6 6 an introduction by habib siddiqui, ph.d. [email protected]
TRANSCRIPT
6
•US has lost 3.3 million manufacturing jobs during March 1998-2006 (before current recession)•Supply chains have lowered inventories by $4.6 trillion dollars•Inquiries on your computer and phone accounts are often answered by outside call centers•Your tax return is completed in India•Your Jet Blue reservation is taken by Betty in her house coat and slippers in Salt Lake City•Dad’s X-rays are read overseas at 2:00 a.m. while you are sleeping•US share of scientific papers in the Physical Review Letters fell from 61% in 1983 to 29% in 2003•Mama-papa stores are vanishing from neighborhoods at an exponential rate. 1/3 of the US population shop in the Wal-Mart every week (Wal-Mart has 60,000 suppliers)•Downstream customers and retailers are placing increasing focus on what goes upstream: how things are made, framed, packaged, transported, displayed and sold worldwide (Life Cycle Thinking)
“We need to be telling our kids to hurry up and eat and get to their homework - there are kids in China and India who are starving for our jobs”
- Tom Friedman, The World is Flat
Have you noticed ?…
Competitive Realities of Our TimeThe Global Business Climate
• Fierce, global competition• Accelerating pace of change• New technologies• Increasing customer demands - performance,
quality, price, “solutions, not products,” ...
Quality & Performance: Higher expectations
Cost:Only low cost
providers will survive
Speed: Short product life cycles,
“e-business” mindset
6
6
This sentiment is echoed by a CEO…
“As we envision the environment in which pharmaceutical companies will be operating in the years ahead, we believe – and we’re hardly alone in this belief – that those companies that develop competitively priced, novel medicines that demonstrably improve the health and wellbeing of patients will prosper.”
- Raymond Gilmartin, Merck & Co. [March 1, 2004]
6
New Initiatives to Meet the Realities
Speede-Engineering,
Lean Enterprise,Streamlining SC
CostTarget Costing, Globalization
Quality & PerformanceSix Sigma, Design for Six Sigma, Lean Enterprise,
Stage Gate/NPI, Risk Management
Lower Cost
BetterQuality
Quicker to Market
Higher Performance
6
“Eighty-five percent of the reasons for failure to meet customer expectations are related to deficiencies in systems and process…
rather than the employee.
The role of management is to change the process rather than badgering individuals to do better.”
In other words, look for the root causes, not the “root who.”
Why Six Sigma? – Dr Deming
6
Six sigma business improvement starts with SIPOC
Suppliers Inputs(Business)
Process
(Process)
Outputs
(Critical)
Customer(Requirements)
Defects
Variation in the output from whatthe customer wants causes defects.
Root cause analysis leads to permanent
elimination of defects
USLLSL
CCR
Output
* *
6
All processes have variability
All variability has causes
Typically only a few causes are significant
To the degree that those causes can be understood- they can be controlled
Designs must be robust to the effects of the process variation
This is true for products, processes, services, information transfer, everything . . .
. . .is that uncontrolled variation is the enemy
The basic Six Sigma The basic Six Sigma premisepremise . . . . . .
6
What is Six Sigma?
Six Sigma is a powerful set of statistical and management tools and methodologies that can create dramatic increases in customer satisfaction, productivity and shareholder value for both service and manufacturing companies/organizations.
It is a disciplined methodology of defining, measuring, analyzing, improving and controlling the quality in every product, process and transaction – with the ultimate goal of virtually eliminating all defects. (Jack Welch, GE)
6
History of Six Sigma
Motorola (mid-’80s) GE – under Jack Welch (mid-’90s) Others doing it:
- Dow, Witco, DuPont, Rohm and Haas- Ford, GM- Johnson & Johnson, Merck (2000-01)- Maytag- Allied Signal (now Honeywell) - 3M, Kodak, Corning- any many others.
Average six sigma project saves $250M
6
Six Sigma: PhilosophyThe Motorola School: (Show me the “Defect”)Relentless Defect Elimination
– Find and Remove Existing Defects
– Prevent New Defects
The GE School: (Show me the “$$$”)Relentless Pursuit of Financial Opportunities
– Identify High Impact Projects
– Use Six Sigma Methodology to Optimize the Process
– Visible Bottom Line Impact
6
• Defects & yield depend on –Spec width (design requirement)
–Standard deviation (process variation)
i.e., on the Sigma Level
• Reduce defects and improve yieldsby reducing process variation, i.e., by reducing std deviation and thereby raising the Sigma Level
Reducing Defects and Improving Yield
TargetLSL
Defects =
Area > USLDefects = Area < LSL
USL
6
Six Sigma : Measurement
Sigma Level
% Good Defective ppm
2 69.15 % 308,537 3 93.32 % 66,807 4 99.379 % 6,210 5 99.9767 % 233 6 99.99966 % 3.4
Historic standard
New standard
20,000x improvement
6
How Important is Quality?
If your water heater operated at Four Sigma performance, you’d be without hot water more than 54 hours each year.
At Six Sigma, you’d be without hot water less than two minutes a year!
If your goal was 99% quality, you'd still have: 15 minutes of unsafe drinking water
every day 2 unsafe plane landings per day at most
major airports 20,000 pieces of lost mail every hour 200,000 wrong drug prescriptions per
year 5,000 incorrect surgical operations per
week(Ref: Control Engineering, Jan. 1999)
Sigma Levels of Some Activities
1
10
100
1000
10000
100000
1000000
1.5 2 2.5 3 3.5 4 4.5 5 5.5 6
Sigma Level
De
fect
s /
Mill
ion
Prescription Writing
Airline Baggage Handling
US Airline Fatality Rate
World-Class Quality
Companies
Ave US Company
Tax Advice at IRS Help Centers
Restaurant Bills
6
6
The Cost of Variation – Taguchi Loss Function
WhereLSL = lower spec limitNom = targetUSL = upper spec limit
Traditional ViewTraditional View
Cos
t
LSLLSL NomNom USLUSL
AcceptableAcceptable
Correct ViewCorrect View
Cos
t
LSLLSL USLUSLNomNom
CTQ
Cost of Poor Quality – COPQ -consists of those costs which are generated as a result of producing defective material.
COPQ is a financial measure of the user or customer’s dissatisfaction with a product’s performance as it deviates from a target value.
$ COPQ = k (CTQ – Target)2
Tangible Costs• Inspection• Scrap• Rework• Warranties
Intangible Costs• Lost Customers • Longer Cycles
The High Cost of Poor Quality
Enormous opportunityEnormous opportunity
Avg. US Co.
World-Class Co
30%
25%
20%
10%
15%
5%
0%
Sigma Level6 5 4 3 2
Cost of Poor Quality (% of Sales)
6
6
The Focus of Six Sigma:Fix Processes, Not Products
Y
• Output
• Dep Variable
• Effect
• Symptom
• Monitor
x1, x2, …, xn
• Inputs
• Indep Variables
• Root Causes
• Problems
• Fix & Control
Output Y= f (Process Variables x1, x2, …, xn)
• Many quality approaches focus on inspecting and fixing outputs (Y’s, e.g., products)
• Six Sigma focuses on fixing and controlling key process variables (x’s) which cause output defects
6
Six Sigma Methodology“Flavors”
Six Sigma Process Optimization (DMAIC) Manufacturing process Business/service/transaction process
Design for Six Sigma (DFSS) new products new processes new services Redesigning an existing product/service.
6
Six Sigma DMAIC Methodology
D
M
I
A
C
Define what’s important
Measure how we’re doing
Analyze what’s wrong
Improve by fixing what’s wrong
Control to guarantee performance
Six Sigma is information dependent.
6
The Six Sigma Filtering EffectThe Six Sigma Filtering Effect
Critical Input Variables
100+ Inputs
8 - 10
3 - 6
1 - 3
25 - 30MEASURE
ANALYZE
IMPROVE
CONTROL
• Capability Study• Measurement Study• C& E Matrix
• FMEA• Multi-Vari Studies
• Design of Experiments (DOE)
• Control Plans
Inexpensively, narrow in on fewer and fewer Variables, Saving more “expensive” tools for when we have fewer Variables
• Process Maps
All Possible Variables
Sustain-Certify • Confirm Benefits $ $
Output Y= f(x1, x2, …, xn)
6
Six Sigma Methodology Road Map
Process Maps Cause and Effect
Diagram (Fishbone)
Dot Plots Box Plots Scatter Plots Control Charts Process Capability Measurement
System Analysis Gage R&R P/T Ratio Discrimination
Ratio Multiple
Subjective Evaluation
Destructive Testing
Full Factorial DOE
Fractional Factorial DOE
Power and Sample Size
Taguchi DOE Response
Surface Modeling
EVOP Attribute DOE Mixture DOE Solution
Generation Simulation Piloting TRIZ Kaizen
Cause & Effect Matrix
Process / Product FMEA
Probability Components of
Variation Multi-Vari Analysis Hypothesis Testing Confidence Intervals Test for Equal
Variance Correlation Linear Regression Multiple Regression One-Way ANOVA Two-Way ANOVA Chi Square Binary Logistic
Regress. Non-Parametric
Tests Value Analysis Theory of
Constraints Discrete Event Sim Is-Is Not Analysis
Design FMEA Process Control
Plan CUSUM Charts EWMA Charts Control Metric ISD Procedures COPS / 5S Visual Factory Poka-Yoke Implementation
Plan PERT Diagrams Communication
Plan
Define
Measure
Control
Improve
Analyze
SustainCertify
Effective Meetings Opportunity
Analysis Cost of Poor Quality Asset Utilization Financial Analysis SIPOCC Brainstorming Nominal Group
Technique Affinity Diagram Check Sheets Run Charts Pareto Histogram Market Research Survey Focus Groups Interviews Kano Model QFD Decision Analysis
Matrix Impact-Effort
Analysis Project Charter Gantt Charts
Project Translation
Documentation
Certification
TOOLS:Statistical
Non statistical
6
Six Sigma: Six Sigma: Conceptual ApproachConceptual Approach
Practical ProblemPractical Problem Statistical ProblemStatistical Problem
Statistical SolutionStatistical SolutionPractical SolutionPractical Solution
y f x x x k ( , , . . . , )1 2
6
When to Use Six Sigma
You have a performance gap in a process Driven by the Business strategy High Impact on $$ or Customers Problems that “have withstood the test of time” Root Cause is unknown Solution is not known
You want a more robust solution In the past, solutions fail to “stick”
6
Let’s be smart about this…
IF
You know the solution to your problem / opportunity
You don’t know the solution but you suspect others do
The impact solving the problem / opportunity is small or not strategically important.
THEN
Implement the Solution .
Identify Best Practice and Implement
. Cancel the project
6
Daily Commute Problem
The employee John Doe lives nearly 25 miles away from his workplace. He likes to come to work at 8 a.m. His average commuting time is nearly 46 minutes and depends on many factors. He rides a car that gives him approx. 20 miles/gal. With high gasoline prices (costing ~$2.80/gal), he is reevaluating how to best optimize his time and miles/gal in commuting to work.
How can we help John Doe with his problem?
Define
6
0 50 100
20
30
40
50
60
70
Observation Number
Indi
vidu
al V
alue
I Chart for Commute_
Mean=45.72
UCL=69.25
LCL=22.19
Variable N Mean Median TrMean StDev SE Mean
Commute_ 100 45.720 46.000 45.711 7.716 0.772
Variable Minimum Maximum Q1 Q3
Commute_ 24.000 66.000 40.000 51.000
Define
6
What are the CTQs and COPQ?
CTQs: - Commuting time (defn: time between leaving home and arriving in office)- MPG
COPQ:- Internal: $100M/yr (lose job) plus price differential in gasoline purchase
- External: $1MM (company)
Define
6
What are the variables that affect these outcomes?
Where to start to identify variables? - Process Mapping
Process Mapping involves- Identify steps- Input variables- Types of input variables (controlled or uncontrolled)
- Outputs
Measure
6
Process Map
Home Activities
Car Start-up Activities
RoadActivities
Parking/Arrival
Outputs
Looking good
Well-fedWell-preparedSafety Elapsed time for home activities
Safety
Drive
Milage per gallon
Commuting time
ParkMilage per gallon
Reach the office
Arrival time
InputsVariable
Type
CWake-up time CGo to bath room CDress up CGet the keys CGet the office badge CHave breakfast CCarry bag to car CLock the house door C
Open the car door CPut the seat belt CStart the car CMove car from driveway to road C
Roads/routes CTime of the day CDay of the week CTraffic condition UWeather USeason of the year CGasoline in car CType of gasoline CTire pressure CCondition of tire CCondition of engine CSpeed of driving C
Find Parking spot UStop the car CWalk out from parking COpen the door to building CWalk to office CTurn on the office light CSwitch on the computer C
Measure
6
Which of these 31 variables are the potentially important ones?
How important are the CTQs? Commute Time: what is desirable? – max. 50 min MPG: what is desirable? – 30 mpg
How do we prioritize the input variables?- Use Cause & Effect matrix to find causal relationship.- Use a scale of 0 (no), 1 (minor), 3 (moderate) and 9 (major relationship).
Measure
6
Rating of Importance to Customer 10 5
1 2 3 4Process Step Process Inputs
Type of variable,
controlled or uncontrolled C
om
muting
Tim
e
MP
G
Total
Road Activities Traffic condition U 9 9 135Home Activities Wake-up time C 9 0 90Home Activities Go to bath room C 1 0 10Home Activities Dress up C 1 0 10Home Activities Car Keys C 3 0 30Home Activities Office Badge C 9 0 90Home Activities Have breakfast C 1 0 10Home Activities Carry bag to car C 1 0 10Home Activities Lock the house door C 1 0 10
Car Start-up Activities Open the car door C 1 0 10Car Start-up Activities Put the seat belt C 3 0 30Car Start-up Activities Start the car C 1 0 10
Car Start-up ActivitiesMove car from driveway to road C
1 1 15
Road Activities Roads/routes C 9 9 135Road Activities Time of the day C 9 9 135Road Activities Day of the week C 3 3 45Road Activities Speed of driving C 9 9 135Road Activities Weather U 3 9 75Road Activities Season of the year C 1 3 25Road Activities Gasoline in car C 3 1 35Road Activities Type of gasoline C 1 9 55Road Activities Tire pressure C 9 9 135Road Activities Condition of tire C 3 3 45Road Activities Condition of engine C 3 3 45
Parking Find Parking spot U 3 1 35Parking Stop the car C 1 1 15Parking Walk out from parking C 1 0 10
Parking Open the door to building C
3 0 30
Parking Walk to office C 1 0 10Parking Turn on the office light C 0 0 0
Parking Switch on the computer
C0 0 0
Total 1030
395
0 0
Lower Spec 0 15
TargetUpper Spec 5
0
30
Measure
6
Total
0
20
40
60
80
100
120
140
160
Roads
/rout
es
Speed
of d
rivin
g
Traffic
con
ditio
n
Office
Bad
ge
Type
of g
asol
ine
Condi
tion
of ti
re
Gasol
ine in
car
Car K
eys
Open
the
door
to b
uildi
ng
Mov
e ca
r fro
m d
rivew
ay to
road
Go to
bat
h ro
om
Have
brea
kfast
Lock
the
hous
e do
or
Start
the
car
Wal
k to
offic
e
Switch
on th
e co
mpu
ter
Measure
6
Potentially important Input Variables are:
Process Step Process Inputs
Type of variable,
controlled or uncontrolled C
omm
utin
g T
ime
MP
G
Total
Road Activities Roads/routes C 9 9 135
Road Activities Time of the day C 9 9 135Road Activities Speed of driving C 9 9 135Road Activities Tire pressure C 9 9 135Road Activities Traffic condition U 9 9 135Home Activities Wake-up time C 9 0 90Home Activities Office Badge C 9 0 90Road Activities Weather U 3 9 75Road Activities Type of gasoline C 1 9 55
Measure
6
Further narrow down of variables
Use Failure Modes & Effects Analysis (FMEA) Rate for severity, S (1=min, 10=max) Rate for occurrence, O (1=least, 10=most) Rate for current control/detection, D (10=none,
1=certain detection) Risk Priority Number, RPN= SxOxD
Analyze
6
What is the process step/
Input under investigation?
In what ways does the Key Input go wrong?
What is the impact on the Key Output Variables (Customer Requirements) or internal requirements?
How
Severe
is t
he e
ffect
to t
he c
usotm
er? What causes the Key Input to
go wrong?
How
oft
en d
oes c
ause o
r
FM
occur? What are the existing controls
and procedures (inspection and test) that prevent either the cause or the Failure Mode? Should include an SOP number.
How
well
can y
ou d
ete
ct
cause o
r F
M?
Route taken Took surface roads Takes longer time5
Too many stop signs; did not carry cash (can't pay toll) 4
Road map2 40
Time of the day Rush hour (left home after 7 a.m.)
Takes too long 9
Too much traffic on the road9
Clock, watch1 81
Time of the day Rush hour (left home after 7 a.m.)
Takes too long 9
Trash collection day3
Weekly event1 27
Traffic Condition Bad Stranded, takes longer7
Accident on the way2
None10 140
Traffic Condition Bad Stranded, takes longer7
Gaper delay2
None10 140
Speed Low speed driving Takes too long, consumes more gasoline 9
Traffic condition is bad3
Speedmeter in car1 27
Speed High speed driving Takes less time, but can get ticketed 9
Less traffic2
Speedmeter in car1 18
Tire pressure Low pressure Consumes too much gasoline, can also get flat - thus requiring repair
9
Lack of inspection
2
Visual
2 36
Office badge Missing badge at work Go to guard house and sign up for a temporary badge 4
Misplacement of badge at home, forgot 2
None10 80
Wake-up time Woke up after 6 a.m. Late for work10
Alarm did not work1
Alarm clock1 10
Wake-up time Woke up after 6 a.m. Late for work10
Went to sleep very late3
Wall clock, spousal complain; alarm clock 1 30
Wake-up time Woke up after 6 a.m. Late for work10
Sick1
Healthy living, good diet; alarm clock 2 20
Weather Snow or rain Late to work9
Nature/seasonality4
None10 360
Type of gasoline Low octane Consumes more gasoline2
Savings per gallon9
Gas pumps1 18
Analyze
6
What is the process step/
Input under investigation?
In what ways does the Key Input go wrong?
What is the impact on the Key Output Variables (Customer Requirements) or internal requirements?
How
Severe
is t
he e
ffect
to t
he c
usotm
er?
What causes the Key Input to go wrong?
How
oft
en d
oes c
ause o
r
FM
occur? What are the existing controls
and procedures (inspection and test) that prevent either the cause or the Failure Mode? Should include an SOP number.
How
well
can y
ou d
ete
ct
cause o
r F
M?
Weather Snow or rain Late to work9
Nature/seasonality4
None10 360
Traffic Condition Bad Stranded, takes longer7
Accident on the way2
None10 140
Traffic Condition Bad Stranded, takes longer7
Gaper delay2
None10 140
Time of the day Rush hour (left home after 7 a.m.)
Takes too long 9
Too much traffic on the road9
Clock, watch1 81
Office badge Missing badge at work Go to guard house and sign up for a temporary badge 4
Misplacement of badge at home, forgot to bring 2
None10 80
Route taken Took surface roads Takes longer time
5
Too many stop signs; did not carry cash 4
Road map
2 40
Tire pressure Low pressure Consumes too much gasoline, can also get flat - thus requiring repair
9Lack of inspection
2Visual
2 36
Wake-up time Woke up after 6 a.m. Late for work
10
Went to sleep very late
3
Wall clock, spousal complain; alarm clock 1 30
Time of the day Rush hour (left home after 7 a.m.)
Takes too long 9
Trash collection day3
Weekly event1 27
Speed Low speed driving Takes too long, consumes more gasoline 9
Traffic condition is bad3
Speedmeter in car1 27
Wake-up time Woke up after 6 a.m. Late for work
10
Sick
1
Healthy living, good diet; alarm clock 2 20
Speed High speed driving Takes less time, but can get ticketed 9
Less traffic2
Speedmeter in car1 18
Type of gasoline Low octane Conumes more gasoline2
Savings per gallon9
Gas pumps1 18
Wake-up time Woke up after 6 a.m. Late for work10
Alarm did not work1
Alarm clock1 10
Analyze
6
Input Variables RPN TypeTraffic Condition 140 UTraffic Condition
140 WWeather 108 U
Time of the day 27 CSpeed 27 C
Route taken 20 CWake-up time 20 C
Type of gasoline 18 CWake-up time
10Wake-up time
10 CTire pressure 9 C
Time of the day9 C
Office badge 8 C
Analyze
6
Actions Recommended
Resp. Actions TakenSEV
OCC
DET
RPN
What are the actions for reducing the
occurrance of the Cause, or improving detection? Should
have actions only on high RPN's or easy
fixes.
Who is Responsible for
the recommended
action?
What are the completed actions taken with the recalculated RPN? Be
sure to include completion month/year
Can't do much; uncontrolled event
God Leave home early, watch weather report 9 4 3 108
Can't do much; uncontrolled event
God Hope 7 2 10 140
Can't do much; uncontrolled event
God Hope 7 2 10 140
Leave before -7 a.m. Employee Leave between 6:40-7 a.m. 9 3 1 27
Put it near eye-glass or inside bag
Employee Put inside attache/bag4 2 1 8
Take Freeway/TPk Employee Take Freeway/TPk
5 2 2 20
Check tire pressue upon arrival at home and before start-up
Employee Check tire pressue upon arrival at home and before start-up
9 1 1 9
Go to sleep before 11 p.m., use alarm clock
Employee Go to sleep before 11 p.m., use alarm clock 10 2 1 20
Leave the trash for pick-up the previous night
Employee Leave trash for pick-up on Sunday 9 1 1 9
9 3 1 27
Exercise 2 days per week, periodic medical checkup
10 1 1 10
9 2 1 18
2 9 1 18
10 1 1 10
Process Step/Input
Potential Failure Mode
What is the process step/
Input under investigation?
In what ways does the Key Input go wrong?
Weather Snow or rain
Traffic Condition Bad
Traffic Condition Bad
Time of the day Rush hour (left home after 7 a.m.)
Office badge Missing badge at work
Route taken Took surface roads
Tire pressure Low pressure
Wake-up time Woke up after 6 a.m.
Time of the day Rush hour (left home after 7 a.m.)
Speed Low speed driving
Wake-up time Woke up after 6 a.m.
Speed High speed driving
Type of gasoline Low octane
Wake-up time Woke up after 6 a.m.
Analyze
6
Experiment with new procedure to check hypothesis
0 50 100
20
30
40
50
Observation Number
Individu
al V
alue
I Chart for New_Time
Mean=35.39
UCL=49.60
LCL=21.18
Variable N Mean Median TrMean StDev SE Mean
New_Time 100 35.390 36.000 35.456 4.788 0.479
Variable Minimum Maximum Q1 Q3
New_Time 24.000 46.000 32.000 38.000
Improve
6
0 100 200
20
30
40
50
60
70
Observation Number
Indi
vidu
al V
alue
I Chart for Time by Experime
Mean=35.39
UCL=49.60
LCL=21.18
Old New
Improve
nt
6
Gas Mileage per Gallon
Initial gas consumption = 20 mpg Desire to get better mileage Key input variables for MPG are
- gasoline octane rating- tire pressure- driving speed
Full Factorial DOE with 3 variables @ 2 levels would require: 2x2x2 = 8 experimental runs.
Improve
6
Problem: Gas Mileage is 20 mpg
What conclusion do you make now?
Speed Octane Tire Pressure Miles per Gallon55 85 30 2565 85 30 2355 91 30 2765 91 30 2355 85 35 2765 85 35 2455 91 35 3265 91 35 25
MPG = f(Speed, Octane, Tire Pressure)
Do we think 32 is best?Do we think 32 is best?
Full Factorial ExperimentImprove
6
Control Plan During rainy and winter seasons: leave home early,
watch weather report before leaving home. In normal times: Leave between 6:40-7 a.m. Put office badge inside wallet Take Freeway + Turnpike whenever possible Check tire pressure upon arrival at home and before
start-up Go to sleep before 11 p.m., use alarm clock to get up
on time. Leave trash for pick-up on Sunday evening For best mileage: ensure tire pressure @35 psig, drive
@ 55 mph, and use 91 octane rating gasoline.
Control
6
Cost Of Poor Quality over the product life cycle (from lab to customer)
Most current six sigma effort is here
R&D/Discovery, Pre-
clinical
Pilot plant/Clinical I-III
Production/Commercialization
Customer
Defects are:
Difficult to see/predictEasy to fix
Easy to seeCostly to fix May lose customers
We ought to do here
We need a paradigm shift in how we do things: from reactive to predictive mode.
6
» Six Sigma Practices in Manufacturing is Not Enough» Cannot Produce a Six Sigma Product via Mfg Alone
Up to 4.5
4.5
5
The “5 Wall”
Achievable via Mfg Improvements
Law of Diminishing Returns in Mfg
Requires Product Designed for 6DFSS
Customers don’t care about the Mfg Processes
Customers want Product Performance, Reliability & Durability
Why DFSS?
D-M-A-I-C has been around for over 10 years, but…
6
Underlying Truism: Knowing what the customer needs
We don't know what we don't know, we can't act on what we don't know, we won't know until we search, we won't search for what we don't question, we won't question what we don't measure, and hence we just don't know.
6
Customer Service 101: Know what is “good” to your customer … (VOC)
Ichiro Ishikawa:
“When I ask the designer what is a good car, what is a good refrigerator and what is a good synthetic fiber, most of them cannot answer. It is obvious they cannot produce good products.
You simply cannot design a good product or service if you do not know what “good” means to the customer.
The designer must create a map that moves the world of customer to the world of the designer”
6
Customers: how important are they?
Customer loyalty for strategic partners is very important. Keeping an old customer happy is more fruitful than finding a new one. (It is easy to retain five satisfied customers than to find a new one.)
One happy customer tells three people, 1 unhappy customer tells 20.
6
Voice of the Customer
Customer Satisfaction = f( Perception, Expectation)
Level 1 : Features and CostLevel 2 : QualityLevel 3: Features, Cost, Quality, Delivery,…. Value added
1
2
3
Customer value =
Higher Quality
Better Servicex
Lower Cost
x Less Time
6
Kano Model (a 2-d concept of quality)
Satisfied
Dissatisfied
Good performancePoor performance
With Time
6
And we often do ….
Insane things.
Insanity (definition):Continue to do things we have always done
and yet expect to get different results.
6
A Specific Product Design Methodology where Customer Requirements Dictate the Critical Parameters and the Variability of the Critical Parameters are Optimized for Predictive Product Performance,
Manufacturability, Reliability and Durability.
Define
Identify
Design
Optimize
Validate
Critical-To-Quality (CTQ) Parameters, Set Technical Requirements & Quality Targets
Concept Design, Develop Transfer Functions between CTQ’s and Design Spec’s… Y=f(x)
Analyze & Optimize for Robust Performance, Predictive Manufacturability, and Reliability
Test & Validate Predictions, Assess Performance, Initial Capability Studies
Marketing
Science/Engg.
Sta
tist
ical
Thi
nkin
gU
nder
stan
d/C
ontr
ol V
aria
tion
s
DFSS Overview: Alternative Roadmap
Customer Requirements Based on Expectations, a.k.a: Voice of the Customer (VOC)
6
What’s Different About DFSS?
Disciplined, comprehensive process Line of sight from customer CTQs to all design levels Statistical design to understand and reduce variation “New” tools: QFD, DOE, Robust Design, DFM, statistical
tolerancing, multi-variable optimization, ... Quality prediction throughout development
6
Integrating DFSS with Stage-Gate/NPI
Tollgates 1 2 3 4 5NPI Stages
DF
SS
Ele
men
ts
DFSS: Identify Design Optimize Validate
• Business needs• Mkt reqts
• Alpha drawings
• Rev 1 drawings
• Rev 2 drawings
CTQs
SystemsEng
• Quality targets
• Updated scorecard: prediction,
capability data
• Scorecard verification
• Product Z estimates
Specs
Detailed Design
Prototype & Test
Mfg Preparation
Prediction
Quality
Design for Perf
Product
Marketing & Concept Design
Design for Prod
• Customer CTQs
• QFD 2: Reqts to part specs
• CTQ estimates
• First piece CTQ
verification
• Final system CTQ
verification
• Tech feasibility
assessment
• Alpha test performance verification
• Pilot run performance verification
• Mfg process concept
• Tolerancing• Proc capability
compatibility
• Tooling procurement
• Final mfg reqts
• Supplier qual
• Improved perf based on eng
analysis & tests
• Improved transfer fcts:
eng analysis & DOE
• System concept
• Functional block diagram
• Prototype confirmation
• Model validation
• System & component
qualifications
• Initial Z allocations & scorecard:
opinion, entitlement
• Prelim perf based on
subsystem specs & initial tests
• Prelim mfg reqts • Key process
capability data
• CTQ flowdown • Subsystem
reqts & transfer functions
Preliminary Design
• QFD 1: Customer CTQs
to tech reqts
• Preliminary product specs
Columns: Items required for tollgate review
Columns: Items required for tollgate review
Stage-Gate/NPI: Bus/mgt process for new product dev
Stage-Gate/NPI: Bus/mgt process for new product dev
Rows: Progressive refinement of DFSS elements
Rows: Progressive refinement of DFSS elements
DFSS: Technical process - how product is
designed
DFSS: Technical process - how product is
designed
6
Chairman Cho of Toyota: Three Keys to Lean Leadership
Go See
“Senior Management must spend time on the front lines”
Ask Why
“Use the “Why?” technique daily”
Show Respect
“Respect your people.”
Fujio Cho, Chairman of Toyota Motor Corp.
Reproduced with permission of John Shook, Lean Enterprise Institute
6
Lean Methodology
Specify ValueSpecify ValueIdentify the
Value StreamIdentify the
Value Stream
PullPullFlowFlow PerfectionPerfection
Reference: Lean Thinking, J. Womak and D. Jones, 1996
A Simple, Intuitive Approach!!
Focus on Value from the eyes of the customer (VA and NVA activities)Eliminate waste (Muda)Reduce cycle time / increase speedIn lean design – create flow and value for the customer / minimize waste
6
Value Add vs. Non-Value Add
Type of Process Step Definition Action
Value Add (VA) Transforms or shapes a product/service which is eventually sold to a customer
Identify
Non-Value Add (NVA) Take time, resources, or space but do not add value to the product or service
Eliminate or redesign
Business Value Add (BVA) Not seen as valuable by the customer, but required in order to complete the process
Remediate or Continuously improve
6
What is Waste?
Activity that consumes resources without creating value for the customer
Unevenness in a process (“Hurry up and wait”)
Overburdening people or equipment
6
7 Types of Waste
Processing
Motion
Defects Over-
Production
Transportation
Inventory
Waiting Time
TypesOf
Waste
Returns
Incoming material rejects
Multiple sign offs
Over design of packaging
Expired, unused promotional materials
Unnecessary Warehousing
Multiple handoffs
Searching for information
Multiple data bases
Excess promotional
material
Over forecasting
Slow decision making
Late sales material
TIMWOOD
6
ManyHand-offs
Poor Quality
Lack ofTraining
Poor AccessTo Data
PoorScheduling
Sea of Excess Transactions, Staffing, Capital
Some Drivers of Waste
6
A Lean Roadmap
Strategy Strategy DeploymentDeployment
ID & Select ID & Select KeyKey
ProcessesProcesses
Project Project CharteringChartering
Understanding Understanding As-IsAs-Is
Plan Event / Plan Event / Design Future Design Future
StateStateProject ClosureProject Closure
• Link company strategy to lean initiatives
• Develop plan to identify key value streams to be addressed
• Determine the type of event and resources required
• Train participants
• Generate Current State
• Identify opportunities
• Create Future State
• Plan for improve-ments
• Track results
• Transfer knowledge
• Conduct events
• Drive for results
Implement Implement ImprovementsImprovements
• VOC
• VOP
• CM
6
Project Selection Guideline
New Process/ Business/Product/Service?
Improve Process/ Business/ Product/ Service by reducing standard deviation and
shifting the mean?
Reduce Waste in Process/Business/Product/Service?
No
No
No
Just do it!
DFSS
DMAIC
Lean
Yes
Yes
Yes
Needs redesign because of reaching entitlement?
Yes
No
6
Which Problem Solving Method? – Another Look
LeanLean• Unacceptable time to get things done
• Large amounts of waste/wasted effort
• High, unexplained variation in key metrics
• Lack of optimal settings for process X’s
DMAICDMAIC
• A current design that cannot achieve desired result
• No current design at all (clean sheet)
DFSSDFSS
If You Are Experiencing . . . You probably need . . .
6
Practical Statistics
The long-range contribution of statistics depends not so much upon getting a lot of statisticians into industry as it does in creating a statistically minded generation of physicists, chemists, engineers, and others who will have a hand in developing and directing the production processes of tomorrow.
– W. A. Shewhart and W. E. Deming (1939)
6
Six Sigma …The right projects
+
The right people
+
The right roadmap and tools
+
The right support
=
The right results