Professional Documents
Culture Documents
1
Six Sigma History
1. Motorola (1987)
2. Texas Instruments (1988)
3. Digital Equipment (1989)
4. IBM (1990)
5. ABB (1993)
6. AlliedSignal & Kodak (1994)
7. GE (1995)
… and now, many, many others.
2
What is Six Sigma?
• Philosophy
• Methodology
• Approach
• Goal
• Vision
• Culture
Measure
CUSTOMER E verything
FOCUSED T hat
R esults
I n
C ustomer DATA
S atisfaction DRIVEN
3
Six Sigma Philosophy
Variation Defect
Prevention Continuous
Reduction
Improvement
Define
Characterize Measure
Analyze
Improve
Optimize
Control
“Before we can improve any system one must listen to the voice of the system (the
voice of the process). Then one must understand how the inputs affect the outputs of
the system. Finally, one must be able to change the inputs (and possibly the system) in
order to achieve the desired results. This will require sustained effort, constancy of
purpose, and an environment where continuous improvement is the operating
philosophy.”
Donald J. Wheeler
4
The Six Sigma Methodology
is based on
DATA
• You need baseline data to understand the process
• Don’t trust historical data!
If it’s all that’s available…question it!
• Analysis & improvement plans must be based on data!
• Improvements must be validated with data
• Monitoring & control plans require data
9
• Identify Opportunities
• Map the Process
Improve
(if needed)
• Perform Failure Mode and Effect Analysis
Control
contribution to Outputs
• Incorporate Controls on Critical Variables
5
Six Sigma Approach
6
What is a Defect?
Examples of Defects
7
Variation in Product Characteristics = Function (Variation in Process Parameters)
or
Y=f(x)
Causes of Variation
The variation in the outputs is a function of the variation in the inputs: Y = f(x)
8
The “Language of Variation”
17
6σ 3σ 2σ 1σ µ 1σ 2σ 3σ 6σ
68.26%
95.46%
99.73%
99.99999975%
Areas under the normal distribution curve
9
Exercise
For each of the following samples of data, what shape of
distribution would you expect?
– Rolling a fair die
– Throwing darts at a target
– Dropping a handful of sand on the floor
– Machining a dimension on a lathe
– Measuring a single part many times with a single gage
– Mean time to component failure
– Cycle time to complete a work order
– Defects per length of weld
– Your commute time to work
Note: There are methods of mathematically transforming other types of populations into a normal distribution.
LSL USL
LCL UCL
6σ 6σ
10
What a 2σ process looks like...
Process
Center
LSL USL
308,537 Defects
Per Million LCL UCL
11
What a 6σ process looks like...
Process
Center
LSL USL
3.4 Defects
Per Million LCL UCL
12
What Does Six Sigma Mean
In Your Daily Life ?
10000
Baggage Handling
1000
100
10 Airline
Safety Rate
1
1 2 3 4 5 6 7
Sigma Level
60 2 308,538
Over the Previous Sigma
4.62X
50 3 66,807
40 10.76X
30 4 6,210
26.69X
20 5 232.7
10 68.42X
6 3.4
0
13
99% as a Goal
Real Data
• If 99% were good enough -- this means:
– 2,000 lost articles of mail per hour
– Your car wouldn’t run for about one hour each week
14
Definitions Of Yield
15
Hidden Factory Really Exists
90
80
70 Actual
60 Defects
Defects
50
40
30
20
Inspection
10 Defects
0
0 5 10 15 20 25 30 35 40
Time
16