Approved for Public Release, Distribution Unlimited, GDLS approved, log 2006-24, dated 04/26/06 2 nd Annual Design for Six Sigma Conference Introduction to Design for Six Sigma James M. Wasiloff, MBB September 14, 2006
Jan 18, 2015
Approved for Public Release, Distribution Unlimited, GDLS approved, log 2006-24, dated 04/26/06
2nd Annual Design for Six Sigma Conference
Introduction to Design for Six SigmaJames M. Wasiloff, MBB
September 14, 2006
2Approved for Public Release, Distribution Unlimited, GDLS approved, log 2006-24, dated 04/26/06
Driving SuccessDriving SuccessUsing DFSS in the Development of Battle Ships @Raytheon
Jon Mckenzie, Director of Six Sigma at Raytheon says, “in early phases of product development we use DFSS in modeling and simulation of how these
products will work. After the contract is awarded and we have firmed up exactly what we are going to build and what the system is going to look like, then we use
DFSS to derive requirements from the customer, and all the way to the critical design elements that a design engineer will need to put the parts together and
make it work”
McKenzie explained that DFSS is embedded in Raytheon’s Integrated Product Development System (IPDS), which he said “governs everything we do
in the company. If you follow IPDS, you are going to get DFSS along the way.”
3Approved for Public Release, Distribution Unlimited, GDLS approved, log 2006-24, dated 04/26/06
Driving SuccessDriving Success"Boeing Picks McNerney as Chief" - Wall Street Journal, July 1, 2005
Jim McNerney has left his post at 3M Company where he had been the top executive for over four years to take the job as chairman, president and chief executive of the world's largest aerospace company. McNerney is globally recognized as a strong advocate for the deployment of Design for Six Sigma. The following are a sampling of quotes from McNerney while serving at 3M:
"Six Sigma is totally changing 3M. Many of the things that had driven the success of our company for the past eighty years no longer apply"
"Six Sigma is not a program. It's our game plan. It will challenge all of us. individually and collectively, to be the very best we can be"
"Major goal is to have for the first time, common approach to problem solving, new product development, and measurement across entire company"
"McNerney preaches Six Sigma to Clients... It changes everybody's lives in the first year. We're betting our performance on Six Sigma. This is something that, if Six Sigma doesn't succeed, the
company doesn't succeed."
"At 3M, Six Sigma is driven by our executive management teams, who are fully engaged in critical business processes and actively deploying Six Sigma methodologies throughout the
organization"
4Approved for Public Release, Distribution Unlimited, GDLS approved, log 2006-24, dated 04/26/06
Driving Success
Subir Chowdhury, author of the book “The Power of Design for Six Sigma”
states in his book:
“ Most companies spend only 5% of their budget on design, when design typically would determine 70% of the cost of the product- partly because 80 % of all quality problems are unwittingly designed into the product itself. In fact, in government contracts, 30 to 40 % of the budget is set aside for testing and correcting the product. Imagine! So they are admitting in advance that one-third of the budget must be devoted to correcting the problems they plan to create with the first two-thirds of the budget. My gut says, any time testing and fixing are planned for up front, it is a virtual certainty that testing and fixing will be performed. Plan for Failure and you’ll get it”
5Approved for Public Release, Distribution Unlimited, GDLS approved, log 2006-24, dated 04/26/06
Six Sigma: High Level PerspectiveDEFINE PROBLEM / ISSUE
StrategyPrevent defect(s)
Eliminate defect(s)
CHARACTERIZE
DEFINE
OPTIMIZE
VERIFY
MEASURE
ANALYZE
IMPROVE
CONTROL
DMAIC Black belt project toimprove mfg. capability
Mfg. process cannot provide sufficient improvement. Need DFSS project to reduce product sensitivity to mfg. noise.
DFSS DMAIC
New products Existing products?
6Approved for Public Release, Distribution Unlimited, GDLS approved, log 2006-24, dated 04/26/06
Introduction to Design For Six Sigma
Across private industry and government/defense sectors, Design For Six Sigma is a Product Development process that:
Effectively translates the Voice of The Customer into a design Models and quantifies the design’s performance and risk Applies statistical tools to understand, optimize, and control key
factors (or develop countermeasures to…) that deliver critical customer attributes robustly in the presence of noise
Quantifies risk and facilitates business discussions regarding product delivery quality and reliability early in the Product Development process
7Approved for Public Release, Distribution Unlimited, GDLS approved, log 2006-24, dated 04/26/06
A Historical Perspective on DFSS
Six Sigma developed at Motorola and adopted by GE and others (Steps: DMAIC)
DFSS concept originally developed at GE in late ’90s GE approach requires enhancement in the DFSS concept for
successful application to automotive industry product development: More focus on achieving Customer Satisfaction by improving
Robustness and High Time in Service performance
Aligned to defense product development practices (e.g., DVP&R, Kano Model, Robust Engineering, Reliability & Robustness checklist, etc.)
8Approved for Public Release, Distribution Unlimited, GDLS approved, log 2006-24, dated 04/26/06
Key Elements of DFSS
Should be viewed as an enhancement to the current design process:
Bundles existing product development tools Teaches tools just-in-time at appropriate development phases Provides common 6-sigma based language for PD Not a “locked-in” process that requires the use of specified tools at
every milestone Each program may select tools according to the ability to fulfill key
DFSS elements
Is a stage-gate process Is a team and project driven process Score Card driven process Integral part of IPS and CMMI
9Approved for Public Release, Distribution Unlimited, GDLS approved, log 2006-24, dated 04/26/06
Characterize
Gate Reviews and Project Reviews
Optimize
Verify
GateReview
GateReview
Define
GateReview
Gate Review
Project Reviews
Project Reviews
Project Reviews
Project Reviews
10Approved for Public Release, Distribution Unlimited, GDLS approved, log 2006-24, dated 04/26/06
DFSS ScorecardSummary: Gate review tool and project summarySummary: Gate review tool and project summary
Project Team/BBs Rate Deliverables
Separates how well the team has done from the answer they got
Project Team/BBs Rate Deliverables
Separates how well the team has done from the answer they got
Key Phase Schedule: J F M A M J J A S O N D J F M A M JConcept DesignDesign DevelopmentOptimizeVerify Capability
Phase Deliverables: Owner Date Risks/Issues
1. Critical Parameter Database G G G2. G G G3. G G G4.Subsystem & Subassembly Reliability Test
Plans & Current Data/Results Summary G G G5. Say/Do Contract book Update G G G6. Marketing Plan Update / Final G G G7. Issue management G G G8. Updated Risk Mitigation Plan G G G9. Updated Integrated Schedule G G G
10. Customer Requirements Validation G G G11. G G G12. SA Checklist Complete/Approved G G G13. Business Case Update / Final G G G14.Post-Launch Product Control Plan G G G
Total Score Gate Approval G
2006
ConfidenceScore
PerformanceScore
Gate Reviewscore
System / Sub-system Robustness VerifiedOptimized Design Performance Verified
Mfg. & Supply Chain Capability Assessment
2005
Phase-Gate DeliverablesPhase-Gate Deliverables
Gate Reviewers Rate
Deliverables
Gate Reviewers Rate
Deliverables
Decision by GatekeepersDecision by Gatekeepers
11Approved for Public Release, Distribution Unlimited, GDLS approved, log 2006-24, dated 04/26/06
Project Categories1. New programs or technologies with large design space and some
constraints (not usually a “clean sheet design”, but hopefully allows concept selection)
2. New product applications (beyond conceptual phase) for which design does not meet customer wants/functional requirements; usually, limited design space
3. Current model applications with very small design space and many constraints (high degree of optimization “tuning”)
4. New model applications with very aggressive Reliability Requirements
Project examples: New products with 10 x Reliability of legacy design
12Approved for Public Release, Distribution Unlimited, GDLS approved, log 2006-24, dated 04/26/06
Project Prioritization Scheme
Low High High High
Low Low High Low
Critical to GDLS
Cri
tica
l to
Cu
sto
mer
Low High
Low
Hig
h
13Approved for Public Release, Distribution Unlimited, GDLS approved, log 2006-24, dated 04/26/06
Proposed Project Selection Criteria
Impact on customer satisfaction Impact on reliability Design “degrees of freedom” Estimated cost avoidance Impact on maintainability Project duration Project complexity/scope Manufacturing location CAE model availability
14Approved for Public Release, Distribution Unlimited, GDLS approved, log 2006-24, dated 04/26/06
DFSS Key Focus
Developing a QFD or other rigorous identification of customer rigorous identification of customer needsneeds to greater depth than is current practice
Defining or enhancing a transfer functiontransfer function, “y=f(x)” that mathematically describes “critical to satisfaction” metrics in terms of design variables
Better leveraging analytical meansanalytical means to identify & optimize critical design, manufacturing, and assembly elements
Summarizing design risk in a scorecardscorecard that captures design & manufacturing capability and enforces process discipline
Assessing field robustness robustness and using that assessment to guide verification planning & implementation
15Approved for Public Release, Distribution Unlimited, GDLS approved, log 2006-24, dated 04/26/06
Understand Customer- and -
Understand History
Identify Critical to Satisfaction Drivers (CTS’s) and Related Functional Targets
• Flow Down to CTS’s to lower level (y’s)
• Relate CTS’s (y’s) to CTQ design parameters (x’s)
• Characterize robustness opportunities including high mileage
• Characterize capability/stability and select robustness strategy
Design for Producibility
Design for Robust Performance
Minimize process sensitivity to product & mfg. variations
Minimize product sensitivity to mfg. & usage conditions
Test & Verify
Assess Performance, Reliability & Manufacturing…Not OK
OK
Perform tradoffs to ensure all
CTS’s are met
DefineCTS’s
CharacterizeSystem
OptimizeProduct/ Process
VerifyResults
Design for Six Sigma: DCOV
DFSS Assess-
ment
Estimate for
process capability and for product function
over time
Capture data in
scorecards
Understand System- and -
Select Concepts
16Approved for Public Release, Distribution Unlimited, GDLS approved, log 2006-24, dated 04/26/06
DFSS (DCOV) Flow of Analysis & Tools
Signifies the OPTIMUM S/N Set point…Signifies the OPTIMUM S/N Set point…
Function ModelingConcept Generation
& Selection
DoERobust Design & Tolerance Design
Functions
VOC KJ QFD
Customer Needs/Statements
Customer Requirements
FMEA
Technical Requirements
Reliability/Robustness Demonstration
0.1 1.0 10.0 100.0
1
2
3
5
10
20
30
40 50 60 70 80
90 95
99
Time to Failure
Pe
rce
nt
Probability Plot for time5Weibull Distribution-95.0% Conf idence Interv als
Censoring Column in censor5
ShapeScale
MeanStDev
MedianIQR
1.2257 7.8398
7.3357 6.0165
5.8135 7.3970
0.1 1.0 10.0 100.0
1
2
3
5
10
20
30
40 50 60 70 80
90 95
99
Time to Failure
Pe
rce
nt
Probability Plot for time5Weibull Distribution-95.0% Conf idence Interv als
Censoring Column in censor5
ShapeScale
MeanStDev
MedianIQR
1.2257 7.8398
7.3357 6.0165
5.8135 7.3970
Concepts
D
C
O
V
17Approved for Public Release, Distribution Unlimited, GDLS approved, log 2006-24, dated 04/26/06
4.0 Verify
5.0 Verify
Functionality
DFSS-DCOV Process Map
DefineCharacterize Optimize
What are the customer requirements?
What approach/ system architecture would best meet customer’s requirements?
How can we optimize forRobustness?
How can weensure thatthe customerneeds will beconsistently metin variousenvironmentsand situations?
How can we design to meet the customer needs?
What measurable system requirements supportthe customer needs?
How will we demonstrate robustness/reliability?
How can weverify that this product is reliableand robust?
Verify
18Approved for Public Release, Distribution Unlimited, GDLS approved, log 2006-24, dated 04/26/06
DFSS (DCOV) Flow of Analysis & Tools
Signifies the OPTIMUM S/N Set point…Signifies the OPTIMUM S/N Set point…
Function ModelingConcept Generation
& Selection
DoE
Robust Design & Tolerance Design
Functions
VOC KJ QFD
Customer Needs/Statements
Customer Requirements
FMEA
18
Technical Requirements
Reliability/Robustness Demonstration
0.1 1.0 10.0 100.0
1
2
3
5
10
20
30
40 50 60 70 80
90 95
99
Time to Failure
Pe
rce
nt
Probability Plot for time5Weibull Distribution-95.0% Conf idence Interv als
Censoring Column in censor5
ShapeScale
MeanStDev
MedianIQR
1.2257 7.8398
7.3357 6.0165
5.8135 7.3970
0.1 1.0 10.0 100.0
1
2
3
5
10
20
30
40 50 60 70 80
90 95
99
Time to Failure
Pe
rce
nt
Probability Plot for time5Weibull Distribution-95.0% Conf idence Interv als
Censoring Column in censor5
ShapeScale
MeanStDev
MedianIQR
1.2257 7.8398
7.3357 6.0165
5.8135 7.3970
Concepts
D
19Approved for Public Release, Distribution Unlimited, GDLS approved, log 2006-24, dated 04/26/06
Define Phase The Define phase begins the process with a formal tie of design to Voice of
the Customer. This phase involves developing a team and team charter, gathering VOC, performing competitive analysis, and developing CTQs.
Crucial Steps: Identify and cascade customer and product requirements Establish the business case Identify technical requirements (CTQ variables and specification limits) Roles and responsibilities Establish project Milestones
Key Tools: VOC (Voice of Customer) QFD (Quality Function Deployment) KJ (Kawakita, Jiro) Kano Model Benchmarking
20Approved for Public Release, Distribution Unlimited, GDLS approved, log 2006-24, dated 04/26/06
Gather, Process & Validate the Voice of the Customer (VOC)
• Who our customer is • The wants or needs of
the customer (Voice of the Customer)
KJ
• An affinity diagram that groups similar things of qualitative nature
• The customer’s wants orneeds translated into SystemLevel Technical Requirements
QFDVOC
21Approved for Public Release, Distribution Unlimited, GDLS approved, log 2006-24, dated 04/26/06
KJ (Kawakita, Jiro) Analysis
Outputs:• Grouped, ranked & prioritized Customer Requirements• Customer feedback – ranked importance for each requirement, ranked
strength of competitors’ ability to fulfill the Customer Requirements
Outputs:• Grouped, ranked & prioritized Customer Requirements• Customer feedback – ranked importance for each requirement, ranked
strength of competitors’ ability to fulfill the Customer Requirements
Summary: integrate “images & visualized needs” from the customers with written Voice of the Customer inputs to obtain prioritized Customer Requirements. Summary: integrate “images & visualized needs” from the customers with written Voice of the Customer inputs to obtain prioritized Customer Requirements.
KJ Questions/Guidelines • How are customers being selected?• How will you interview the customers?• How are you facilitating selection of the NUDs? • How are the requirements being prioritized?
Conducting KJ Analysis: create affinity diagrams that gather and group VOC wants/needs and rank needs based on what is New, Unique and/or Difficult (NUDs) …
22Approved for Public Release, Distribution Unlimited, GDLS approved, log 2006-24, dated 04/26/06
KJ Results – Voice of Consumers
Easy to use
Device must have easy to use buttons
Easy to navigate internet
Device must have easy text entry method
Few or No Dropped Calls
Dependable connection, in various environments and terrains
Good Imaging
Device must have ability to expand memory
Device must have easy picture messaging
Can go 3 days without Recharging
Provides securem-Commerce
Must support banking solution
Must provide security regarding identity
Optimized Display Size
High Resolution
Max. Screen to Phone
Supports High SpeedInternet Connectivity
The device must be able to handle data-intensive serviceDevice should be a compelling replacement for wired high speed internet
Fits lightly into A Pocket or Purse
Small size
Lightweight
Prefer Customizable at Point of Sale
Can Email Quality PhotosEasily and Quickly
The device must have a push email solution.
Device must have a simple email solution
Camera as good as stand alone
Device must have great picture quality on display
Dependable connection within building, with weak signals or strong signals
Good Entertainment
Device must have a good multiplayer gaming solutionDevice must have high quality music solution
Device must have video on demand
Must have confidence that phone will have enough power for urgent situations.
23Approved for Public Release, Distribution Unlimited, GDLS approved, log 2006-24, dated 04/26/06
Cell Phone VOCs
Few or nodropped calls
Fits lightly intoa pocket or purse
Cell Phone Project
Can go 3 days without recharging
Can email quality photos easily &
quickly
Supports high speed internet connectivity
Provides secure m-Commerce
24Approved for Public Release, Distribution Unlimited, GDLS approved, log 2006-24, dated 04/26/06
Quality Function Deployment (QFD)
NUDs from KJ Analysis are translated to System level Technical Requirements through Quality Function Deployment (QFD).
VOC KJ QFD
Customer Need
Statements
Customer Requirements
(NUDs)
System Level Technical
Requirements
25Approved for Public Release, Distribution Unlimited, GDLS approved, log 2006-24, dated 04/26/06
QFD/House of Quality
Summary: Translates the Voice Of the Customer (VOC) into New, Unique and/or Difficult Technical Requirements. Identifies conflicts / tradeoffs among Technical Requirements. Links System Technical Requirements to Subsystems, Subassemblies and Components
Summary: Translates the Voice Of the Customer (VOC) into New, Unique and/or Difficult Technical Requirements. Identifies conflicts / tradeoffs among Technical Requirements. Links System Technical Requirements to Subsystems, Subassemblies and Components
Cus
tom
er Im
port
ance
HOWs (Title)
Sof tware will prov ide world class out of box experience
Ac
hiev
e b
enc
hmar
kin
g to
ol s
uite
goa
l
Win
driv
er C
ertif
icat
ion
Su
ppor
t cu
sto
mer
's
requ
ired
ve
rsio
ns o
f op
erat
ing
sys
tem
s
Mul
tiple
ope
ratin
g
syst
ems
supp
ort
Improve Perf ormance
Spe
cm
ark
impr
ovem
ent
Red
uce
Cos
t
Impr
ove
Iop
s
Im
prov
e pr
oces
sor
sp
eed
Im
prov
e m
emor
y ba
ndw
idth
Tw
o in
depe
nden
t ho
stbr
idge
d pr
oces
sor
s
Expandable I/O
Sel
ect
indu
stry
st
anda
rd e
xpa
nsio
n
Opt
imiz
e bu
s in
terf
ace
to I
/O
Quasar must meet market window and product lif ecy cle reqts
Pro
ject
sta
ffed
fo
r de
sign
com
ple
tion
by
reqd
dat
e
Customer Assessment
Direction of Improvement
WH
AT
s (T
itle
)
So
ftw
are
will
pr
ovid
e w
orld
cl
ass
out
of
box
expe
rienc
e
Optimize board support package
Software compatibility and migration should be easy
Robust sof tware env ironment
Impr
ove
pe
rfor
ma
nce
4x greater perf ormance or 4x lower cost
Perf ormance =>PC
Increase perf ormance to simplify design and reduce cost
Loosely coupled processors
Ex
pand
able
I/O
Throughput as f ast or f aster than prior generation product
Qua
sar
mus
t m
eet
m
ark
et w
indo
w a
nd
prod
uct
life
cycl
e
requ
irem
ents
Release in 2004
How Much
Weighted Importance
Relative Importance
F F G H F F G F H F F F
F F F H H G H
H H H H H G F
F F F F F F F G F F
F H F F F F F G F F
F F F F F F F F F F
G G F H F F F F F H G
F H F F F F F G F F
F
I K K I J J I I I K K I K
10.0
10.0
10.0
6.0
6.0
6.0
6.0
3.0
2.0
100%
C
ompl
ian
ce
100%
100%
S
upp
ort
for
3 O
S's
>3
OS
's
>80
%
<$4
00
>1M
>2
GH
z
>10
0 M
B/s
Ach
ieve
100%
BE
R <
1e-1
2
Pro
ject
ed M
3 be
fore
1/
12/2
005
285.
0
216.
0
211.
0
168.
0
333.
0
108.
0
333.
0
253.
0
333.
0
333.
0
219.
0
317.
0
423.
0
VV
VVVV
VVVV
VV
WW
VV
WW
VV
RR
RR
SSSS
RR
SS
VVVV
VV
SSSS
SSRR
WWSS
WW
VVWW
VV
VVVV
VV
WW
WebSurfboard NUDs top level May 13, 2004
VOC NUD’s(from KJ Analysis)
VOC NUD’s(from KJ Analysis)
Conflicts / Tradeoffsamong
Technical Requirements
Conflicts / Tradeoffsamong
Technical Requirements
Ranked Importancefor Technical Requirements
Ranked Importancefor Technical Requirements
Technical RequirementsTechnical Requirements
Targets, Tolerances for Technical Requirements
Targets, Tolerances for Technical Requirements
1
3
2
4
5
26Approved for Public Release, Distribution Unlimited, GDLS approved, log 2006-24, dated 04/26/06
4.0 Verify
5.0 Verify
Functionality
DFSS-DCOV
DefineCharacterize
Optimize
What are the customer requirements?
What approach/ system architecture would best meet customer’s requirements?
How can we optimize forRobustness?
How can weensure thatthe customerneeds will beconsistently metin variousenvironmentsand situations?
How can we design to meet the customer needs?
What measurable system requirements supportthe customer needs?
How will we demonstrate robustness/reliability?
How can weverify that this product is reliableand robust?
Verify
27Approved for Public Release, Distribution Unlimited, GDLS approved, log 2006-24, dated 04/26/06
DFSS (DCOV) Flow of Analysis & Tools
Signifies the OPTIMUM S/N Set point…Signifies the OPTIMUM S/N Set point…
Function ModelingConcept Generation
& Selection
DoERobust Design & Tolerance Design
Functions
VOC KJ QFD
Customer Needs/Statements
Customer Requirements
FMEA
Technical Requirements
Reliability/Robustness Demonstration
0.1 1.0 10.0 100.0
1
2
3
5
10
20
30
40 50 60 70 80
90 95
99
Time to Failure
Pe
rce
nt
Probability Plot for time5Weibull Distribution-95.0% Conf idence Interv als
Censoring Column in censor5
ShapeScale
MeanStDev
MedianIQR
1.2257 7.8398
7.3357 6.0165
5.8135 7.3970
0.1 1.0 10.0 100.0
1
2
3
5
10
20
30
40 50 60 70 80
90 95
99
Time to Failure
Pe
rce
nt
Probability Plot for time5Weibull Distribution-95.0% Conf idence Interv als
Censoring Column in censor5
ShapeScale
MeanStDev
MedianIQR
1.2257 7.8398
7.3357 6.0165
5.8135 7.3970
Concepts
C
28Approved for Public Release, Distribution Unlimited, GDLS approved, log 2006-24, dated 04/26/06
Characterize Phase The Characterize phase emphasizes CTQs and consists of identifying
functional requirements, developing alternative concepts, evaluating alternatives and selecting a best-fit concept, deploying CTQs and predicting sigma capability.
Crucial Steps:Formulate and select concept design
Identify potential risks using FMEA For each technical requirement, identify design parameters (CTQs) using
engineering analysis such as simulation Robustness Strategy Use DOE (design of experiments) and other analysis tools to determine CTQs
and their influence on the technical requirements (transfer functions)
Key Tools: Risk assessment and FMEA Engineering analysis and simulation DOE (Design of Experiments) Critical Parameter Management (CPm) Design tools like TRIZ, Axiomatic Design, Functional Modeling, Pugh
29Approved for Public Release, Distribution Unlimited, GDLS approved, log 2006-24, dated 04/26/06
CPM - a Disciplined MethodologyCPM is a disciplined methodology to capture the product performance into a structured repository.
Marketing
Design
SystemsEngineering
SubjectMatter
Experts
MFG
Suppliers
Standards
Partners
CriticalParameter
Management(CPM)
30Approved for Public Release, Distribution Unlimited, GDLS approved, log 2006-24, dated 04/26/06
CPM – The Strategy
Outputs:• Capability Growth Index (CGI)• Predicted capability (Cp, Cpk) for meeting Critical Customer Requirements
Outputs:• Capability Growth Index (CGI)• Predicted capability (Cp, Cpk) for meeting Critical Customer Requirements
Product Launch
Higher Layer Cp/Cpk
System Capability Cp/Cpk
Transfer Function
System Requirements
CPM Flow Down
Lower Layer Cp/Cpk
Subsystem Requirements
Modeling
Opt
imize
Verify
Design Requirements
KJ QFDVOCConcept
Summary: Creates a formal link between the product’s design & customer needsSummary: Creates a formal link between the product’s design & customer needs
31Approved for Public Release, Distribution Unlimited, GDLS approved, log 2006-24, dated 04/26/06
Function Modeling
Summary: Develop an architecture-independent (unbiased regarding possible solutions) hierarchy of functions through functional diagrams that illustrate the flow of functions within and across the technical requirements. The results are used during system architecting and design.
Summary: Develop an architecture-independent (unbiased regarding possible solutions) hierarchy of functions through functional diagrams that illustrate the flow of functions within and across the technical requirements. The results are used during system architecting and design.
Output:• Functional Flow Diagrams
Output:• Functional Flow Diagrams
Function Modeling Guidelines • Be sure to consider all of the key functions.
• Be sure to capture all the key flows.
32Approved for Public Release, Distribution Unlimited, GDLS approved, log 2006-24, dated 04/26/06
Theory of Inventive Problem Solving (TRIZ)
Summary: Generate alternative Concepts to fulfill the Technical Requirements, using innovative approaches such as the TRIZ (Theory of Inventive Problem Solving) to overcome tradeoffs among the technical requirements. The candidate concepts are evaluated using the Pugh approach.
Summary: Generate alternative Concepts to fulfill the Technical Requirements, using innovative approaches such as the TRIZ (Theory of Inventive Problem Solving) to overcome tradeoffs among the technical requirements. The candidate concepts are evaluated using the Pugh approach.
33Approved for Public Release, Distribution Unlimited, GDLS approved, log 2006-24, dated 04/26/06
Pugh Concept Selection Matrix
Summary: Compares and selects best ideas & concepts using a simple system of “better than”, “worse than”, and “same” scoring. Identifies best features from each concept and creates hybridized solutions.
Summary: Compares and selects best ideas & concepts using a simple system of “better than”, “worse than”, and “same” scoring. Identifies best features from each concept and creates hybridized solutions.
Criteria / Concept 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Ease of achieving 105-125 DbA S - + - + + - - - - S +Ease of achieving 2000-5000Hz S S N + S S + S - - - S +Resistance to corrosion, erosion & water - - O S - - S - + - - - SResistance to vibration, shock & acceleration D S - T S - S - - S - - - -Resistance to temperature A S - S - - - S S - - S SRsponse time T S - + - - - - S - - - -Complexity: number of stages U - + E S + + - - - + + - -Power consumption M - - V + - - + - - - - S +Ease of maintenance S + A + + + - - S + + S -Weight - - L + - - - S - - - - +Size - - U S - - - - - - - - -Number of parts S S A + S S - - + - - S -Life in service S - T + - S - - - - - - -Manufacturing cost - S E - + + - - S - - - -Ease of installation S S D S S + - S - - - S -Shelf life S S S S - S S S S S S
0 6 10
2 9 5
8 1 7
3 9 4
5 7 4
3 12 1
0 11 5
2 8 6
2 13 1
2 13 1
0 8 8
4 9 3
+ - S
Each concept compared
against “best in class” datum
Each concept compared
against “best in class” datum
Concepts listed along top of
matrix
Concepts listed along top of
matrix
Customer & GDLS desired criteria listed along side of
matrix
(from top of QFD HOQ)
Customer & GDLS desired criteria listed along side of
matrix
(from top of QFD HOQ)
Concepts are compared in terms
of fulfillment of criteria
Concepts are compared in terms
of fulfillment of criteria
34Approved for Public Release, Distribution Unlimited, GDLS approved, log 2006-24, dated 04/26/06
FMEA (Failure Modes & Effects Analysis)
Outputs:• Ranked group of failure modes • Impacts of failures on customer, product or process• Risk Priority Numbers (RPN) before and after corrective action • Corrective action to remove or reduce the risk or impact of a failure mode
Outputs:• Ranked group of failure modes • Impacts of failures on customer, product or process• Risk Priority Numbers (RPN) before and after corrective action • Corrective action to remove or reduce the risk or impact of a failure mode
Potential Failure Modes
Potential Failure Effects S
EV Potential
Causes OC
C
RP
N
Controls
FMEA worksheet for identifying the failure modes with the highest
RPN
FMEA worksheet for identifying the failure modes with the highest
RPN
Summary: Failure Modes and Effects Analysis is a structured method for identifying & ranking the significance of various failure modes of the program & their effects on the product or customer
Summary: Failure Modes and Effects Analysis is a structured method for identifying & ranking the significance of various failure modes of the program & their effects on the product or customer
FMEA Guidelines • The FMEA session should include adequate
representation of key people with input on risks.• When will you implement corrective actions for
the highest RPN failure modes?
35Approved for Public Release, Distribution Unlimited, GDLS approved, log 2006-24, dated 04/26/06
Design for Manufacturability & Assembly
Identifies part consolidation opportunities Exposes manufacturing, assembly, quality,
service and cost problems early in the design process
Objectively assesses design simplification opportunities
Drives optimization of product costs Allows setting of target costs
36Approved for Public Release, Distribution Unlimited, GDLS approved, log 2006-24, dated 04/26/06
Optimizing for Manufacturing & Assembly Reduces Overhead Costs
Fewer parts Less material to inventory Fewer assembly stations Less automatic assembly equipment Less dedicated fabrication tooling Less paperwork and drawings
Fewer subassemblies and operations Fewer work holders Fewer assembly tools and fixtures
37Approved for Public Release, Distribution Unlimited, GDLS approved, log 2006-24, dated 04/26/06
Design for Manufacturability & Assembly (DFM) – Key Points to Remember
Design parts for ease of fabrication Simple geometry Few process steps Make with existing equipment & tooling or can be
easily contracted out Geometry allows easy machine tool access to
machined surfaced Not too small or too large for existing processes Usual tolerances (not too tight)
38Approved for Public Release, Distribution Unlimited, GDLS approved, log 2006-24, dated 04/26/06
Eliminate Parts by Integration
Welding
Weldingor
Caulking
Eliminate Welding Operation
39Approved for Public Release, Distribution Unlimited, GDLS approved, log 2006-24, dated 04/26/06
Reduce the Number of Parts
Label Letters are molded
Assembly Cost Reduction: 100%
Mold message onto surface Eliminate parts by integration Reduce assembly time
40Approved for Public Release, Distribution Unlimited, GDLS approved, log 2006-24, dated 04/26/06
Reduce Fastener Count
Note: NCR study estimated that each screw eliminated
saves $12K in life cycle costs!
41Approved for Public Release, Distribution Unlimited, GDLS approved, log 2006-24, dated 04/26/06
4.0 Verify
5.0 Verify
Functionality
DFSS-DCOV
Define CharacterizeOptimize
What are the customer requirements?
What approach/ system architecture would best meet customer’s requirements?
How can we optimize forRobustness?
How can weensure thatthe customerneeds will beconsistently metin variousenvironmentsand situations?
How can we design to meet the customer needs?
What measurable system requirements supportthe customer needs?
How will we demonstrate robustness/reliability?
How can weverify that this product is reliableand robust?
Verify
42Approved for Public Release, Distribution Unlimited, GDLS approved, log 2006-24, dated 04/26/06
DFSS (DCOV) Flow of Analysis & Tools
Signifies the OPTIMUM S/N Set point…Signifies the OPTIMUM S/N Set point…
Function ModelingConcept Generation
& Selection
DoERobust Design & Tolerance Design
Functions
VOC KJ QFD
Customer Needs/Statements
Customer Requirements
FMEA
Technical Requirements
Reliability/Robustness Demonstration
0.1 1.0 10.0 100.0
1
2
3
5
10
20
30
40 50 60 70 80
90 95
99
Time to Failure
Pe
rce
nt
Probability Plot for time5Weibull Distribution-95.0% Conf idence Interv als
Censoring Column in censor5
ShapeScale
MeanStDev
MedianIQR
1.2257 7.8398
7.3357 6.0165
5.8135 7.3970
0.1 1.0 10.0 100.0
1
2
3
5
10
20
30
40 50 60 70 80
90 95
99
Time to Failure
Pe
rce
nt
Probability Plot for time5Weibull Distribution-95.0% Conf idence Interv als
Censoring Column in censor5
ShapeScale
MeanStDev
MedianIQR
1.2257 7.8398
7.3357 6.0165
5.8135 7.3970
Concepts
O
43Approved for Public Release, Distribution Unlimited, GDLS approved, log 2006-24, dated 04/26/06
Optimize Phase The Optimize phase requires use of process capability information and a
statistical approach to tolerancing. Developing detailed design elements, predicting performance, and optimizing design, take place within this phase.
Crucial Steps: Assess process capabilities to achieve critical design parameters and meet
CTQ limits Optimize design to minimize sensitivity of CTQs to process parameters Design for robust performance and reliability Establish statistical tolerancing Optimize sigma and cost
Key Tools: Design for Manufacturability and Assembly Process capability models Robust design Monte Carlo Methods Tolerancing and Tolerance Design Optimization tools like DOE, Response Surface Methodology, Multiple Y
44Approved for Public Release, Distribution Unlimited, GDLS approved, log 2006-24, dated 04/26/06
Response Surface Methods
Summary: Designed Experiment for optimizing the performance of a Response (Y). Summary: Designed Experiment for optimizing the performance of a Response (Y).
Output:• Set points for the vital x’s to optimize the mean value of the Response (Y)• Set points for the vital x’s to optimize the standard deviation of the Response• Mathematical Model (transfer function) of the Response as a function of the vital x’s.
Output:• Set points for the vital x’s to optimize the mean value of the Response (Y)• Set points for the vital x’s to optimize the standard deviation of the Response• Mathematical Model (transfer function) of the Response as a function of the vital x’s.
NP
N B
eta=
50
NP
N B
eta=
102.
5
NP
N B
eta=
155
NP
N B
eta=
207.
5
NP
N B
eta=
260
NP
N B
eta=
312.
5
NP
N B
eta=
365 Impl SR=1500
Impl SR=1700
Impl SR=1900
Impl SR=2100
Impl SR=2300
Impl SR=2500
14.2
14.4
14.6
14.8
15
15.2
15.4
Response Surface - Regulation Voltage
44
45Approved for Public Release, Distribution Unlimited, GDLS approved, log 2006-24, dated 04/26/06
Goals of Robust Engineering
1 - Identify the ideal function for the product or process design
2 - Through multi-variable experimentation, chose the nominal design parameters that optimize performance in the presence of factors that cause variability at the lowest cost. This is done through two step optimization process: Step 1: Minimize Variability Step 2: Shift mean to Target
46Approved for Public Release, Distribution Unlimited, GDLS approved, log 2006-24, dated 04/26/06
Robust Design – P-Diagram
3) Customer Usage and Duty Cycle:
4) External Environment Conditions:
5) Internal Environment (System Interaction):2) Wear Out:
1) Piece to Piece Variation:
Ideal FunctionItemSignal
Error StatesControl Factors
(Customer Intend) (Intended Result)
(Noise Factors)
(Unintended Result)
12
3
4
5
47Approved for Public Release, Distribution Unlimited, GDLS approved, log 2006-24, dated 04/26/06
Step 1 -Minimize Variability
Req
uire
men
t
Tar
get
Pro
babi
lity
Den
sity
Performance
48Approved for Public Release, Distribution Unlimited, GDLS approved, log 2006-24, dated 04/26/06
Step 2 -Shift Mean to Target
Req
uire
men
t Tar
get
Pro
babi
lity
Den
sity
Performance
49Approved for Public Release, Distribution Unlimited, GDLS approved, log 2006-24, dated 04/26/06
4.0 Verify
5.0 Verify
Functionality
DFSS-CDOVConcept
Generation&
Selection
Design Optimize
What are the customer requirements?
What approach/ system architecture would best meet customer’s requirements?
How can we optimize forRobustness?
How can weensure thatthe customerneeds will beconsistently metin variousenvironmentsand situations?
How can we design to meet the customer needs?
What measurable system requirements supportthe customer needs?
How will we demonstrate robustness/reliability?
How can weverify that this product is reliableand robust?
Verify
50Approved for Public Release, Distribution Unlimited, GDLS approved, log 2006-24, dated 04/26/06
DFSS (DCOV) Flow of Analysis & Tools
Signifies the OPTIMUM S/N Set point…Signifies the OPTIMUM S/N Set point…
Functional ModelingConcept Generation
& Selection
DoERobust Design & Tolerance Design
Functions
VOC KJ QFD
Customer Needs/Statements
Customer Requirements
FMEA
Technical Requirements
Reliability/Robustness Demonstration
0.1 1.0 10.0 100.0
1
2
3
5
10
20
30
40 50 60 70 80
90 95
99
Time to Failure
Pe
rce
nt
Probability Plot for time5Weibull Distribution-95.0% Conf idence Interv als
Censoring Column in censor5
ShapeScale
MeanStDev
MedianIQR
1.2257 7.8398
7.3357 6.0165
5.8135 7.3970
0.1 1.0 10.0 100.0
1
2
3
5
10
20
30
40 50 60 70 80
90 95
99
Time to Failure
Pe
rce
nt
Probability Plot for time5Weibull Distribution-95.0% Conf idence Interv als
Censoring Column in censor5
ShapeScale
MeanStDev
MedianIQR
1.2257 7.8398
7.3357 6.0165
5.8135 7.3970
Concepts
V
51Approved for Public Release, Distribution Unlimited, GDLS approved, log 2006-24, dated 04/26/06
Verify Phase The Verify phase consists of testing and verifying the design. As increased
testing using formal tools occurs, feedback of requirements should be shared with manufacturing and sourcing, and future manufacturing and design improvements should be noted.
Crucial Steps: Prototype test and validation Assess performance, failure modes, reliability, and risks Design iteration Final phase review
Key Tools: Accelerated testing Reliability/Robustness Demonstration and Growth FMEA Disciplined New Product Introduction (NPI)
52Approved for Public Release, Distribution Unlimited, GDLS approved, log 2006-24, dated 04/26/06
Design Verification
Product Design Verification Functional Performance Verification Operational Environment requirements verification Reliability requirements verification Usage requirements Verification Safety requirements verification
Manufacturing Process Verification Process Capability verification Production Throughput Verification Production cost verification
53Approved for Public Release, Distribution Unlimited, GDLS approved, log 2006-24, dated 04/26/06
Reliability Assessment
Measure reliability performance of Robust subsystems, subassemblies, and components Predicted reliability vs. Measured Reliability Reliability Growth Activities
• Measure Reliability, Availability and Maintainability (RAM) of equipment, manufacturing processes and assemblies.
54Approved for Public Release, Distribution Unlimited, GDLS approved, log 2006-24, dated 04/26/06
Questions to Ask to Ensure I have Selected an Appropriate Individual as a Lean Six Sigma Black Belt Candidate
High (2) Med (1) Low (0)
Executes both effectively and efficiently?
Demonstrated change agent?
Remains on task, not easily sidetracked?
Thinks logically, connects Y to projects?
Possesses executive presence?
Thinks strategically and acts tactically?
Understands business and customer needs?
Aptitude for LSS and statistics?
Follows program management methodologies?
Propensity to both learn and teach?
Strong candidates will have scores of 15 and higherJ. Wasiloff 4 Jan 2006
55Approved for Public Release, Distribution Unlimited, GDLS approved, log 2006-24, dated 04/26/06
Strategic Project Selection
Typically need four to five ideas to generate one good Design for Six Six Sigma project
Some projects may be cancelled during one of the phases of the process
Employ a methodology for consistent ranking of projects Ensure you have a plan for gathering ideas.
Top down Best impact on customer or ROI Aligns with strategic goals
Bottom up Knows the problem processes Knows the customer best Influences cultural change
56Approved for Public Release, Distribution Unlimited, GDLS approved, log 2006-24, dated 04/26/06
Training References (External)
American Supplier Institute (ASI) American Society for Quality (ASQ) Eastern Michigan University University of Michigan Moresteam University (on-line) PDSS Six Sigma Management Institute Sigma Pro SBTI
57Approved for Public Release, Distribution Unlimited, GDLS approved, log 2006-24, dated 04/26/06
Literature References Creveling, Clyde M., Slutsky, Jeffrey Lee and D. Antis, Design for Six Sigma in
Technology and Product Development. Prentice Hall, 2002
Yang, Kai and El-Haik, Basem. Design for Six Sigma: A Roadmap for Product Development
Fowlkes, William Y. and Creveling, Clyde. Engineering Methods for Robust Product Design. Addison-Wesley, 1995.
O'Connor, Patrick D. T. Practical Reliability Engineering. John Wiley and Sons, 1991.
Reklaitis, G. V., Ravindran, A. and Ragsdell, K. M. Engineering Optimization. John Wiley and Sons, 1983.
Suh, Nam P.. The Principles of Design. Oxford University Press, 1990.
58Approved for Public Release, Distribution Unlimited, GDLS approved, log 2006-24, dated 04/26/06
Q&A
Thanks for your participation in this technical interchange session