Transcript Six Sigma Overview

Six Sigma
Achieving Our Objectives
指導教授:盧淵源
第二組 組長:蔡孟行
組員:詹美琴 陳文德
吳宗啟 周永宏
簡總益
1
The Focus of Six Sigma

Identifying critical aspects of the business with
problems or opportunities for improvement.

Targeting those critical areas and designating
improvement efforts as Six Sigma Black Belt
projects.

Selecting top people to work on the projects--full
time.

Ensuring these people have the time, tools, and
resources they need to succeed.
2
What Types of Problems Should We Target?






High Defect Rates
Low Yields
Excessive Cycle Time
Excessive Machine Down Time
High Maintenance Costs
Bottlenecks
3
What Types of Problems Should We Target?






High Defect Rates
Low Yields
Excessive Cycle Time
Excessive Machine Down Time
High Maintenance Costs
Bottlenecks
Non-Conformance
4
Cost Of Poor Quality (COPQ)
Traditional Quality Costs
(tangible)
Inspection
Warranty
Scrap
Rework
Rejects
Administration /
Disposition
Concessions
More Setups
Expediting costs
Lost sales
Late delivery
Lost Customer Loyalty
Excess inventory
Long cycle times
Engineering change orders
Additional Costs of Poor
Quality
(intangible)
Lost Opportunity
(Difficult or impossible to measure)
Hidden Factory
Average COPQ approximately 15% of Sales
5
What is Cost of Poor Quality?

In addition to the direct costs associated with finding and fixing
defects, “Cost of Poor Quality” also includes:
• The hidden cost of failing to meet customer expectations the first time
• The hidden opportunity for increased efficiency
• The hidden potential for higher profits
• The hidden loss in market share
• The hidden increase in production cycle time
• The hidden labor associated with ordering replacement material
• The hidden costs associated with disposing of defects

For most companies today, the cost of poor quality is likely to be
25 % of sales. For Seagate, that’s over $1 billion each year.

In almost every company where the COPQ is unknown, the COPQ
exceeds the profit margin.
6
The Role of Measurement
 If we cannot express what we know in the form of numbers, we really
don’t know much about it.
 If we don’t know much about it, we cannot control it.
 If we cannot control it, we are at the mercy of chance.
Certainty + Uncertainty = 100%
Known
+ Unknown
Belief + Disbelief
Confidence + Risk
= 100%
= 100%
= 100%
Yield + Defect Rate = 100%
7
Customer Focus: A Model For Success
People
People
Processes
Processes
Technology
Organization
Technology
Organization
Capability
Capability
• Business survival is dependent upon how well we satisfy
our customers.
• Customer satisfaction is a function of quality, price, and
delivery.
• Quality, cost, and prompt delivery are dependent upon
capability.
8
The Customer Supplier Interaction
Customer
Supplier
Delivery
Cycle Time
Price
Do
Need
Cost
Quality
Defects
We strive for Six Sigma capability
on Cycle Time, Cost, and
Conformance to meet customer
expectations on Delivery, Price,
and Quality.
9
How Does Six Sigma Make the Difference?






Vision
Philosophy
Aggressive goal
Metric (standard of measurement)
Method
Vehicle for:
»
»
»
»
Customer focus
Breakthrough improvement
Continuous improvement
People Involvement
10
Six Sigma Vision
The Vision of Six Sigma is to delight
customers by delivering world-class quality
products through the achievement of Six
Sigma levels of performance in everything we
do.
11
Six Sigma Philosophy
The philosophy of Six Sigma is to apply a
structured, systematic approach to achieve
breakthrough improvement across all areas of
our business.
12
Six Sigma - Aggressive Goal

Process
Capability
PPM
2
3
4
5
6
308,537
66,807
6,210
233
3.4
Defects per
Million Opp.
Sigma is a statistical unit of measure that reflects process
capability. The sigma scale of measure is perfectly
correlated to such characteristics as defects-per-unit,
parts-per million defective, and the probability of a
failure/error.
13
Statistical Definition of Six Sigma
- 6st
- 3 st
+ 6st
mo
+ 3st
scale
Process Width
Design Width
scale
LSL
LSL
TT
.001 ppm < LSL
USL
USL
.001 ppm > USL
scale
LSL
LSL
T
T
USL
USL
14
The Standard Deviation
m
Point of Inflection
1
p(d)
p(d)
T
1
2
3

3
This is a 6 Sigma Process
4
5
6
USL
Six Sigma - Performance Target
Sigma
Long-Term Yield
Standard
3 Sigma
93.32 %
Historical
4 Sigma
99.379 %
Current
5 Sigma
99.9767 %
Intermediate
6 Sigma
99.99966 %
Long-Run
16
Six Sigma -- Practical Meaning
99% Good (2.8 Sigma)
99.99966% Good (6 Sigma)
20,000 lost articles of mail per hour
Seven articles lost per hour
Unsafe drinking water for almost 15
minutes each day
One unsafe minute every seven months
5,000 incorrect surgical operations
per week
1.7 incorrect surgical operations per
week
Two short or long landings at most
major airports each day
One short or long landing every five
years
200,000 wrong drug prescriptions
each year
68 wrong prescriptions per year
No electricity for almost seven hours
each month
One hour without electricity every 34
years
17
The Strategy
USL
LSL

Characterize

Optimize

Breakthrough
T
USL
LSL
T
USL
LSL
T
LSL’
USL’
18
The Breakthrough Phases
Phase 1:
Measurement
Characterization
Phase 2:
Analysis
Breakthrough
Strategy
Phase 3:
Improvement
Optimization
Phase 4:
Control
19
Narrow
the scope of input variables --> ID leveraged KPIV’s
Six
Sigma--Methodology
Process Map
30 - 50
Inputs Variables
10 - 15
Key Process Input
Variables (KPIVs)
C&E Matrix and FMEA
Gage R&R, Capability
Measure
Multi-Vari Studies,
Correlations
Analyze
8 - 10
KPIVs
T-Test, ANOM, ANOVA
Screening DOE’s
DOE’s, RSM
Improve
4-8
Critical KPIVs
3-6
Key Leverage
KPIVs
Quality Systems
SPC, Control Plans
Control
Optimized Process
20
The Focus of Six Sigma
Y=
f (X)
To get results, should we focus our behavior on the Y or X?






Y
Dependent
Output
Effect
Symptom
Monitor






X1 . . . XN
Independent
Input-Process
Cause
Problem
Control
If we are so good at X, why do we constantly test and inspect Y?
Focus on X rather than Y, as done historically
21
The Breakthrough Strategy
Recognize
Define
Breakthrough Cookbook
Measure
Analyze
Improve
Control
1
2
3
4
5
6
7
8
9
10
11
12
Application Projects
A
B
C
Select Output Characteristic
Define Performance Standards
Validate Measurement System
Establish Product Capability
Define Performance Objectives
Identify Variation Sources
Screen Potential Causes
Discover Variable Relationships
Establish Operating Tolerances
Validate Measurement System
Determine Process Capability
Implement Process Controls
Goal:
Yi =f X1 , ..., XN
D
E
F
G
Focus on the Y’s
Product
Capability
Analysis
Focus on the X’s
Process
Capability
Analysis
22
This Drives Breakthrough Improvement
Performance
Six Sigma
Breakthrough
3 Sigma
BAD
6 Sigma
GOOD
Time
23
The Foundation of the Six Sigma Tools
Data is derived from objects, situations, or
phenomenon in the form of measurements.
Data is used to classify, describe, improve , or
control objects, situations, or phenomenon.
Levels of Analysis:
1. We only use experience, not data.
2. We collect data, but just look at the numbers.
3. We group the data so as to form charts and graphs.
4. We use census data with descriptive statistics.
5. We use sample data with descriptive statistics.
6. We use sample data with inferential statistics.
What level are you at?
24
Impact of Variation on Cost
The Traditional View
Loss
No Loss
Goal
Post
Mentality
Lower
Specification
Limit
Target
Upper
Specification
Limit
The Enlightened View
Loss
Loss
Lower
Specification
Limit
Target
Upper
Specification
Limit
25
Six Sigma Metrics
Existing Metrics
Yield
Scrap
Rework
?
?
?
?
?
Leadership Must Ask the Right Questions
Six Sigma Metrics
CTX’s(Cost, Quality, Delivery, Satisfaction)
Defects Per Unit
Complexity
Defects Per Million Opportunities
Rolled Throughput Yield
Rolled Throughput Yield Normalized
Sigma Score
Process Baseline
Process Entitlement
Process Benchmarking
KPIV’s
KPOV’s
Shift & Drift
What Gets Measured Gets Managed
26
Six Sigma Metrics - Definitions
CTx’s:
» Critical to Customer Satisfaction parameters. Typically,
these include, but are not limited to, cost, quality and
delivery
KPOV’s:
» Key Process Output Variables. The results of the collective
action of KPIV’s in a process on a product
KPIV’s:
» Key Process Input Variables. Those process variables that
directly and/or in conjunction with other KPIV’s, drive a
change in an output variable
27
CTX’s

CT = “Critical To…”
» CTS - Critical To achieving customer
Satisfaction. Typically, this includes,
but is not limited to, those parameters
which are
• CTQ - Critical to Quality
• CTD - Critical to Delivery
• CTC - Critical to Cost
» Six Sigma Leads you to the “Critical” to
Increase the Efficiency of the
Improvement Process…..Work on What
is Important
28
Key Process Variables

KPOV’s: Key Process Output Variables
» The process outputs critical to achieving the CTX’s,
• In Golf as an Example
» Distance hit or degrees off line from tee to hole

KPIV’s: Key Process Input Variables
» Those process variables that directly or in combination with
other input variables produce a direct effect in a KPOV
• Club selection, stance, back swing velocity, club face angle, wrist
action, follow-through
» Six Sigma Leads you to the “Criticals” to Increase the
Efficiency of the Improvement Process…..Work on What is
Important
29
Six Sigma Metrics - Definitions
Process Baseline:
» The average, long term defect level of a process when all
input variables in the process are running in an
unconstrained fashion
Process Entitlement:
» The best case, short term defect level of a process when
all input variables in the process are centered and in
control
Process Benchmark:
» The defect level of the process deemed by comparison to
be the best process possible
30
Process Baseline
Process Baseline: The
average, long term defect level of
a process when all input variables
in the process are running in an
unconstrained fashion
Long-term
Baseline
31
Process Entitlement
Process Entitlement: The best
case, short term defect level of a
process when all input variables in the
process are centered and in control
32
Process Benchmark
Process Benchmark:
The defect level of the process
deemed by comparison to be
the best process possible
Factory C
Factory B
Factory A
33
Six Sigma Metrics - Definitions
Shift:
» Step function change in the mean or average of a
population, often driven by a special cause or movement in
a key process input variable
» Sudden
Drift:
» Sustained trend in a mean or average of a population,
often due to a progressive change to an key process input
variable
» Gradual
34
Shift & Drift
Shift: Step function change
in the mean or average of a
population, often driven by a
special cause or movement in
a key process input variable
35
Shift & Drift
Drift: Sustained trend in a
mean or average of a
population, often due to a
progressive change to an
key process input variable
36
Six Sigma Metrics - Definitions
Defects per Unit:
» The total number of defects observed on a unit of output
Opportunities:
» The number of possibilities for defect creation in a process
or sequence of processes.
DPMO:
» Defects per million opportunities
37
Defects and the Hidden Factory
Each defect must be detected, repaired and
placed back in the process. Each defect
costs time and money.
Inputs
Operation
Rework
Hidden Factory
Scrap
OK
Inspect
NOT
OK
First Time
Yield
90%
Customer Quality
Yield After
Inspection or Test
• Wasted Time
• Wasted Money
• Wasted Resources
• Wasted Floor space
Manufacturing Variation Causes A "Hidden Factory"
Increased Cost - Lost Capacity
38
RTY Versus FTY
Inputs
Operation
Rework
Hidden Factory
OK
First Time
Yield
Inspect
=
90%
Customer Quality
Yield After
Inspection or Test
NOT
OK
Scrap
Process
A
90%
Yield
Rolled Yield
C
Final Test
D
90%
Yield
90%
Yield
90%
Yield
81 %
73 %
66 %
B
Rolled-Throughput Yield
66%  90%
... why not?
Classical First-Time Yield
Using “final test (or first time) yield” ignores the hidden factory. Final test
performance is a function of inspection & test not actual defect data.
39
Two Types of Defect Models
Conclusion:
Conclusion: The
Theuse
useof
ofaarandom
random
model
modelto
todescribe
describe
the
theoccurrence
occurrenceof
of
Uniform
defects
defectsisisplausible.
plausible.
Random
Universe
UniverseofofDefects
Defects
Uniform Defect: The same type of defect appears
within a unit of product; e.g., wrong type of steel.
Random Defect: The defects are intermittent and
unrelated; e.g., flaw in surface finish.
40
Defects per Unit
Defects per Unit (DPU): Average number of defects per unit produced
1. OD Dimension
2. ID Dimension
3. Flatness
4. Roughness
5. Coercivity
6. Carbon Thickness
7. Lube Thickness
8. Glide Height
Total Defects per Disc
1
x
Disc Number
2
3
4
5
x
x
x
x
x
3
x
1
2
1
0
DPU: 7 Defects / 5 Units = 1.4 Defects per Unit
41
Opportunities
Opportunities: The number of possibilities for defect creation in
any unit of product, process or sequence of processes.
1. OD Dimension
2. ID Dimension
3. Flatness
4. Roughness
5. Coercivity
6. Carbon Thickness
7. Lube Thickness
8. Glide Height
8 Opportunities
42
Opportunities
Opportunities: The number of possibilities for defect creation in
any unit of product, process or sequence of processes.
1. OD Dimension
2. ID 8Dimension
x 5 = 40 Opportunities
3. Flatness
4. Roughness
5. Coercivity
6. Carbon Thickness
7. Lube Thickness
8. Glide Height
DPMO: Defects per Million Opportunities
43
Six Sigma Metrics - Definitions
Classical Yield:
» The number of good units divided by the number of units
tested or inspected
Rolled Throughput Yield (RTY):
» The probability of a unit going through all process steps with
zero defects. This is used to identify and quantify the “Hidden
Factory”
Hidden Factory:
» The amount of work above and beyond the requirements
necessary to produce a unit of output
44
Classical Yield
In
Factory
Out
Out
Yield =
In
Scrap
The number of good units produced which have no defects, divided
by the number of units started, tested or inspected.
45
Classical Yield
Not all Yields are alike!
100
Factory A
85
100
Factory B
85
15
Scrap
46
Classical Yield
Not all Yields are alike!
100
Factory A
85
100
15
Scrap
Factory B
50
85
35
Rework
Factory C: The Hidden Factory
15 Scrap
47
Classical Yield
Not all Yields are alike!
100
Factory A
85
100
15
Scrap
Factory B
50
85
35
Rework
Equal Yields … Unequal Costs
Factory C: The Classical
Hidden Yield
Factory
does not correlate to cost, cycle time
or inventory levels
15 Scrap
48
Rolled Throughput Yield
1000
Process 1
950
The probability of going
through all process steps
Process 2 930 97.9% with zero defects
95.0%
50
Process 3
820
88.2%
20
110
Process 4
810
98.8%
10
Rework
90
47.4%
900
90.0%
49
Rolled Throughput Yield
1000
Process 1
950
The probability of going
through all process steps
Process 2 930 97.9% with zero defects
95.0%
50
820
Process 3
88.2%
20
Process 4
110
Yrt = (.950)*(.979)*(.882)*(.988) = 81.0%
Rework
810
98.8%
10
90
47.4%
900
90.0%
50
Rolled Throughput Yield
1000
Process 1
950
The probability of going
through all process steps
Process 2 930 97.9% with zero defects
95.0%
Correlates
to cost,
cycle
time,
and
50
820
Process 3
88.2%
inventory20levels
Process 4
110
Yrt = (.950)*(.979)*(.882)*(.988) = 81.0%
Rework
810
98.8%
10
90
47.4%
900
90.0%
51
Yield Comparison
Final Classical Yield
- Rolled Throughput Yield
=
Hidden Factory
52
An Average Measure
How do we measure the relative performance of a process?
Factory A
Factory B
Which factory
is Better?
Yrt = 80.1%
Yrt = 77.3%
53
An Average Measure
It’s a TRICK QUESTION
Factory A
90%
89%
Factory B
Factory B runs
higher average
yields at each step
95%
94%
96%
98%
92%
Yrt = 80.1%
Yrt = 77.3%
54
Six Sigma Metrics - Definitions
Normalized Throughput Yield:
» The yield for a total process averaged over all process steps
Complexity:
» A measure of how complicated a process or product is…the
more opportunities for defects a process or product has, the
more complex it is.
55
Normalized Thruput Yield
The yield for a total process averaged over all process steps (Yna)
Factory A
Yna = (Yrt)1/n
Factory B
95%
90%
94%
89%
96%
Yna = (.801)1/2 = 89.5%
Yna = (.773)1/5 = 95.0%
Yrt = 80.1%
98%
92%
Yrt = 77.3%
56
Normalized Throughput Yield
The yield for a total process averaged over all process steps (Yna)
Factory A
90%
Yna = (Yrt)1/n
Factory B
95%
94%
89%
96%
“n” is the COMPLEXITY. As process steps or the number
1/2 = 89.5%
Yna = (.801)
of features/functions
increase
98%
• the opportunities for defects usually increase linearly
92%
Yna = (.773)1/5 = 95.0%
• Rolled Thruput Yield decreases
Yrt = 80.1%
Yrt = 77.3%
57
Six Sigma Metrics - Definitions
Sigma Value (Z- Score):
» The sigma value is derived from the probability of a defect
in a process and is used to compare performance across
products or processes
» z-Score is most accurately determined by using the
equation of z = e(-DPU)
58
Z Score

Z Score
» The universal metric used to
compare performance across
products or processes
» Derived from the probability of
producing a defect
• Based upon Normalized Throughput
Yield for complex processes
• Related to comparable Sigma value of
an equivalent normal distribution
Normalized
Thruput
Yield
Defective
 scale
3
Z
59
Z Score

Conversion to Std Normal
» i.e ~N(0,1)
~N(0,1)
3
Z
~N(20,10)
60
Harvesting the Fruit of Six Sigma
Sweet Fruit
Design for Manufacturability
5  Wall, Improve Designs
Bulk of Fruit
Process Characterization
and Optimization
---------------------------------------
4  Wall, Improve Processes
Low Hanging Fruit
Seven Basic Tools
---------------------------------------
We don't know what we don't know
We can't act on what we don't know
We won't know until we search
We won't search for what we don't question
We don't question what we don't measure
Hence, We just don't know
3  Wall, Beat Up Suppliers
Ground Fruit
Logic and Intuition
© 1994 Dr. Mikel J. Harry - V4.0
61
Potential Project Deliverables: Measure Phase







Project definition
» Problem description
» Project Metrics
Project Exploration
» Process Flow Diagram
» C&E Matrix, Process FMEA, Fishbone Diagrams
» Data Collection System
Measurement System Analysis
» Attribute/Variable Gage Studies
Capability Assessment (on each Y)
» Cp, Cpk, Pp, Ppk,  level, DPU, RTY
Graphical & Statistical Tools
Project Summary
» Conclusions
» Issues and Barriers
» Next Steps
Completed local Project Review
62
企業經營循環









企業之生存在於不斷地追求成長。
成長主要決定於顧客滿意度。
顧客滿意取決於交期、品質及價格。
交期、品質及價格是由製程能力所掌握 。
製程能力為變異數所影響 。
製程變異數因不良率、成本與時程增加而變差。
為了消除變異數，我們需要運用正確的知識。
想要運用正確的知識，我們必須先學會它。
唯有不斷地學習與運用新知識，企業才能永續經營。
成長即是硬道理
63
64