Design for Reliability Readiness Plan in DFSS

ABOUTthe
AUTHORS
“Readiness Plan,” p. 9
Matthew Hu (matthew.hu@hp.com) is a senior quality program manager for
Hewlett-Packard Co. in Houston. He earned his doctorate in quality and reliability
engineering from Wayne State University in Detroit. Hu is a senior member of
ASQ and holds ASQ certifications as a reliability engineer and a quality engineer.
“No Silver Bullet,” p. 17
Francisco A. Hernandez Jr. (hernandezjr_francisco@bah.com) is an associate at
Booz Allen Hamilton in Washington, D.C. He earned an MBA from the University of San
Francisco. Hernandez is an ASQ member and an ASQ-certified Six Sigma Black Belt.
“As Easy as 1, 3, 9?” p. 23
Dan Zwillinger (zwilling@az-tec.com) is a consultant in Boston. He
earned a doctorate in applied mathematics from the California Institute of
Technology in Pasadena. Zwillinger is an ASQ-certified Six Sigma Black Belt.
s i x s i g m a f o r u m m a g a z i n e I a u g u s t 2 0 1 3 I 3
call for articles
Six Sigma Forum Magazine is seeking articles for publication.
For information on the review process, types
of articles considered, and the submission
requirements, go to www.asq.org/pub/sixsigma.

S
ystem reliability is a key requirement for a system to function suc-
cessfully under the full range of conditions experienced in the oil
industry. From a probabilistic viewpoint, reliability is defined as the
probability a system will meet its intended function under stated conditions
for a specified period of time; therefore, to predict reliability, you must
know three things:
1. Function.
2. Stated conditions.
3. The specified useful life or time period.
A typical textbook that addresses reliability will present a set of proba-
bilistic concepts, such as a survival function, failure rates and mean times
between failures. These concepts are related to a model of the causes of
failure, such as component reliabilities or material and environmental
variability. To quantify, specified operating conditions are defined as an
agreed-upon range of allowable conditions or an estimated probability
density function for uncertain or variable parameters. This approach is well
suited to calculating predicted failure rates when all of the data are available.
To improve reliability prediction capability when useful data are not avail-
able or not sufficient, an alternative approach can be:
• Identify all potential function failure modes, make a risk assessment
and implement countermeasures.
• Make the product insensitive to user environments.
• Identify shortfalls in verification test plans and enhance verification
tests to ensure detection of all failure modes.
• Execute efficient verification tests that demonstrate a product is mis-
take free and robust under real-world use conditions.
System reliability requires fulfilling two critical conditions: mistake avoid-
ance and robustness.1

Mistake, in this case, is defined as the error due to design decision and
manufacturing operations. Examples of mistakes in product development
include missing components, installing a component backwards or interpret-
ing a software command as being expressed in inches when it’s actually in
centimeters. Product reliability can be improved by reducing the incidence
of such mistakes through a combination of knowledge-based engineering
and problem-solving processes, such as Six Sigma’s define, measure, analyze,
improve and control (DMAIC).
Robustness is the ability of a system to function (that is, insensitive to the
user’s environment to avoid failure) under the full range of conditions that
may be experienced in the field.
System design faces two different challenges:
1. Developing a system that functions under tightly controlled conditions,
such as in a laboratory.
2. Making that system function reliably throughout its life cycle as it
experiences a broad set of real-world environmental and operating
conditions.
dfss
Readiness Plan
Transfer function-
based design to
improve product
reliability and
robustness
in design for
six sigma
By Matthew Hu,
Hewlett-Packard Co.

Readiness Plan
10 I a u g u s t 2 0 1 3 I W W W . AS Q . ORG
An example of this real-world challenge is effective
system reliability engineering. The most cost effective
and least time consuming way to make a reliable prod-
uct—one that’s insensitive to the user environment, or
robust—is to start in the development or design phase
by discovering and preventing failure modes soon after
they are created, and implementing countermeasures
before production.
This article covers the second challenge—robust-
ness—by proactively factoring design for reliability
(DFR) efforts through transfer function-based robust-
ness improvement in the design for Six Sigma (DFSS)
approach. DFSS is a method that calls on many of the
fundamental design tools such as robust design. By
using DFSS along with a well-defined reliability plan,
you can know when to use which tool and how to
integrate each together to produce a reliable product
in the shortest amount of time.
A transfer function is a useful tool, if it’s validated
properly, that you can leverage to understand physics,
explore design space and optimize a design in terms of
reliability and robustness. Knowing the transfer func-
tion Y = f(X) between input and output, you’re able
to simulate the design performance with minimum
hardware requirements or without building prototypes
or building minimum prototypes. The variables in the
transfer function can be characterized from an engi-
neering viewpoint. Transfer functions then can enable
engineers to introduce variation into the models to
understand how the distribution of variation can alter
the desired performance by the following:
• Find the combination of control factors settings
that allow the system to achieve its ideal function.
• Remain insensitive to those variables that can-
not be controlled or that are not intended to be
controlled.
This approach allows engineers to predict what will
happen in actual applications. The essence of the
robust design approach is to design built-in quality.
Instead of trying to eliminate or reduce the causes
for product performance variability, it is preferable to
adjust the product design so product performance is
insensitive to the effects of uncontrolled (noise) varia-
tions through transfer function deployment.
Transfer function overview
A transfer function is a relationship between input
(lower-level requirements) and output (higher-level
requirements). Transfer functions are set up as equa-
tions and are expressed in Y = f(X) terms. Transfer
functions are either developed analytically or experi-
mentally that directly measure the customer needs.
Y is the output response measurement such as prod-
uct strength or customer satisfaction. The transfer
function explains the transformation of the inputs
into the output. X is any input process step that is
involved in producing the output, and Y is the intend-
ed design functions cascaded from critical to satisfac-
tion (CTS) and others. The transfer function may be
mathematically derived (for example, spring force
and displacement [Y = kx]), and empirically (induc-
tive) obtained from a design of experiment (DoE) or
regression based on the historical data (for example,
Y = a0+a1x1+a2x22+… polynomial approximation).
In general, a transfer function is established through
an analytical or empirical approach. For a proper
transfer function development, a rational structure
of a design is needed to assess where to start the
transfer function development. The transfer function
?
Select
measurable Y
Develop Y = f(X)
Assess Y = f(X)
Prediction
Are there any
previously known?
Can model
and pattern be
confirmed?
Yes
No
Induction
(from the specific to the general)
Deduction
(from the general to the specific)
Generated data
Design of experiment
Observed data
Regression
Logical foundations
• Physics equation
for example:
f = ma, y = f(x)
• Axioms
Engineering logic
CAE models
Finite elements
Function decomposition
Y=a0
+a1
x1
+a2
x2
2
+ ...
CAE = computer-aided engineering
No
Yes
Adapted from Matthew Hu and Kai Yang, “Transfer Function
Development in Design for Six Sigma Framework,” Society for
Automotive Engineering Journal, April 11, 2005.
Figure 1. Transfer function development
process flow

Readiness Plan
development process is similar to the inductive and
deductive feedback loop. The process of developing or
updating a transfer function is highly iterative, moving
frequently between the inductive and deductive paths.
Occasionally, the transfer function is known explicitly
and can be determined through the understanding of
the physics of the system. At other times, the transfer
function is unknown and must be estimated empiri-
cally through directed experiments or by the analysis
of already available data. Figure 1 shows how a transfer
function can be established.2
Deductive reasoning is the process by which an engi-
neer makes conclusions based on previously known
facts such as:
• Logical foundations—for example, physics
equations, the study of structure, change and
space patterns, and axioms.
• Engineering logic—for example, finite element
and mathematical modeling-proposed engi-
neering design.
This method of reasoning is a step-by-step process
of drawing conclusions based on previously known
truths from engineering validation. Although deduc-
tive reasoning seems rather simple, it can be mislead-
ing in more than one way. When deductive reasoning
leads to faulty conclusions, the reason is often that the
premises were incorrect; thus, the model validation is
important.
Transfer functions can be schematically represented
by the P-diagram used in robust engineering design,
as shown in Figure 2. A product can be divided into
functionally oriented operating systems. Function is
a key word and basic need for describing your prod-
uct or behavior. Regardless of what method is used
to facilitate a design, they all have to start with the
understanding of functions. Questions include: “What
is the definition of function?” and “How is the func-
tion defined in these disciplines of a specific design?”
Understanding the specific meanings of function (or
the definition of function) within each of these dis-
ciplines could help take the advantages of tools to
improve design efficiency and effectiveness.
Transfer functions can enable engineers and scien-
tists to introduce variation into the models to under-
stand how the distribution of variation can alter the
desired performance. A flowchart showing develop-
ment of a transfer function using the computer-aided
engineering (CAE) model is shown in Figure 3 (p. 12).
Inductive reasoning is the process of arriving at a
conclusion based on a set of observations (from the
specific to the general—for example, through DoE or
regression analysis). Inductive reasoning is valuable
because it allows engineers or scientists to form ideas
about groups of things in real life. In engineering,
inductive reasoning helps organize what is observed
into engineering hypotheses that can be proved using
more reliable methods. The process of inductive rea-
soning almost always is the way ideas are formed about
things. After those ideas form, it is possible to system-
atically determine (using formal validation) whether
the initial ideas were right, wrong or somewhere in
between.
Robust design overview
Robust design, also known as Taguchi parameter
design, can be used to achieve robust reliability; that
is, to make a product’s reliability insensitive to uncon-
trollable user environments. Robust design is the heart
of DFSS.
An important development in reliability engineering
is robust design pioneered by Genichi Taguchi.3
For
any design concept, there is a potentially large space
of control factor settings that will nominally place the
function at the desired target value. Taguchi’s method
Figure 2. P-diagram
Output response
Y=f(XIS
, XCF
, XNF
)
= ß M+[f(M, XCF
, XNF
) – ßM]
= ß M ideal functional
relationship
Control factors (XCF
)
Noise factors (XNF
)
Input signal
M
Error states
Y
M
ß
Ideal function
Y=ß M
P-diagram
Adapted from Matthew Hu and Kai Yang, “Transfer Function
Development in Design for Six Sigma Framework,” Society for
Automotive Engineering Journal, April 11, 2005.

Readiness Plan
employs orthogonal arrays to explore the design space.
At the same time, outer arrays or compounded noises
are used to explore the range of possible operating
conditions. Further case studies and research show
that compound noise factor theory turns out to be
the sufficient conditions for robustness and reliability
improvement. In a reliability engineering test, com-
pound noise strategy can be considered an effective
way of improving reliability confidence tests.
Robust design requires the evaluation of product
control factors in the noisy environments from which
classical multi-factor designed experiments seek isola-
tion. Taguchi recommended that noise factors be con-
sidered in any experiment to improve reliability where
it is practical. Robust reliability design is closely related
to accelerated life testing and worst case analysis in this
requirement for exposure of design to combinations of
extreme noise conditions under experimenter control.
Taguchi and other authors have written extensively
on designing quality into products and processes.4, 5
Their concepts have been widely adapted to design for
reliability. The first concept of Taguchi that must be
discussed is what he refers to as noise factors, which are
viewed as the causes of performance variability, includ-
ing why products fail. Figure 4 shows the reliability
bathtub curve and Taguchi’s type of noise.
By consciously considering the noise factors (envi-
ronmental variation during the product’s use, manu-
facturing variation and component deterioration) and
cost of failure in the field, the robust design method
helps ensure customer satisfaction. Robust design
focuses on improving the fundamental function of
the product or process; thus,
facilitating flexible designs and
concurrent engineering. When
variability occurs, Taguchi said
this is because the physics active
in the design and environment
promote change. Taguchi cat-
egorized noise into five catego-
ries:
1. Piece-to-piece variation,
such as rubber thickness.
2. Change over time, such as
failure from material wear,
or changes in force or
dimension with time.
3. Customer use, such as
open-hole wellbore size.
4. The environmental con-
dition, such as tempera-
ture variation.
5. System interactions, such as
elements outside dimension
variations and open-hole
size.
The result of noise may be
degradation in quality (soft
failure) or a malfunction failure
(hard failure). A product is said
to be robust when it’s insensi-
tive to the effects of sources
of variability, even though the
sources themselves have not
been eliminated.
Figure 4 illustrates how
Taguchi’s noise factors neatly
Figure 3. Transfer function development using
CAE model flowchart
Step 1: Develop and validate a CAE model
for a given design.
Step 2: Develop a P-diagram with identified
measurable ideal response (CTQ).
Step 3: Generate a matrix for experiments
over concerned design boundary.
Step 4: Use CAE model to calculate
response based on the experiment matrix.
Step 5: Develop response surface capturing
relationship between input and output using
surface response modeling, for example,
Kriging model.
A designed
computer experiment
critical faction.
Calculate response
using CAE model.
Develop a transfer function
using surface response model.
X2 Y2
X1 Y1
CAE = computer-aided
engineering
CTQ = critical to quality
Initial design
Approximation
Feasible
Infeasible
(failed)
Constraint
boundary
Oilfilmthickness
Taper
4
3
2
2
1
0
Bearing
phase
Optimal 2
Optimal 1
Adapted from Matthew Hu and Kai Yang, “Transfer Function Development in Design for Six
Sigma Framework,” Society for Automotive Engineering Journal, April 11, 2005.

fit within the accepted model of product failures in
reliability and their relation to the bathtub curve.
Robustness and reliability improvement
Categorically, there are five strategies for improving
robustness and thus reliability:
1. Change the design concept or technology.
2. Make the design insensitive to noise factors.
3. Reduce or remove the noise factors.
4. Use a compensation device (for example,
dynamically tuned absorbers).
5. Send the failure mode to another part of the
system (trade-off) where it will do less harm.
As noted earlier, the second strategy for making the
design insensitive to noise factors is the focus of this
article.
M.S. Phadke stated that there are three fundamental
ways to improve the reliability of a product during the
design stage:6
1. Reduce the sensitivity of the product’s function
to the variation in the product parameters.
2. Reduce the rate of change of the product
parameters.
3. Include redundancy.
The most cost-effective approach for reliability
improvement is to find appropriate continuous quality
characteristics and reduce their sensitivity to all noise
factors. Phadke provides simple examples of a robust
design approach. In actual application, however, more
than one strategy may be necessary.
DFR overview
DFR is a process. Specifically, DFR describes the
entire set of tools that support product and process
design (typically from early in the concept stage all
the way through to product obsolescence) to ensure
that customer expectations for reliability are fully met
throughout the life of the product with low overall
life cycle costs. In other words, DFR is a systematic,
streamlined, concurrent engineering program in
which reliability engineering is woven into the total
development cycle.
The purpose of the DFR process is to provide
requirements for DFR activities, which are intended
to be an integral part of every product development
effort to continuously improve product reliability and
robustness. The reliability process integrates with a
generic technology and product development process,
and can be tailored as specified in the technology and
product development process. The product develop-
ment process defines the scope and applicability. The
reliability plan documents the tailoring of the DFR
activities.
The reliability plan is created by the design team.
It is the responsibility of the design team to imple-
ment the DFR by completing the activities outlined in
this plan. The team must leverage a set of reliability
engineering tools along with a proper understanding
of when and how to use these tools throughout the
design cycle. This process encompasses a variety of
tools and practices, and describes the overall order of
deployment that an organization must follow to design
reliability into its products. The reliability is part of the
DFSS scorecard. DFR tasks can be well aligned with
and embedded in a DFSS roadmap.
To make reliability a key product requirement and
understand where reliability efforts stand in terms of
the DFR process for designing and manufacturing for
reliability, a DFR assessment scorecard can be help-
ful. The DFR assessment drives reliability goal setting,
understanding the quality history, tool selection activi-
ties, testing strategies and reliability demonstration
through DFR gates review.
The DFR process can follow the DFSS roadmap—for
example, the identify, design, optimize and validate
(IDOV) framework. With reliability in mind, prod-
uct program teams can identify the boundary and
scope of system requirements and design the product.
Meaningful test progression strategies can be devel-
Readiness Plan
Figure 4. Reliability bathtub curve and
types of noise mapping
Failure rate
Stress
Strength
Time
Affected by
outer noises
Affected by
inner noises
Failures occur
in overlap
DFR
IFR
CFR
a. Infant
mortality
noise #1
c. Wear
out
noise #2
b. Useful life
noise #3/4/5
Affected by customer use variation
Affected by manufacturing variation
CFR = constant failure rate
DFR = decreased failure rate
IFR = increased failure rate

Readiness Plan
oped and emphasized through optimizing the design
over the time domain and functional validation of the
product.
DFR activities are part of various elements in tech-
nology and product development activities during the
complete product life cycle. Goals of the DFR process
are:
• Integrate voice of customer (VOC) into product
requirements to improve reliability and robust-
ness of the product.
• Provide requirements for activities involved in
the DFR/DFSS process. Optimize the design over
the time domain and functional validation of the
product using a test progression strategy.
• Identify methods for defining product reliability
requirements and activities involved at each stage
of product development.
• Provide the practitioner a means of prioritizing
the reliability projects and studies that must be
undertaken.
• Continuously improve product reliability and
robustness over time.
DFSS overview
DFSS describes the application of Six Sigma tools to
product development and process design. The goal is
to “design in” Six Sigma performance capability. DFSS
is an approach to designing (or redesigning) a product
or service. It is equally useful in developing business
processes or technical products. DFSS is a defined
method—a culture and a way of viewing value creation.
The focus of DFSS begins with critical VOC analysis
and rational business planning. After gaining an under-
standing of the market and customer needs, design
personnel work to understand and characterize critical
design parameters and functionality. To achieve a cul-
tural shift—focused on continuous improvement—you
must go beyond DMAIC by leveraging a full suite of
performance improvement tools. The time to develop
new products is a critical success factor in almost any
business today. DFSS helps reduce development time
by deploying lessons learned throughout the develop-
ment and manufacturing setup process.
DFSS provides many tangible benefits to organiza-
tions. For instance, the DFSS approach results in long-
term cost reductions for a product. There are many
ways these savings are realized. Instead of debugging
products and processes that already exist, DFSS is a
re-examination of the function and design parameters.
DFSS starts from scratch with the goal of designing
virtually error-free products or processes. This strategy
effectively replaces the trial and error or built-test-fix
processes, and results in product designs that consis-
tently meet customer requirements. There are several
different DFSS roadmap models:
• Invention, innovation, develop, optimize and
verify (I2DOV).
• Define, concept, design, optimize and verify
(DCDOV).
• Identify, define, develop, optimize and verify
(IDDOV).
• Define, measure, analyze, design and verify
(DMADV).
• Identify, characterize, optimize and verify
(ICOV).
Each has a different focus on generic technology
development or product commercialization. The road-
map names are not important,7
but the contents and
tasks at each phase as defined to enhance product
development process are.
A typical DFSS approach includes the four ICOV
phases:
1. Identify—Identify market needs. Define customer
requirements and project goals. Identify critical to
satisfaction (CTS) and related functional targets.
Reliability is often a key CTS on the reliability aspects
of a product.
The purpose of this stage for the reliability effort
is to clearly and quantitatively define the reliability
requirements and goals for a product, as well as the
end-user product environmental and use conditions.
These can be at the system, assembly, component or
even the failure-mode level. Requirements can be
determined in many ways or through a combination
of those different ways. Requirements can be based on
contracts, benchmarks, competitive analysis, customer
expectations, cost, safety and best practices. Some of
the tools worth mentioning that help quantify the VOC
include Kano models, affinity diagrams and pair-wise
comparisons. Of particular interest to DFR are the
requirements that are critical to reliability (CTR).
The system reliability requirement goal can be allo-
cated to the assembly, component or even the failure-
mode level. After the requirements have been defined,
they must be translated into design requirements and
into manufacturing requirements.
2. Characterize—Understand the system and select
design concepts. Map CTS characteristics to lower-
level y factors. Relate y factors to critical to quality
(CTQ) or CTR x design factors. Determining use and
environmental conditions is an important early step of
a DFR program. Know what it is to be designed for and
what types of stresses the product should withstand.

The conditions can be determined based on customer
surveys, environmental measurement and sampling.
The tendency for the potential failure-mode occur-
rence is aggravated by noise factors, which are those
that engineers have little or no control and negatively
influence designed system performance. Fundamental
to designing for reliability and robustness using trans-
fer function is the inclusion of noise factors during
analysis that challenge the design and uncover poten-
tial failure modes.
After uncovered, these failure modes can be avoided
by developing appropriate counter measures—either
in the design or manufacturing process. Including
noise factors in up-front design analysis has encour-
aged engineers developing transfer function to con-
sider appropriate noise factors and realistic levels, as
well as strategies to include them in simulations.
It is important to estimate the product’s reliabil-
ity, even with a rough first-cut estimate, early in
the design phase. This can be done with estimates
based on engineering judgment and expert opinion,
physics of failure analysis, transfer functions-based
simulation models, prior warranty and test data from
similar products and components (using life data
analysis techniques), or standards-based reliability
prediction.
3. Optimize—Design for robust and reliable perfor-
mance. That minimizes product or process sensitivity
to uncontrollable user environment to have better
manufacturability and higher reliability.
In this stage, robust parameter design helps fur-
ther factor reliability tasks into the design process by
optimizing design function in the presence of noise
factors to:
• Identify important variables.
• Estimate their effect on a certain product charac-
teristic.
• Optimize the settings of these variables to
improve design robustness.
Noise screen experiments may be necessary to iden-
tify high-impact noise factors to single out significant
factor results in more realistic reliability tests and more
efficient accelerated tests (because resources are not
wasted on including insignificant stresses in the test)
prior to the robust optimization efforts.
Within the DFR concept, you are mostly interested
in the effect of stresses on your test units. Robust
design plays an important role in DFR because it assists
in identifying the factors that are significant to the
product’s life, especially when the physics of failure
are not well understood. The robustness of the given
concept design can be used to assess the limitation of
the given concept design from a reliability improve-
ment perspective.
4. Verify—Assess the integrated system and subsys-
tem effects on performance. Use reliability and manu-
facturing verification to assess design performance and
the ability to meet customer requirements.
If the design has been “demonstrated,” the product
can be released for production. When reaching the
manufacturing stage, the DFR efforts should focus
primarily on reducing or eliminating problems intro-
duced by the manufacturing process. Manufacturing
introduces variations in material, processes, manu-
facturing sites, human operators and contamination.
Because manufacturing piece-to-piece variation has
been considered as part of noise factors and was
optimized in the optimize phase, the product’s per-
formance should be insensitive to manufacturing
variation if the noise factors were identified and
incorporated in the optimize phase for the robust-
ness study.
However, reliability may be re-evaluated in light
of additional process variables. Design modifica-
tions might be necessary to improve robustness. For
example, a design should require the minimal pos-
sible amount of nonvalue-added manual work and
assembly. Whenever possible, it should use common
parts and materials to facilitate manufacturing and
assembling. It should also avoid tight design toler-
ances beyond the natural capability of the manufac-
turing processes.
Managing a DFSS project is not a trivial matter, and
all of the key enablers must be in place to realize maxi-
mum benefit. DFSS is the way for an organization to
realize Six Sigma’s full potential. DFSS has substantial
Readiness Plan
DFSS is a powerful method that can be incorporated into an
organization’s product development process to provide customers with
sustained value while generating growth, revenue and healthy profits.

Readiness Plan
effects on long-term profitability through improved
products and efficiencies. It results in increased
customer satisfaction, improved market share and
increased profit potential.
As you already have seen, reliability is a function of
time and, therefore, depends on age. This implies that
the useful life of a particular item may be defined. It
turns out this concept is useful in Six Sigma because—
by definition—DFSS is interested in designing a prod-
uct to a specified life. The assessment of reliability
usually involves testing and analysis of stress, strength
and environmental factors, and should always include
improper use by the end user. A reliable design should
anticipate all that can go wrong. DFR can be viewed as
a means to maintain and sustain Six Sigma capabilities
over time and is one tool set in the DFSS method.
Using a structured process to gain insight to the
customer’s needs and translate them to tangibles, CTQ
product specifications significantly reduces cycle time
and ensures a higher probability of success. Using
metrics, data and a rigorous approach, you can gain
fundamental knowledge about the critical parameters
of the product. This shared knowledge is instrumental
in producing and selling high quality, consistent, cost
competitive and profitable products.
DFSS is a powerful method that can be incorporated
into an organization’s existing product development
process to provide its customers with sustained value
while generating growth, revenue and healthy profits
for itself.
Reliability and DFSS
Reliability is one of the most important characteristics
of an engineering system. Reliability can be measured
as robustness over time. A reliable product is insensi-
tive to noise (uncontrollable user conditions) over
time. Insufficient data or lack of useful reliability field
data presents challenges of conducting meaningful
reliability analysis, prediction and, therefore, proper
decision making.
Analytical reliability and robustness using transfer
functions enable engineers to introduce variation (for
example, manufacturing piece-to-piece variation and
aging) into the analytical models to understand how
the distribution of variation can alter the desired per-
formance. Reliability and robustness can be analyzed
and optimized through transfer functions. Potential
failure modes may be uncovered and discovered
through a properly developed transfer function. Noise
factors can be identified and included in transfer
functions to uncover potential failure modes for reli-
ability improvements in the up-front design phase.
The design of swell packers for use in the energy
industry is a perfect example of being challenged for
proper reliability prediction when useful data are not
available.
Product development has a huge impact on rev-
enue stream and reliability. Enhancing product devel-
opment process with DFSS disciplines will improve
the product delivery process to develop a customized
DFR process with required tools to support specific
reliability tasks. It’s more cost effective and less time
consuming to make design insensitive to uncontrol-
lable user environments using transfer function.
DFR tasks can be best accomplished through a DFSS
roadmap.
EDITOR’S NOTE
Six Sigma Forum Magazine will publish the second installment of Hu’s article
in the November 2013 edition. That article will present a case study of swell
packer reliability improvement using transfer function.
REFERENCES
1. Don Clausing and Daniel D. Frey, Improving System Reliability by Failure-
Mode Avoidance Including Four Concept Design Strategies, Wiley InterScience,
2006.
2. Matthew Hu and Kai Yang, “Transfer Function Development in Design
for Six Sigma Framework,” Society for Automotive Engineering Journal, April
11, 2005.
3. Genichi Taguchi and Yoshiko Yokoyama, Taguchi Methods: Design of Experi-
ments, American Supplier Institute, 1993.
4. Madhav S. Phadke, Quality Engineering Using Robust Design, Prentice-Hall,
1989.
5. Genichi Taguchi, Subir Chowdhury and Yuin Wu, Taguchi Quality Engineer-
ing Handbook, Wiley, 2004.
6. Phadke, Quality Engineering Using Robust Design, see reference 4.
7. Hu and Yang, “Transfer Function Development in Design for Six Sigma
Framework,” see reference 2.
BIBLIOGRAPHY
Box, George E.P., “Scientific Methods: The Generation of Knowledge and
Quality,” Quality Progress, January 1997, pp. 47-50.
Cabadas, Joseph, “Robust Engineering Eliminates Unnecessary Expenses at
Ford,” U.S. Auto Scene, April 12, 1999.
Davis, Tim, “Measuring Robustness as a Parameter in a Transfer Function,”
Society of Automotive Engineers (SAE) International technical paper,
presented at SAE World Congress and Exhibition, March 8, 2004.
Hu, Matthew, John M. Pieprzak and John Glowa, “Essentials of Design
Robustness in Design for Six Sigma (DFSS) Methodology,” SAE Interna-
tional technical paper, presented at SAE World Congress and Exhibition,
March 8, 2004.

Design for Reliability Readiness Plan in DFSS

More Related Content

What's hot (20)

Similar to Design for Reliability Readiness Plan in DFSS (20)

Recently uploaded (20)

Design for Reliability Readiness Plan in DFSS