to live
in it—an enterprise which, according to modern science, took a very
long time—I could well imagine that He reasoned with Himself as
follows: “If I make everything predictable, these human beings,
whom I have endowed with pretty good brains, will undoubtedly learn
to predict everything, and they will thereupon have no motive to do
anything at all, because they will recognize that the future is
totally determined and cannot be influenced by any human action. On
the other hand, if I make everything unpredictable, they will
gradually discover that there is no rational basis for any decision
whatsoever and, as in the first case, they will thereupon have no
motive to do anything at all. Neither scheme would make sense. I
must therefore create a mixture of the two. Let some things be
predictable and let others be unpredictable. They will then,
amongst many other things, have the very important task of finding
out which is which.”
I love noise ! I love noise ! I love noise !
Black Belt Training
Module 1 (Define) Six Sigma Overview The Six Sigma Problem Solving
Process (overview) Seven Basic Tools for Quality Improvement
DMAIC Introduction Basic Statistics (includes Data Transformation)
Customer Focus (includes Process Mapping)
Project Reviews / DMAIC Review
Black Belt Training Topics
Module 2 (Measure) DMAIC Review Drill Down Chart / Process Mapping
Control Charts Capability Studies Analysis of Variance Measurement
Systems Analysis Change Point Analyzer Project Reviews / DMAIC
Review
Black Belt Training Topics
Module 3 (Analyze) DMAIC Review Fault Tree Analysis Multi-Vari
Charts Confidence Intervals Linear & Polynomial Regression
Design of Experiments Project Reviews / DMAIC Review
Black Belt Training Topics
Module 4 (Improve) DMAIC Review Multiple Regression Response
Surface Methodology VarTran Augmentation Robust Design Project
Reviews / DMAIC Review
Black Belt Training Topics
Module 5 (Control) DMAIC Review FMEA Mistake Proofing Hypothesis
Testing Acceptance Sampling Process Validation DMAIC Workshop
Project Reviews / DMAIC Review
Black Belt Training Topics
Course Curriculum at GE
INTRODUCTION OF SIX SIGMA AT GE CEO Jack Welch
Decides on company-wide quality initiative (“most important ever”)
Sold on Six Sigma (by Larry Bossidy); launches in September ‘95
Tied to promotion and management incentive compensation
Implementation centered within each GE business Champion (senior
manager) reporting to business leader, provides resources Master
Black Belts (MBBs)—full time mentor, trainer, change agent Black
Belts (BBs)—full time project leader Green Belts (GB)—Implementer
(part-time)
Plus small corporate group to provide coordination, share best
practices, etc.
Start-up Brought in Mikel Harry for MAIC (trained first MBBs)
(later brought in Maurice Berryman for Design for Six Sigma (DFSS))
Rapid transition to GE leaders
Training and certification All GE employees trained in strategies,
tools and techniques All professionals green belt (or black belt)
trained: 4 one week sessions plus two successful
projects--administered by MBBs Master black belt: Special training
beyond GB/BB & mentor 20 projects
GG/BB DMAIC CURRICULUM-WEEK 1: DEFINE/MEASURE
Classical Training Define/Measure
Overview and process improvement planning—The MAIC roadmap Quality
Function Deployment (QFD) – Matrix to relate what customer wants to
design requirements via “house of quality” Failure Mode and Effects
Analysis (FMEA) Organizational effectiveness concepts Basic
Statistics using Minitab Process capability Measurement systems
analysis
Revised Recommendation (Hoerl 2001) Context
Why Six Sigma? DMAIC & DFSS process Project management
fundamentals Team effectiveness fundamentals
Define Project selection Scoping projects Developing project plan
Multi-generational projects Process identification
Measure QFD (and House of Quality)
Identifying customer needs Developing measurable CTQ metrics
Sampling (data quality and quantity) Measurement system analysis
(not just gauge R&R) SPC Part I
Statistical Control (process stability) Implications of
instability
Capability analysis
Classical Training Statistical thinking Hypothesis testing (F, t,
etc.) Correlation Simple regression
Revised Recommendation (Hoerl 2001) Basic graphical improvement
tools (“Magnificent 7”: Flow chart, Pareto chart, cause &
effect diagram, histogram, scatter plot, stratification, and run
chart) Management and planning tools Confidence intervals
(emphasized) Hypothesis testing (de- emphasized) Analysis of
variance (de- emphasized) Regression Developing conceptual designs
in DFSS
GG/BB DMAIC CURRICULUM-WEEK 3: IMPROVE
Classical Training Design of Experiments (DOE)
Factorial experiments Fractional factorials Balanced block designs
Response surface designs
Analysis of Variance (ANOVA) Multiple Regression Facilitation
Tools
Revised Recommendation (Hoerl 2001) DOE, focusing on
Two level factorials Screening designs (fractional factorials)
Response surface designs
Piloting (of DMAIC improvements) FMEA Mistake-proofing DFSS design
tools
CTQ flowdown Capability flowup Simulation
(Note: Discussion might spill into week 4)
GG/BB DMAIC CURRICULUM-WEEK 4: CONTROL
Classical Training Control plans Mistake-proofing Team development
Parallel, discrete, continuous process, administration and design
tracks
Note: 1) Taught by MBBs/BBs 2) ~9 day classroom training,
plus
~3 weeks project for each phase
Revised Recommendation (Hoerl 2001) Development control plans
Statistical process control-Part II: Using control charts Piloting
new designs in DFSS
Note: 1) Taught by MBBs/BBs 2) ~9 day classroom training,
plus
~3 weeks project for each phase
ABOVE IS MANUFACTURING ORIENTED: CHANGE FOR DFSS, FINANCE,
ETC.
More Readings
See the following web sites: www.isixsigma.com www.asq.org
See the following texts: Books on Quality Improvement by Thomas P.
Ryan or Douglas Montgomery Books on Six Sigma by Peter Pande or
Forrest Breyfogle
See the following article: Snee, Ronald (1999), “Why should
statisticians pay attention to Six Sigma”, Quality Progress
32(9):100- 103.
WHERE TO GO FOR MORE INFORMATION BOOKS LISTED ON WEB
(www.isigma.com/books): 2000 and before
Breyfolge et al, Managing Six Sigma: A Practical Guide (2000) Pande
et al, The Six Sigma Way: How GE, Motorola and other Top Companies
are Honing their Performance (2000) Rath & Strong, Rath and
Strong’s Six Sigma Pocket Guide (2000) Arthur, Six Sigma Simplified
(2000) Snyder: Understanding the Essentials of the Six Sigma
Quality Initiative (2000) Naumann, Hoisington, Customer Centered
Six Sigma:Linking Customers, Process Improvement and Financial
Results (2000) Oriel Incorporated, Guiding Successful Six Sigma
Projects (2000) Tennant, Six Sigma: SPC & TQM in Manufacturing
and Services (2000) Eckes, GE’s 6 Sigma Revolution: How GE &
Others turned Process into Profits (2000) Brue: Six Sigma for Team
Members (2000) Harry and Schroeder, Six Sigma: The Breakthrough
Management Strategy Revolutionizing the World’s Top Corporations
(1999) (also audio) Breyfogle, Implementing Six Sigma: Smarter
Solutions using Statistical Methods (1999) Perez-Wilson, Six Sigma:
Understanding the Concept, Implications and Challenges (1999)
Lawson, Measuring Six Sigma and Beyond (1997) Prins, Six Sigma
Metrics (1993) Harry, Vision of Six Sigma (1997) -- 8 volumes Harry
et al, Six Sigma Mechanical Design Tolerancing (1988) Harry, The
Nature of Six Sigma Quality (1988)
WHERE TO GO FOR MORE INFORMATION BOOKS LISTED ON WEB
(www.isigma.com/books): 2001
Stamatis, Six Sigma and Beyond: 8 Volumes: Foundations, Problem
Solving & Basic Maths, Statistics & Probability,
Statistical Process Control, Design of Experiments, Design for Six
Sigma, Implementation Process (2001/2) Breyfogle et al, Wisdom on
the Green: Smarter Six Sigma Business Solutions (2001) Eckes,
Making Six Sigma Last: Managing Cultural and Technical Change
(2001) Brassard & Ritter, Sailing Through Six Sigma (2001)
Chowdhury, The Power of Six Sigma: How 6 Sigma is Transforming the
Way we Work (2001) Arthur, Six Sigma Instructor Guide (2001)
Pyzded, Six Sigma Handbook: Complete Guide for GBs, BBs and
Managers (2001) Pande et al, What is Six Sigma? (2001) Lowenthal,
Six Sigma Project Management: A Pocket Guide (2001) Tayntor, Six
Sigma Software Development (2001) Mills et al, The Path to
Integration of Lean Enterprise and Six Sigma (2001) Munro, Six
Sigma for the Shop Floor: A Pocket Guide (2001) Juran Institute,
Implementing Juran’s 6-Step Quality Improvement Process and Six
Sigma Tools (2001) Harry, Six Sigma Knowledge Guide (2001) Bhote,
The Ultimate Six Sigma: Beyond Quality Excellence (2001) American
Productivity & Quality Center, Deploying Six Sigma (2001)
Perez-Wilson, Six Sigma: Understanding the Concept, Implications
and Challenges (2001) Dodson et al, Six Sigma Study Guide (2001)
Keller, Six Sigma Deployment: A Guide for Implementing in Your
Organization (2001) Tenant, Six Sigma and TQM in Manufacturing and
Services (2001) ?, The Power of Six Sigma: An Inspiring Tale of an
Extraordinary Process Transforming the Way we Work (2001)
WHERE TO GO FOR MORE INFORMATION BOOKS LISTED ON WEB
(www.isigma.com/books): 2002 (and 2003)
Harry, The Six Sigma Fieldbook: How to Successfully Implement Six
Sigma Breakthrough Strategy (2003) Harrington, Six Sigma Toolkit
Set (2003) Wheat et al, A Parable of the Journey to Six Sigma and
Lean Enterprise (2003) Eckes, Six Sigma for Everyone (2003) Rath
& Strong, Six Sigma Leadership Handbook (2003) Andell, Six
Sigma Leadership (2002) Snee and Hoerl, Leading Six Sigma: A Step
by Step Guide (2002) Ehrlich, Transactional Six Sigma and Servicing
(2002) Barney & McCarty, The New Six Sigma: A Leader’s Guide
(2002) Gaudard & Ramsey, Six Sigma Companion Guide:
Experimental Design for Practitioners (2002) Smith et al, Strategic
Six Sigma: Best Practices from the Executive Suite (2002) Craveling
et al, Design for Six Sigma in Technology and Product Development
(2002) Shina, Six Sigma for Electronics Design and Manufacturing
(2002) Eckes, Six Sigma Team Dynamics: The Elusive Key to Project
Success (2002) Plotkin, Sis Sigma: What It Is and How to Use It
(2002) George, Lean Six Sigma, Combining Six Sigma Quality with
Lean Production Speed (2002) Walters, The Six Sigma Journey from
Art to Science (2002) Tennant, Design for Six Sigma: Launching New
Products and Services without Failure (2002) Lientz & Rea,
Reach Six Sigma Goals Without the Pain (2002) Campbell, Applying
Project Management to Six Sigma Projects (2002) Adams et al, Six
Sigma Deployment (2002) Dedeke, What Makes Six Sigma Work (2002)
Barry et al, The Six Sigma Book for Healthcare (2002) Conner, Six
Sigma and other Continuous Improvement Tools for the Small Shop
(2002) Gordon, Six Sigma Quality for Business and Manufacture
(2002) ? Six Sigma and Accelerated Testing (2002)
WHERE TO GO FOR MORE INFORMATION (Continued) Six Sigma Forum,
American Society for Quality publication, (First issue
November
2001); articles to date (August 2002) include: Gregory Watson:
Cycles of Learning: Observations of Jack Welch Janet Young: Driving
Performance Results at American Express Larry Smith: Six Sigma and
the Evolution of Quality in Product Development Mark Goldstein: Six
Sigma Program Success Factors Larry Bossidy et al: The Honeywell
Edge Dennis Attenello and John Uzzi: Achieving Six Sigma in
Financial Services Debbie Phillips-Donaldson: Six Sigma Payoff
Charles Gowen: How to Implement Six Sigma for Maximum Benefit Tracy
Thurkow: Manage Behavioral Changes Tim Young: Merging Six Sigma and
IT Matt Barney: Motorola’s Second Generation Andrea Kabcenell &
Donald Berwick: Pursuing Perfection in Healthcare Greg Stock:
Taking Performance to a Higher Level Gerry Hahn: 20 Key Lessons
Learned Roger Hoerl: An Inside Look at Six Sigma at GE Robert
Galvin: In the Beginning (at Motorola) Rick Edgeman & David
Bigio: Six Sigma in the Academic World Douglas Gorman & Keith
Bower: Measurement System Analysis and Destructive Testing Charles
Huber: Straight Talk on DFSS Gregory Watson: Selling Six Sigma to
Upper Management Larry Smith: Are Your Champions Doing a Good
Job?
GUIDE TO ABBREVIATIONS ABB: Asea, Brown, Boveri, Ltd. ANOVA:
Analysis of Variance ASQ: American Society for Quality ASQC:
American Society for Quality Control AT&T: American Telephone
and Telegraph Company BB: Black Belt CART: Classification and
Regression Trees CEO: Chief Executive Officer CT: Computer
Tomography Imaging CTQ: Critical to Quality characteristic DFSS:
Design for Six Sigma DMAIC: Define, Measure, Analyze, Improve,
Control DMADOV: Define, Measure, Analyze, Design, Optimize and
Verify (in DFSS)