MATERI 6

SIX SIGMA
Chapter 14
Six-Sigma Management and
Tools
S. Thomas Foster, Jr.
Boise State University
PowerPoint
prepared by
Dave Magee
University of Kentucky
Lexington Community College

©2004 Prentice-Hall
 Chapter Overview
Slide 1 of 2

 What is Six Sigma?

 Organizing Six Sigma
 DMAIC Overview
 Define Phase
 Measure Phase
 Analyze Phase
 Improve Phase
 Control Phase
 Chapter Overview
Slide 2 of 2

 Taguchi Design of Experiments

 Background of the Taguchi Method
 The Taguchi Process
 Design for Six Sigma
 Lensing Six Sigma From A Contingency Perspective
Differences of Six Sigma from Traditional
Continuous Improvement
 Six Sigma represents a well thought out packaging
of quality tools and philosophies in an honest
effort to provide rigor and repeatability to quality
improvement efforts.
 Six Sigma is much more cost reduction-oriented
than traditional continuous improvement.
 Six Sigma is organized around creating
champions, black belts, green belts, and in some
situations, yellow belts.
What Is Six Sigma?
Slide 1 of 4

 Sigma stands for the Greek symbol σ that

designates a standard deviation in statistics.
 Six refers to the number of standard deviations
from a mean specifications should be.
 Began at Motorola in 1982 as an effort to reduce
costs and improve quality.
 It now involves planning, organization, training,
human resources planning, and pay-for-
knowledge. It requires both organizational and
individual cooperation to achieve a goal.
What Is Six Sigma?
Slide 2 of 4

Six-Sigma Variation Figure 14.4

σ σ σ σ

3-sigma 6-sigma
producing some producing almost
product out of no product out of
specification specification
What Is Six Sigma?
Slide 3 of 4

Sigma Levels and ppm Defects

Sigma Level Long-Term ppm* Defects

1 691,462
2 308,538
3 66,807
4 6,210
5 233
6 3.4
*ppm = parts-per-million
What Is Six Sigma?
Slide 4 of 4

Six Sigma Effectiveness

Outside specialists
(Less than 1%)

Six Sigma
(Less than 90%)

Basic Tools of Quality

(90%)
Organizing Six Sigma
Slide 1 of 5

 Cost of training can run $10,000 to $16,000 for a

single black belt
 Expected returns from projects can run into the
$100’s of thousands
 Key Players in Six Sigma Efforts
 Champion
 Master black belt
 Black belt
 Green belt
 Yellow belt
Organizing Six Sigma
Slide 2 of 5

 Champion
 Job is to work with black belts and potential black belts
to identify possible projects.
 Provides continuing support for the project and validates
the results at the end of the project.
 Would probably be the CEO in smaller companies and a
senior VP in larger companies.
 Master Black Belt
 Experienced black belts who serve as mentors and
trainers for new black belts.
 Brings training in-house and can reduce costs.
Organizing Six Sigma
Slide 3 of 5

Champion Decision Making Figure 14.3

VOC VOE Financial

VOB Strategic
Data
Direction
IDEAS

Scarce
Scarce Champion
Resource
Resource

Blackbelt Projects
Organizing Six Sigma
Slide 4 of 5

 Black Belt
 Key to Six Sigma
 Specially trained individuals who are committed full-
time to completing cost reduction projects.
 Training usually lasts about 4 months.
 Each project lasts from 2 months to a year depending on
the project scope.
 Individuals usually spend about two years as a black belt
and are then moved into management jobs.
Organizing Six Sigma
Slide 5 of 5

 Green Belt
 Trained in basic quality tools and work in teams to
improve quality.
 Assigned part-time to work on process and design
improvement.
 In some cases, activities are the same as black belts.
 In other companies, green belts are involved in less
critical projects.
 Yellow Belts
 Some companies have employees who are familiarized
with improvement processes.
DMAIC Overview
Slide 1 of 2

 DMAIC Process
 Very similar to the PDCA cycle.
 DMAIC Phases
 Define
 Measure
 Analyze
 Improve
 Control
DMAIC Overview
Slide 2 of 2

Overview of Six Sigma Process Figure 14.4

Define Measure Analyze Improve Control

•Business
•BusinessCase
Case •Solution
•Project
•ProjectDesirability
•Graphical
•GraphicalData Analysis •SolutionSelection
DesirabilityMatrix
Matrix Data Analysis
•Solution
Selection
•Problem/Objective
•Problem/ObjectiveStatement
•Multi-Vari
•Multi-VariAnalysis
Analysis •SolutionImplementation
Implementation
Statement •Inferential Statistics •Control Plans
•Primary/Secondary Metric •Inferential Statistics •Control Plans
•Primary/Secondary Metric •Control
•Change Management
•Change Management
•Hypothesis
•HypothesisTests
Tests •ControlCharts
Charts
•VOC/QFD •Correlation •Hypothesis
•HypothesisTests
•VOC/QFD •Correlation Tests
•SIPOC •Regression •Process Capability
•SIPOC •Regression •Process Capability
•Process •Process
•ProcessModeling and Simulation Assessment
•ProcessMapping
Mapping Modeling and Simulation Assessment
•Best Practice
•Best Practice
Sharing/Translation
Sharing/Translation
•XY Matrix
•XY Matrix
•Process FMEA •Hypothesis Tests
•Process FMEA •Hypothesis Tests
•Basic
•BasicStatistics •Design
Statistics
•Decision-Making •DesignofofExperiments (DOE)
Experiments (DOE)
•Decision-Making •Advanced Regression
•Advanced Regression
•Non-Normal Data Graphical Analysis •Process
•Non-Normal Data Graphical Analysis
•Measurement System Analysis •ProcessModeling
Modelingand
andSimulation
Simulation
•Measurement System Analysis •Pilot and Test
•Pilot and Test
•Process Capability
•Process Capability
Define Phase
Slide 1 of 7

 Define Phase
 Projects are identified and selected.
 Project selection is performed under the direction and
with the participation of the champion
 Master black belts and black belts or green belts are also
involved in selection.
 Phases of the Define Phase
 Developing the business case.
 Project evaluation
 Pareto analysis
 Project definition
Define Phase
Slide 2 of 7

 Developing the Business Case includes:

 Identifying a group of possible projects
 Writing the business case
 Stratifying the business case into problem statement and
objective statement.
Define Phase
Slide 3 of 7

 The mnemonic device “RUMBA” is used to check

the efficacy of a business case.
 Realistic - Are the goals attainable, is the time line
feasible?
 Understandable - Do I understand the case?
 Measurable - Do we show the measures?
 Believable - That is a lot of money. Can it be done?
 Actionable - Can it be implemented?
 If the business case meets all of these
requirements, it probably will be a good project.
Define Phase
Slide 4 of 7

 Project Evaluation
 There are several methods for evaluating a project
including risk and return assessment.
 Risk and Return Assessment
 Risk assessment evaluates a potential project across
several dimensions to establish an overall risk factor for
the project which will be used to determine the
attractiveness of the project.
 Project return assessment evaluates a potential project on
3 dimensions - growth, urgency, and impact.
 The risk and return scores are placed on a grid to
determine the attractiveness of the potential project.
Define Phase
Slide 5 of 7

Project Risk and Return Figure 14.7

100
90
80
70
Stars ?
Return Factor

60
50
40
30 Low hanging
Dogs
20 fruit
10

0
0 10 20 30 40 50 60 70 80 90 100
Risk Factor
Define Phase
Slide 6 of 7

 Pareto Analysis
 Part of the responsibility of the champion is to perform a
cost of poor quality (COPQ) analysis.
 This is based on the PAF categorization of costs
discussed in Chapter 4.
 Performing a study of internal and external failure costs
will help to determine where the most benefit can be
found.
 Problem Definition
 Project definition consists of a problem statement,
project goals/objectives, primary metrics, secondary
metrics, and team member identification.
Define Phase
Slide 7 of 7

Problem Definition Figure 14.9

Problem
ProblemStatement:
Statement:InIn2002, plant A lost $5.6
2002, plant A lost $5.6
million
millionononCOPQ.
COPQ.OfOfthis,
this,almost
almost$3.5
$3.5million
millionoccurred
occurredinin
operation
operation 2. This has resulted in a loss of profitabilityfor
2. This has resulted in a loss of profitability for
the
thefirm.
firm.

Project
ProjectGoals/Objective:
Goals/Objective:Reduce
ReduceCOPQ
COPQfor
for
operation
operation22by
by30%
30%by
byyear-end.
year-end.

Primary
PrimaryMetrics:
Metrics:- -COPQ
COPQ
- Rework (% of sales)
- Rework (% of sales)
- Scrap (% of sales)
- Scrap (% of sales)

Secondary
SecondaryMetrics:
Metrics: - -Downtime
Downtimefor
forprocess
process
- Plant sales
- Plant sales
- -Labor
Laborproductivity
productivity

Team
TeamMembers:
Members:Pat
PatShannon
Shannon
Phillip Fry
Phillip Fry
Lyman Gallup
Lyman Gallup
Jerry LaCava
Jerry LaCava
Measure Phase
Slide 1 of 4

 Two Major Steps in the Measure Phase

 Selecting process outcomes
 Verifying measurements
 Measure Phase Tools
 Process Map
 XY Matrix
 FMEA
 Gage R&R
 Capability Assessment
Measure Phase
Slide 2 of 4

 Selecting Process Outcomes

 To define process outcomes, you first need to
understand the process. This involves process mapping.
A process map is a flowchart with responsibility. The
goal with a process map is to identify non-value added
activities.
 Two important measures that are monitored are defects
per unit (DPU) and defects per million opportunities
(DPMO).
 The XY matrix is used to identify inputs (X’s) and
outputs (Y’s) from a project you have mapped and are
desiring to pursue.
Measure Phase
Slide 3 of 4

 Verifying Measurements
 It is necessary to use gauges, calipers and other tools
when measuring critical characteristics of processes.
 Measurement system analysis (MSA) is used to
determine if measurements are consistent.
 Product and process capability analysis is another
approach for verifying measurements.
 Gage Repeatability and Reproducibility Analysis (Gage
R&R) is the most commonly used MSA.
Measure Phase
Slide 4 of 4

 Verifying Measurements (continued)

 Reasons for problems in measurement
 The measurement gauges are faulty.
 Operators are using gauges improperly.
 Training in measurement procedures is lacking.
 The gauge is calibrated incorrectly.
 Statistical experiments using analysis of variance
(ANOVA) are useful in performing Gage R&R.
 Two-way ANOVA is used to determine whether
variation comes from the part being measured,
differences in operator measurements, or from the
measurement instrument.
Analyze Phase
Slide 1 of 4

 Analyze Phase
 Involves gathering and analyzing data relative to a
particular Black belt project.
 Steps in the Analyze Phase
 Define your performance objectives.
 Identify independent variables (X’s).
 Analyze sources of variability.
Analyze Phase
Slide 2 of 4

 Defining Objectives
 Attempting to determine what characteristics of the
process need to be changed to achieve improvement.
 Capability analysis is reviewed to determine where the
processes are incapable and prioritized in order of
importance.
Analyze Phase
Slide 3 of 4

 Identifying X’s
 This involves identifying independent variables where
data will be gathered.
 These variables significantly contribute to process or
product variation.
 Primary tools used in identifying X’s
 Process maps
 XY matrices
 Brainstorming
 FMEA
Analyze Phase
Slide 4 of 4

 Analyzing Sources of Variation

 The goal of this step is to use visual and statistical tools
to better understand the relationships between dependent
and independent variables (X’s and Y’s) for use in future
experimentation.
 Tools for Analyzing Sources of Variation
 Histograms
 Box plots
 Scatter plots
 Regression analysis
 Hypothesis tests
Improve Phase
 Improve Phase of the DMAIC Process
 Involves off-line experimentation.
 Off-Line Experimentation
 Involves studying the variables we have identified and
using analysis of variance to determine whether these
independent variables significantly affect variation in our
dependent variables.
 The Taguchi method is an important method for
performing off-line experiments.
Control Phase
 Control Phase of the DMAIC Process
 Involves managing the improved process using process
charts.
 Process charts were covered in Chapters 12 and 13
Taguchi Design of Experiments
Slide 1 of 3

 Taguchi Method
 A standardized approach for determining the best
combination of inputs to produce a product or service.
 This is accomplished through design of experiments (DOE)
for determining parameters.
 It provides a method for quantitatively identifying just the
right ingredients that go together to make a high-quality
product or service.
Taguchi Design of Experiments
Slide 2 of 3

 Robust Design
 The Taguchi concept of robust design states that products
and services should be designed so that they are inherently
defect free and insensitive to random variation.
 Taguchi’s approach for creating robust design is through a
three-step method consisting of concept design, parameter
design, and tolerance design.
 Concept design
 The process of examining competing technologies to
produce a product.
 Includes process technology and process design choices.
Taguchi Design of Experiments
Slide 3 of 3

 Parameter design
 The selection of control factors and the determination of
optimal levels for each of the factors.
 Control factors are those variables in a process that
management can manipulate.
 Optimal levels are the targets or measurements for
performance.
 Tolerance design
 Deals with developing specification limits.
 Occurs after parameter design has been used to reduce
variation and the resulting improvement has been insufficient.
Background of the Taguchi Method
Slide 1 of 6

 History and Impact

 The Taguchi method was first introduced by Dr. Genichi
Taguchi to AT&T Bell Laboratories in 1980.
 Thanks to its wide acceptance and utilization, the
Taguchi method for improving quality is now commonly
views as comparable in importance to statistical process
control(SPC), the Deming approach, and the Japanese
concept of total quality control.
Background of the Taguchi Method
Slide 2 of 6

 Taguchi Definition of Quality

 In Taguchi terms, ideal quality refers to a reference point
or target value for determining the quality level of a
product or service.
 Ideal quality is delivered if a product or tangible service
performs its intended function throughout its projected
life under reasonable operating conditions without
harmful side effects.
Background of the Taguchi Method
Slide 3 of 6

 Quality Loss Function

 If a measurement is taken of the critical product
characteristic and it is within the specification limits,
the traditional conclusion was that it wasn’t a problem.
 However, if it is closer to being out of specification than
to being at the target measurement, it might cause a
problem. Taguchi calls this potential for problem a
potential loss to society.
 Loss to society is the cost of a deviation from a target
value.
Background of the Taguchi Method
Slide 4 of 6

Classical QC-Step Function Figure 14.16

Q(c)
Scrap Cost

LSL USL

Target

A
Background of the Taguchi Method
Slide 5 of 6

 Quality Loss Function (continued)

 To quantify loss to society, Taguchi used the concept of
a quadratic loss function.
 Any variation from the target of six (where T = 6)
results in some loss to the company.
 The quality loss function focuses on the economic and
societal penalties incurred as a result of purchasing a
nonconforming product.
 Losses may include maintenance costs, failure costs, ill
effects to the environment such as pollution, or
excessive costs of operating the product.
Background of the Taguchi Method
Slide 6 of 6

Taguchi Quadratic Loss Function Figure 14.17

The Taguchi Process
Slide 1 of 6

Taguchi Process Figure 14.18

The Taguchi Process
Slide 2 of 6

 Step 1: Problem Identification

 The production problem must be identified. The
problem may have to do with the product process or the
service itself.
 Step 2: Brainstorming Session
 To identify variables that have a critical affect on service
or product quality takes place.
 The critical variables identified in the brainstorming
sessions are referred to by Taguchi as factors.
 These may be identified as either control factors
(variables that are under the control of management) or
noise factors (uncontrollable variation).
The Taguchi Process
Slide 3 of 6

 Step 3: Experimental Design

 Using the factors, factor levels, and objectives from the
brainstorming session, the experiment is designed.
 The Taguchi method uses off-line experimentation as a
means of improving quality.
 This contrasts with traditional on-line (in process)
quality measurement.
The Taguchi Process
Slide 4 of 6

 Step 4: Experimentation
 There are different Taguchi analysis approaches that use
quantitatively rigorous techniques such as analysis of
variance (ANOVA), signal-to-noise ratios (S/N), and
response charts. These approaches, although not always
theoretically sound, are useful in engineering related
projects involving engineered specifications, torques,
and tolerances.
The Taguchi Process
Slide 5 of 6

 Step 5: Analysis
 Experimentation is used to identify the factors that result
in closest-to-target performance.
 If interactions between factors are evident, two
alternatives are possible.
 Either ignore the interactions (there is inherent risk to this
approach)
 Provided the cost is not prohibitive, run a full factorial
experiment to detect interactions.
 The full factorial experiment tests all possible
interactions among variables.
The Taguchi Process
Slide 6 of 6

 Step 6: Confirming Experiment

 Once the optimal levels for each of the factors have
been determined, a confirming experiment with factors
set at the optimal levels should be conducted to validate
the earlier results.
 If earlier results are not validated, the experiment may
have somehow been significantly flawed.
 If results vary from those expected, interactions also
may be present, and the experiment should, therefore, be
repeated.
Design for Six Sigma
Slide 1 of 2

 Design for Six Sigma (DFSS)

 Used in designing new products and services with high
performance as measured by customer-based critical to
quality metrics.
 Requires DMADV methodology instead of DMAIC.
 Design
 Measure
 Analyze
 Design
 Verify
 DMADV pertains to developing new processes and
products, whereas DMAIC pertains to improving existing
processes and products.
Design for Six Sigma
Slide 2 of 2

 Design for Six Sigma (DFSS)

 IDOV is another method.
 Identify
 Design
 Optimize
 Verify
 IDOV is focused on final engineering design
optimization.
 These methods are customer focused, encompassing the
entire business-to-market process and pertain to both
services and products.
Lensing Six Sigma from a Contingency Perspective
Slide 1 of 2

 Six Sigma
 Very popular approach for improving the robustness of
designs and processes.
 Very technical and requires special expertise.
 Can be useful for companies that need to improve their
cost and efficiency through quality efforts.
 Requirements for Six Sigma Success
 Culture
 Leadership
 Commitment
 Availability of data for projects
Lensing Six Sigma from a Contingency Perspective
Slide 2 of 2

 Reasons for Six Sigma Failure

 Lack of leadership by champions
 Misunderstood roles and responsibilities
 Lack of appropriate culture for improvement
 Resistance to change and the Six Sigma structure
 Faulty strategies for deployment
 Lack of data