Introduction To Six Sigma: Pdfmachine Trial Version
Introduction To Six Sigma: Pdfmachine Trial Version
n
io
rs
Introduction to Six Sigma
ve
al
tri
ne
hi
ac
fM
pd
1
© Raza Industrial Engineering Solutions, Islamabad
A NUST Technology Incubation Company
[email protected]
6
Defining Six Sigma
n
io
rs
The simplest definition of Six Sigma is to eliminate
ve
defects / waste and to mistake proof the process that
al
tri
create value for customers
ne
n
io
Mail Delivery
rs
99 % defect free (3.8 ) means 20,000 lost articles
ve
per hour
al
While delivery at 6 (99.99966 % defect free)
tri
ne
ac
Electricity
fM
n
io
Surgeries
rs
99 % defect free (3.8 ) means 5000 errors
ve
al
per week
tri
hi
ac
n
io
Six Sigma emerged as a natural evolution in business
The Current business environment now demands – and
rs
ve
rewards – innovation more than ever before
al
tri
Customer Expectations
ne
Technological Change
Global Competition
hi
ac
Market Fragmentation
fM
n
io
Population is the universe under study. A population
rs
ve
could consist of all invoices sent to customers, all
al
customers inquires seeking technical support, or a
tri
production run
ne
hi
pd
n
io
rs
ve
al
tri
ne
hi
ac
fM
pd
n
io
An inspector checks the resistance value of
rs
ve
five coils and records the values in ohms:
al
tri
3.35, 3.37, 3.28, 3.34, and 3.30. Determine
ne
the average.
hi
ac
n
io
rs
ve
A bell shape curve that follows the
al
following mathematical relation
tri
ne
hi
ac
fM
pd
n
io
rs
ve
The standard normal curve is the normal
al
curve with a mean of zero and a standard
tri
deviation of one.
ne
hi
ac
fM
pd
n
io
rs
ve
al
tri
ne
hi
ac
fM
pd
n
io
rs
ve
al
tri
ne
hi
ac
fM
pd
n
io
rs
ve
al
tri
ne
hi
ac
fM
pd
n
io
rs
ve
al
tri
ne
hi
ac
fM
pd
n
io
What factors does process of shooting
rs
ve
arrows involve ?
al
Selecting the arrow (raw material)
tri
Holding the bow steady and smoothly
ne
hi
(machinery)
pd
n
io
rs
ve
al
tri
ne
hi
ac
fM
pd
n
io
1. Discrete Data (a.k.a. Attribute Data)
rs
Type – 1: One can count both the number of
ve
al
occurrences and non-occurrences
tri
Heads / Tails, Yes / No, Pass / Fail
ne
n
Manufacturing Process: Making Sheets of Glass
io
rs
ve
al
tri
ne
hi
ac
fM
pd
n
io
rs
Performance
ve
CTQ Type Standard Continuous Discrete
al Source
tri
Good/Bad
ne
Pass/Fail
ac
fM
Standards,
Under/Over
Time Customer, Actual Time
pd
Estimate
Quotes/Bids
Quotes/Bids Under/Over
Money Budgets
Actual Cost
Budget
n
Cycle time for a “Credit Check” process
io
1.
rs
ve
2. Percent cream content in milk bottles
3. al
Sales Hit Rate (number of sales proposals that were won)
tri
reported each month
ne
5.
fM
n
io
Define Measure Analyze Improve Control
rs
ve
A. Identify Your al B. Develop a C. Define and Build a
tri
Project’s/ Customer’s
Team Charter High Level Process Map
ne
CTQs
hi
ac
fM
pd
n
io
rs
ve
CTQ means Critical to Quality .What does
al
your customer think is essential and how
tri
do you set-up a project to meet their
ne
hi
desires.
ac
fM
pd
n
io
rs
ve
Y = f (X)
Where:
al
tri
ne
processes
pd
n
io
rs
flow
ve
thinking
al
tri
ne
Requirements Requirements
hi
C O P I S
ac
fM
n
io
Define Measure Analyze Improve Control
rs
ve
1.Select CTQ al 2. Define Performance 3. Establish Data Collection
tri
Characteristics Standards Plan, Validate Measurement
System, and Collect Data
ne
hi
ac
fM
pd
n
io
rs
Functional
ve
Requirements
(HOW’s) Part Characteristics
(HOW’s)
al Manufacturing
tri
Processes
Fundamental Requirements
House
Customer CTQs
(HOW’s)
(WHAT’s)
Of House
ne
Process Variables
(WHAT’s)
(HOW’s)
Part Characteristics
Quality Of House
hi
(WHAT’s)
#1 Quality
Manufacturing Processes
Of
ac
#2 Quality House
fM
(WHAT’s)
#3 Of
pd
Quality
Key #4
Functional Key Part
Requirements Characteristics Key
Manufacturing
Processes Key
Process
© Raza Industrial Engineering Solutions, Islamabad 26
Variables
A NUST Technology Incubation Company
[email protected]
6
Process Mapping
n
io
A tool to represent a process graphically and come
rs
ve
to common understanding of the different factors
al
or steps involved in a process
tri
ne
hi
ac
Process
fM
Input Output
pd
n
io
Purpose
rs
To provide a visual display of all possible causes of a specific problem
ve
When:
al
To expand your thinking to consider all possible causes
tri
To gain group’s input
ne
Cause
Cause
Categories Effect
Effect
fM
pd
Problem
Statement
Causes
n
io
Which one does not belong here?
rs
ve
al
tri
People
ne
hi
ac
Distractions
fM
Insufficient Training
pd
Uniform
Lack of experience
Workload
n
io
Vital Few
rs
ve
al
Frequency or Cost
tri
Trivial Many
ne
hi
ac
fM
pd
Causes
© Raza Industrial Engineering Solutions, Islamabad 30
A NUST Technology Incubation Company
[email protected]
6
Failure Modes and Effects Analysis
n
io
rs
ve
al
tri
Concept
Concept
ne
hi
ac
Identify
Identifyways
waysthe
theproduct
productororprocess
processcan
can
fM
fail.
fail.Then
Thenplan
planto
toprevent
preventthose
thosefailures.
pd
failures.
n
io
rs
ve
al Effect 1
Effect 1
tri
Cause 1
Cause 1
ne
Effect 2
Effect 2
hi
Cause 1 Effect 1
Cause 1 Effect 1
ac
Cause 1
fM
Effect 1
Effect 1
Cause 2
Cause 2
n
io
rs
ve
Operational Definition
al
Target Performance
tri
ne
Specification Limit
hi
ac
Definition of Defect
fM
pd
n
io
rs
Customer: On time means that my flight leaves
ve
the ground at the time printed on my ticket
al
tri
ne
hi
n
io
How can we best measure the airline on-
rs
ve
time performance?
al
Record the number of flights that leaves the
tri
a.
ne
month
ac
fM
n
io
rs
ve
Specification Limit
al
Target Performance
tri
ne
Definition of Defect
hi
ac
fM
n
io
rs
ve
Lower Spec. Limit
al Target Upper Spec. Limit
tri
ne
hi
ac
9:30 9:45
n
io
rs
ve
al
Lower Spec. Limit / Upper Spec.
tri
Target Limit
ne
hi
Defect Defect
ac
9:30 9:45
fM
pd
n
io
Define Measure Analyze Improve Control
rs
ve
al
4. Establish Process 5. Define Performance 6. Identify Variation
tri
Capability Objectives Sources
ne
hi
ac
fM
pd
n
io
rs
Which route will minimize your chances of
ve
arriving late? You have 35 minutes to reach
al
office.
tri
ne
Choose:
pd
n
io
rs
Obs. No. Route 1 Route 2
ve
1 25 20
al 2 27 20
tri
3 23 20
ne
4 25 20
hi
5 25 21
ac
6 29 25
fM
7 21 15
pd
8 25 13
9 27 14
10 23 40
So on
n
io
rs
ve
30
Mean StDev
al 25.4 4.3
tri
ne
Frequency
20
hi
ac
fM
10
pd
20 Route 1 30
© Raza Industrial Engineering Solutions, Islamabad
USL 40
42
A NUST Technology Incubation Company
[email protected]
6
Variation and Standard Deviation
n
io
rs
ve
30
Mean StDev
al 20.6 7.4
tri
ne
Frequency
20
hi
ac
fM
10
pd
20 Route 2 30
© Raza Industrial Engineering Solutions, Islamabad
USL 40
43
A NUST Technology Incubation Company
[email protected]
6
Process Capability
n
io
rs
ve
Process capability measures the ability of
al
your process to meet customer
tri
requirements. It compares process
ne
hi
variability.
pd
n
io
rs
ve
al
tri
ne
hi
ac
fM
pd
n
io
rs
ve
al
tri
ne
hi
ac
fM
pd
n
io
rs
ve
al
tri
ne
hi
ac
fM
pd
n
io
rs
ve
al
tri
ne
hi
ac
fM
pd
n
io
rs
ve
al
tri
ne
hi
ac
fM
pd
n
io
rs
ve
al
tri
ne
hi
ac
fM
pd
n
Classical Yield = YC= ¾ =75%
io
rs
Classical yield is the number of defect-free parts for the whole
ve
process divided by the total number of parts inspected. If we say
the yield is ¾ or 75%, we lose valuable data on the true
al
tri
performance of the process. This loss of insight becomes a barrier
ne
to process improvement.
First time Yield = YFT= ¼ = 25%
hi
ac
total number of parts inspected for the first time. If we say the yield
is ¼ or 25%, we are really talking about the First Time Yield
pd
n
io
rs
ve
al
tri
ne
hi
ac
fM
pd
n
io
rs
We want our performance objective to
ve
exceed process entitlement. What does
al
tri
this mean?
ne
1.
ac
in our process
pd
n
io
rs
Y = f (X) To get result, should we focus
ve
our behavior on the result Y or X?
al
tri
• Y • X1 . . . Xn
ne
• Independent
hi
• Dependent
ac
• Output • Input-Process
fM
• Effect • Cause
pd
• Symptom • Problem
• Monitor • Control
Historically the Y, … with Six Sigma the Xs
© Raza Industrial Engineering Solutions, Islamabad 54
A NUST Technology Incubation Company
[email protected]
Improve Phase
6
n
io
Define Measure Analyze Improve Control
rs
ve
al
7. Screen Potential 8. Discover Variable 9. Establish Operating
tri
Causes Relationships Tolerances
ne
hi
ac
fM
pd
n
io
Types of DOE
rs
ve
Screening DOE is used to identify the vital
X’s for the CTQ
al
tri
ne
n
io
rs
ve
Budget for testing
al
Full factorial or fractional factorial design
tri
ne
Availability of personnel
hi
ac
n
io
Define Measure Analyze Improve Control
rs
ve
10. Define and Validate
al 11.Determine 12. Implement
Measurement System on
X’s in Actual Application
tri
Process Capability Process Control
ne
hi
ac
fM
pd