Quality Tools

MTSU

The PDSA Cycle

Plan
Act
Do
Study

MTSU

The Process Improvement

Cycle
Select a
process
Document
Study/document
Evaluate

Seek ways to
Improve it

Implement the
Improved process

MTSU

Design an
Improved process

Process Improvement
Tools
There are a number of tools
that can be used for problem
solving and process
improvement
Tools aid in data collection and
interpretation, and provide the
basis for decision making
MTSU

Seven Basic Quality Tools

1.
2.
3.
4.
5.
6.
7.
MTSU

Check sheets
Flowcharts
Scatter diagrams
Histograms
Pareto analysis
Control charts
Cause-and-effect diagrams
5

Check Sheet
Billing Errors

Monday

Wrong Account
Wrong Amount

A/R Errors
Wrong Account
Wrong Amount

MTSU

Flowchart

Process

Good?

Process

Good?

Process

Process

MTSU

Scatter Diagram

Variable B

Variable A
MTSU

Histogram

frequency

MTSU

80%
80%of
of the
the
problems
problems
may
may be
be
attributed
attributed to
to
20%
20%of
of the
the
causes.
causes.

MTSU

Number of defects

Pareto Analysis

Off
Smeared Missing Loose Other
center print
label
10

Control Chart
1020

UCL

1010
1000
990

LCL

980
970
0

MTSU

9 10 11 12 13 14 15

11

Cause-and-Effect Diagram
Methods
Cause

Environment

Materials
Cause
Cause

Cause

Cause

Cause
Cause

People

MTSU

Cause

Cause

Cause
Cause

Effect

Cause

Equipment

12

Tracking Improvements
UCL

UCL

LCL
LCL

Process centered
Process not centered and stable
and not stable

MTSU

UCL

LCL
Additional improvements
made to the process

13

Seven Management Tools

1.
2.
3.
4.
5.
6.
7.
MTSU

Affinity Diagram
Interrelationship Diagraph
Tree Diagrams
Prioritization Matrices
Matrix Diagram
Process Decision Program Chart
Activity Network Diagram
14

Methods for Generating

Ideas
Brainstorming
Quality circles
Interviewing
Benchmarking
5W2H

MTSU

15

A Brief History of Quality

MTSU

16

America Re-discovers
Deming

MTSU

17

 and Juran

MTSU

18

Rediscovering the Gurus

Deming
Emphasis on Statistical Control
14 Points for Management

Juran
Quality Planning and Analysis
Managerial Breakthrough
Quality Control Handbook
MTSU

19

Sta
ti

MTSU

stic
s

20

MTSU

21

MTSU

22

MTSU

23

MTSU

24

MTSU

25

MTSU

26

MTSU

27

MTSU

28

MTSU

29

Sta
ti

MTSU

stic
s

30

MTSU

31

What Types of Companies

Can Benefit from Six
Sigma?

Companies that benefit from structured

organizational improvement
Companies that need to improve customer
satisfaction
All types of companies can benefit:

MTSU

Manufacturing
Service
Non-profit
Educational
32

What is Six Sigma?

A. Customer Focus Focus on what
is critical to customers
B. Data Driven Extensive use of
statistical tools
C. Robust Methodology Tools plus
implementation methods to make
success more likely
MTSU

33

What is Six Sigma?

Key Concepts
1. Critical to Quality: What attributes
are most important to the
customer? (CTQ, CTC, CTD)
2. Defect: Failing to deliver what the
customer wants / expects (DPMO)
3. Variation: The level of
unpredictability the customer
experiences
MTSU

34

What is Six Sigma?

Key Concepts
5. Process Capability: What your
process can deliver consistently
6. Stable Operations: Stable ops are
predictable
7. Design for Six Sigma: Designing
to meet customer needs and
process capability
MTSU

35

DMADV - DMAIC
Existing Processes

New Processes

Measure

Design

Analyze

MTSU

Define

Control

Define

Verify

Measure

Improve

Analyze

36

Statistics Lite
Centered Process

LSL

MTSU

USL

37

Statistics Lite
Centered Process

LSL

MTSU

USL

38

Statistics Lite
Centered Process

LSL

MTSU

USL

Non-conforming
Product
1,300 DPMO

39

Statistics Lite
Centered Process

LSL

MTSU

USL

Non-conforming
Product
2,600 DPMO

40

Statistics Lite
Shifted Process

LSL

USL

1.5 mean shift

MTSU

41

Statistics Lite
Shifted Process

LSL

USL

Non-conforming
Product
66,800 DPMO

1.5 mean shift

MTSU

42

Statistics Lite
Shifted Process
Cost to your company 15-30%
of sales
LSL

USL

Non-conforming
Product
66,800 DPMO

1.5 mean shift

MTSU

43

Cost of Poor Quality

5-8% of
Sales

Inspection
Overtime
Downtime
Rejects
Rework
Lost sales

Long cycle times

Cost of Capital
Redundant Operations
Expediting costs

(less obvious)

MTSU

Late delivery

Lost Opportunity

Lost Customer Loyalty

15-22%
of Sales
Inaccurate Reports

Excessive Planning

44

Statistics Lite
Centered 6Process

LSL

MTSU

USL

45

Statistics Lite
Shifted 6Process

LSL

USL

1.5 mean shift

MTSU

46

Statistics Lite
Shifted 6Process

LSL

USL

Non-conforming
Product
3.4 DPMO

1.5 mean shift

MTSU

47

Why Six Sigma?

99% (3.8 Sigma)

99.99966% (6 Sigma)

20,000 lost articles of mail per

hour

Seven articles lost per hour

5,000 incorrect surgical

operations per week

2 incorrect 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

70 wrong prescriptions per year

No electricity for almost seven

hours each month
MTSU

One hour without electricity every

48
34 years

Six Sigma
Breakthrough Strategy
Define
Measure
Analyze
Improve
Control

MTSU

49

DMADV - DMAIC
New Processes

Define

Verify

Measure

Design

Analyze

MTSU

Existing Processes

Define

Control

Measure

Improve

Analyze

50

Define - Selecting Projects

The project must relate to
customer satisfaction
The projects results must reduce
defects by some threshold amount
The project should achieve some
threshold of cost savings.

MTSU

51

Criteria for Project

Selection
Does it involve recurring
events?
Is the scope narrow?
Do measures exist?
Do you have control of the
process?
MTSU

52

Define Phase, Continued

If these criteria are met then:
1) Identify the customers involved, both
internal and external to the function.
2) Find out what the customers CTs
are
(Critical to Quality, Critical to
Delivery,
Critical to Cost, etc).
3) Define the project scope and goals.
4) Map the process to be improved.
MTSU

53

Define Phase - Tools

Project Charter
Stakeholder
Analysis
Affinity Diagram
SIPOC
Voice of the
Customer
CT Tree
MTSU

Kano Model
SWOT Analysis
Cause-and-Effect
Diagrams
Supplier
Segmentation
Project
Management
54

Charter
Problem statement
Business case
Goals, milestones,
success criteria, &
deliverables
Project scope /
boundaries
Roles & responsibilities
Stakeholder support /
approval needed

MTSU

55

Define Phase

Charter Development
Charter An agreement between
management and project team
members about what the team will
accomplish.

MTSU

56

Charter: Providing
Direction
You have to be
careful if you
dont know
where youre
going, ...
because you
might not get
there.
MTSU

57

Define Phase

Charter What it does

Clarifies expectations (what and
why)
Keeps team focused
Reduces tampering
Reduces wandering
Reduces goal creep

Transfers ownership from

management to team
MTSU

58

Define Phase

Charter What it does

Provides overview of purpose
Describes why you are working on
this project (business case)
Defines scope of project
Determines deliverables
Defines measures of success
Determines resources available
MTSU

59

Business Case
Potential
Improvement
Improve quality
Improve OTD
Select better suppliers
Implement rating
system
MTSU

Potential Impact
Reduce cost, inventory
Shortages , inventory

Q, $, LT, reduce
inventory
Improve supply
efficiency, better
60
suppliers

Define Phase - Tools

Project Charter
Stakeholder
Analysis
Affinity Diagram
SIPOC
Voice of the
Customer
CT Tree
MTSU

Kano Model
SWOT Analysis
Cause-and-Effect
Diagrams
Supplier
Segmentation
Project
Management
61

SIPOC Example
Suppliers

Inputs

Processes Outputs Customer

s

Ops Mgt

Supplier
Perf.

Survey

Ops Mgt

Buyers

Complain
ts

Rating
system

Buyers

Engrg.

Tech
Reqts

Mfg.

Rating
system

MTSU

Supplier
Evaluation Improve
d
Supplier
Perform.

Engineerin
g

Commit. Mfg.
to
62
suppliers

SWOT Analysis

MTSU

Positive

Negative

Internal

Strengths

Weaknesses

External

Opportunitie
s

Threats
63

Define Outputs
Once completed, the Define Phase
should answer the following
questions:
1)
2)
3)
4)
5)
MTSU

Who is the customer?

What matters?
What is the scope?
What defect am I trying to reduce?
What are the improvement targets?
64

Define Phase

Toll Gate Review

Submit project storyboard to
sponsor
Toll Gate review presentation
Sponsor provides feedback
Project corrections are made prior
to proceeding to the next phase

MTSU

65

The Measure Phase

Purpose
To collect current performance of the
process identified in the Define phase
This data is used to determine
sources of variation and serve as a
benchmark to validate improvements

MTSU

66

Measurements
Benefits of having good data need
to outweigh the costs of getting it

What does this measure do for the

Project?

MTSU

67

The Measure Phase

Upon completion of the measure phase,
Project Teams will have:
A plan for collecting data that specifies the
type of data needed and techniques for
collecting the data
A validated measurement system that
ensures the accuracy and consistency of
the data collected
A sufficient data set for problem analysis
MTSU

68

Measure - Key Concepts

Measurement
Variation
Exists naturally in any process and is the
reason Six Sigma projects are undertaken

Data
Data Collection Plan
Measurement System Analysis
Ensures measurement techniques are
reproducible and repeatable
MTSU

69

Recording Measurements
3 stages
The output stage
These tell how well customer needs are
being met

Parts of the process

These are taken at critical points in the
process

The input stage

These evaluate contributions to the process
that are turned into value for the customer
MTSU

70

Recording Measurements
Output Stage

MTSU

Shortages
Line shutdowns
Quality discrepant material
Material price variances
Internal customer survey

71

Recording Measurements
Parts of the process

MTSU

Project milestones
Supplier ship on time performance
Supplier OTD
Supplier internal throughput yield
Supplier suggested cost reductions

72

Recording Measurements
The input stage

MTSU

Supplier base size

% Buyers with degrees
% of spend covered by LTCs
% of spend from reverse auction
Supplier FMEAs

73

Operational Definitions
Walter Shewhart, the inventor
of statistical process control,
believed his work on
operational definitions to be of
greater importance than his
work on control charts.
MTSU

74

Operational Definitions
An operational definition is a

procedure agreed upon for

translation of a concept into
measurement of some kind.

Deming, 1986

Operational definitions should be

valid and reliable.
MTSU

75

Operational Definitions?

Slump, I aint in no
slump. I just aint
hitting.

MTSU

76

Operational Definitions
On-Time Delivery
On-Time Payment
Late
Defective
Clean
Good communication
Engineering support
MTSU

Do you, your management, and your

suppliers agree on these definitions? 77

Determining Data Type

What do we want to know?
Review materials developed during
design phase
What characteristics do we need to
learn more about?
MTSU

78

Data Collection Plan

What data will be collected?
Why is it needed?
Who is responsible?
How will it be collected?
When will it be collected?
Where will it be collected?
MTSU

79

Measurement System
Analysis
After Data Collection Plan is
complete, it needs to be verified
before actual data is collected
MSA is performed on a regular basis
MSA ends when a high level of
confidence is reached that the data
collected accurately depicts the
variation in the process
MTSU

80

Analyze Phase

The analyze phase allows the

Project Team to target
improvement opportunities by
taking a closer look at the data.

MTSU

81

Analyze Phase
Capability Analysis - establishing current
performance level
Graphical Analysis - a visual indication
of performance using graphs
Root Cause Analysis developing a
hypothesis about the causes of variation
Root Cause Verification verifying that
the planned action will generate the
desired improvement
MTSU

82

Process Capability
When selecting a process to
perform an operation, the inherent
variability of process output should
be compared to the range or
tolerances allowed by the
designers specifications.

MTSU

83

Process Capability
process distribution
Lower
Specification

Upper
Specification

Much of the process output

fits within specification width

Almost all of the process output

fits within the specification width

In other words, is the

process capable of
producing the item
within
specifications?

A significant portion of the process output

MTSU
falls outside of the specification width

84

Analyze Phase

Cause Hypotheses
Identifying Obvious Process Problems

MTSU

Disconnects
Bottlenecks
Redundancies
Unnecessary distance
Rework
Decision points

Process Map Review

85

Analyze Phase

Cause Hypotheses
Quantifying Value-Added Steps
Value-Adding
Process Map Review
Value-Enabling
Non-Value-Adding

MTSU

86

Analyze Phase

Cause Hypotheses
Process Time Analysis

MTSU

Work time (value, but often only 5%)

Wait (& queue) time (usually dominates)
Setup time (tremendous leverage)
Move time (process dependent)

87

Analyze Phase
Tools

Brainstorming . . . And beyond!

Process Maps
Cause-and-Effect Diagrams
Focused Problem Statement
Statistical Tools

MTSU

88

Cause-and-Effect Diagrams
Supplier Failure - Causes
Environment

Equipment
Business down
Strike
Flood
Earthquake

Lawsuit
Fire
Business up

People
Incapable equip

New employees

Tornado

Lack of PM
Process upset
Bad specs
Different standards
MTSU
Measurement

Processes

Strike
Fraud
Incompetence
Turnover
Illness

New source
Commodity allocation
Bankruptcy

Inaccurate PO
Materials

Mat'l price increase

89

Elements of Improve
Phase

MTSU

Generate Improvement
Alternatives
Create a Should Be Process Map
Conduct FMEA
Perform Cost/Benefit Analysis
Pilot
Validate Improvement
90

Improve Phase
Elements

MTSU

Generate Improvement
Alternatives
Create a Should Be Process Map
Conduct FMEA
Perform Cost/Benefit Analysis
Pilot
Validate Improvement
91

Failure Modes Effects

Analysis
Identify failure modes How can this
product or process fail?
Identify failure effects What
happens when this failure occurs?
Identify potential causes of the
effects & their probability of
occurring.
Rate the likelihood of detecting the
occurrence.
MTSU

92

FMEA - Output
Ranked list of products that
contribute to risk
List of actions and persons
responsible for addressing the risk
Revised ranked priority list

MTSU

93

Improvement Phase
Methods
UCL

UCL

LCL
LCL

Process centered
Process not centered and in control
and not in control

MTSU

UCL

LCL
Evidence of additional
improvements

94

Review of Improvement
Phase
Generate Improvement
Alternatives
Create a Should Be Process Map
Conduct FMEA
Perform Cost/Benefit Analysis
Pilot
Validate Improvement
MTSU

95

Generating Improvement
Alternatives

MTSU

Define Improvement Criteria

Generate Possible Improvements
Evaluate Improvements and Make
Best Choice

96

Pilot
Benefits of Pilot
Determine best way to implement the
improvement
Lowers risk of failure
Increases opportunity for feedback
Obtain buy-in from affected personnel
Provides opportunity to revise the
improvement before full
implementation
MTSU

97

Stakeholder Analysis
People or
Groups
Level of
Commitment

Buy

Mfg

Eng

Enthusiastic Support
Help it work
Compliant
Hesitant

Indifferent
Uncooperative
MTSU

Opposed

X
98

Review of Implementation
Phase
Generate Improvement
Alternatives
Create a Should Be Process Map
Conduct FMEA
Perform Cost/Benefit Analysis
Pilot
Validate Improvement
MTSU

99

Control Phase
Why is it important?
The Control Phase begins as the
project team tries to eliminate
errors by Mistake Proofing
their improvement alternative.
Mistake Proofing attempts to
eliminate the opportunities for
error.
MTSU

100

Control Phase
Why is it important?
Mistake Proofing tries to make it
impossible for an operation to be
performed incorrectly, and/or
correct errors before they are
passed to the next worker, where
they might become a defect.

MTSU

101

Control Phase #2
During the Control Phase the Project
team will:
1) Develop a plan to make sure the
measurement system will remain
relevant over the long term.
2) Establish Control Charts the process
owner will use to manage the process.
3) Create a Reaction Plan to address
situations that might cause the process to
move out of control.
MTSU

102

Control Phase #3
The Control Phase ends when:
1) Standard Operating Procedures have
been updated.
2) Process Operators, the people who do
the job, have been trained for the new
process.

Once completed, the Control Phase

should sustain the gains the project
made while implementing ongoing
process controls.
MTSU

103

Control Phase #4
When is a project complete?
1) When other Black Belts can see
the ongoing controls work
2) When the customer sees the
results
3) When the business sees the
money.

MTSU

104

Six Sigma
Six Sigma People
Executives
Champions (deployment, project)
Master Black Belts
Black Belts
Green Belts

MTSU

105

Control Phase
Methods

During the Control Phase the Project

team will:
1) Develop a plan to make sure the
measurement system will remain
relevant over the long term.
2) Establish Control Charts the process
owner will use to manage the process.
3) Create a Control Plan to address
situations that might cause the process to
move out of control.
MTSU

106

Statistical Process Control

(SPC)
Chance variations are the many sources of
variation within a process that is in
statistical control. They behave like a
constant system of random chance causes.
If only natural causes of variation are
present, the output of a process forms a
distribution that is stable over time and is
predictable.

MTSU

107

Statistical Process Control

(SPC)
Assignable variation in a process can
be traced to a specific reason.
Machine wear
Misadjusted equipment
Fatigued or untrained workers

If assignable causes of variation are

present, the process output is not
stable over time and is not predictable.

MTSU

108

SPC - Assignable Causes

The operational definition of
assignable variation is variation
that causes out-of-control points
on a control chart.

MTSU

109

Natural Patterns or Variations

Natural patterns exhibit the following characteristics:
Most of the points are near the centerline.
A few points spread out and approach the control
limits.
None (or only on rare occasions) of the points
exceeds the control limits.
Reference: Statistical Quality Control Handbook, Western Electric

MTSU

110

Unnatural Patterns or
Variations
Unnatural patterns exhibit the following characteristics:
Absence of points near the centerline produces a
pattern known as a mixture.
Absence of points near the control limits produces an
unnatural pattern known as stratification.
Presence of points outside of the control limits
produces an unnatural pattern known as instability.
Reference: Statistical Quality Control Handbook, Western Electric

MTSU

111

Tests for Unnatural Patterns

Instability
A single point falls outside of the 3 sigma control
limits.
Two out of three successive points fall in the outer one third
of the control limits.
Four out of five successive points fall in the outer two thirds
of the control limits.
Eight successive points fall on one side of the centerline.

Systematic variable
A long series of points are high, low, high, low without
interruption.
Reference: Statistical Quality Control Handbook, Western Electric

MTSU

112

Statistical Process Control

Why use averages?

To create a normal distribution

Averages are more sensitive to
change than individuals

MTSU

113

Central Limit Theorem

Simulation
The distribution of
a sample
approaches
normal even
when the parent
population is not
normally
distributed.
MTSU

114

Control Phase
Mean and Range Charts
Process
Distribution

process mean drifting upward

UCL

Drift Detected

x-Chart
LCL
UCL

No drift detected

R-chart
LCL

MTSU

115

Control Phase
Mean and Range Charts
Process
Distribution

process variability increasing

UCL

x-Chart

No shift detected
LCL
UCL

Increase detected

R-chart
LCL

MTSU

116

Statistical Process Control

Tolerance or specification limits
Defined by an engineer
Related to product design
requirements

Control limits
Defined by the process
Related to the variation in the process
Unrelated to product needs
MTSU

117

Control Phase
Control Plans

Structured approach for designing

value added control methods
Control actions necessary to
ensure output quality
May include controls anywhere in
the process

MTSU

118

Control Phase
Radar Chart
Product Features -

Data Entry Errors

MTSU

Days Late

Savings by Project

119

Control Phase
Radar Chart
Defect Levels by Supplier

MTSU

120

Final Toll Gate Review

Submit project storyboard to sponsor
Toll Gate review presentation
Did the team carry out the agreed upon
tasks?
Did the team achieve the desired results?

Sponsor provides feedback

Project corrections are made prior to
termination or proceeding to next cycle
MTSU

121

Why TQM Fails

Lack of Top Management Commitment
Commitment
Up Front
Appropriate
Training

Wasted Education & Training

Lack of Short-Term Results

Biz Case
Threshold

Failure to Empower Employees

Charter

Brown, Hitchcock, and Willard - 1994

MTSU

122

Why Six Sigma Fails

(and it does)
Lack of Top
Management
Commitment
Interdepartmental /
cross-functional issues
Communication
People

MTSU

123

Quality Tools

Kimball Bullington, Ph.D.

MTSU

124

References
Books:

Six Sigma Pocket Guide (Rath & Strongs)

The Black Belt Memory Jogger (GOAL / QPC)
Six Sigma (Harry and Schroeder)
Implementing Six Sigma (Breyfogle)
The Six Sigma Way Team Fieldbook (Pande, et al)
The Vision of Six Sigma: A roadmap for breakthrough
(Harry)
Why TQM Fails and What To Do About It (Brown, Hitchcock,
& Willard)

Web sites:

MTSU

www.isixsigma.com
www.ge.com/sixsigma
www.asq.org
www.aiag.org

125