ASQ Six Sigma

Workshop on Structured Problem Solving Using DMAIC Methodology
Agenda
Day 1 TQM Introduction to Six Sigma Define Phase Project Charter Day ! Seven QC Tools Process Mapping
Day " FM A Process Capa!ilit" Control Charts Day # #ean nterprise Day $ $SM xercise %TPC Case Stud" xam & 'rap (p
The volution of Qualit"

) Provides a frame*or+ for understanding the histor" of the ,ualit" movement) xpands the definition of ,ualit"
.uran/s Definition of Qualit"

Defined as 0fitness for use1 !ased on2 ) Customer/s perceptions of product design) Degree to *hich a product conforms to design) Product/s availa!ilit"3 relia!ilit"3 and maintaina!ilit") Availa!le customer service
IS4 Definition of Qualit"

) Degree to *hich a set of characteristics fulfills re,uirements
5e,uiremen ts2 Convenienc e and speed
Cros!"/s Definition of Qualit"
) Qualit" is conformance to re,uirements) 5e,uirements are ans*ers to +e" organi6ational ,uestions2

7o* ,uic+l" *ill orders ship8 'hat is our return polic"8 'hat forms of pa"ment are accepta!le8
Qualit" volution2 Medieval 9uilds
9uilds2 uDeveloped strict rules for products and servicesu(sed stamps to identif" fla*less goods-
1 2 0 0
Qualit" volution2 Product 4rientation
uMaster
cra%tsmen trained apprentices& uIndustrial 'evolution divided trades into speciali(ed tasks) inspectors guaranteed *uality& u+aylor system increased productivity) inspection departments %ound de%ects&
Qualit" volution2 Process 4rientation
uProcesses !ecame uShe*hart identified
,ualit" controluDeveloped strict rules for products and servicesuQualit" !ecame relevant for process3 not just product-
criticalstatistical
Qualit" volution2 'artime
uQualit"
issueuThe militar" developed a sampling inspection s"stem and trained suppliers-
!ecame a safet"
Qualit" volution2 Total Qualit" Movement
uDeveloped
in response to .apanese ,ualit" movementuFocused on improving all processes through people *ho used them-
'- d*ards Deming

Qualit" +e"s2 ) (nderstanding customer needs ) Process improvement ) Statistical anal"sis ) xpertise of *or+ers ) PDCA c"cle
.oseph M- .uran
Qualit" +e"s2 ) Features that satisf" customers ) Freedom from deficiencies ) .uran Trilog":
; Qualit" planning ; Qualit" control ; Qualit" improvement
<aoru Ishi+a*a
Qualit" +e"s2 ) Compan"=*ide participation ) Qualit" control circles ) Advanced statistical methods and tools ) %ation*ide ,ualit" control promotion
Armand $- Feigen!aum
Qualit" +e"s2 ) Total ,ualit" control ) Integration of ,ualit" development3 maintenance3 and improvement ) Focus on internal and external customers
9enichi Taguchi
Qualit" +e"s2 ) Qualit" should !e designed in) Qualit" should minimi6e deviations from a target) D4 optimi6es performance
Philip Cros!"
Qualit" +e"s2 ) Conformance to re,uirements ) Prevention ) >ero Defects ) Price of nonconformance
Total Qualit" Management
Total ,ualit" management ?TQM@2
) A management approach ) Centered on ,ualit" ) Aased on compan"=*ide participation ) Aimed at long=term success ) Through customer satisfaction
B Cs of TQM
Customer relationships Continuous improvement Compan"=*ide participation
"
Customer Definitions
1
#evels of Customer Satisfaction

1
%oria+i <ano identified three levels2 ) xpected ,ualit" ) Desired ,ualit" ) xcited ,ualit"
Customer Feed!ac+
1
7as t*o parts2 N fforts to capture *hat customers sa" a!out compan"/s productsCservices N fforts to drive feed!ac+ !ac+ into organi6ation
Partnering *ith customer2 N xtension of listening to customer feed!ac+ N Most direct route to customer satisfaction
PDCA
! ) A *ell=+no*n model for continuous process improveme nt is the Plan=Do= Chec+=Act c"cle-
Compan"='ide Participation
"
) #eadership must come from management) All emplo"ees must !e involved) mplo"ee involvement usuall" re,uires emplo"ees to *or+ in cross=functional teams-
mplo"ee Involvement
"
,ene%its ) Improved productivit" and cost reduction ) Increased participation and jo! satisfaction ) 4pportunities for professional development
) )
) )
,arriers 0It 'on/t 'or+ 7ere1 Perception of loss of management authorit" mplo"ees feeling 0used1 0Flavor of the month1
Qualit" Aenefits
) +angible
; Increase in earnings ; Decrease in -aste ; Increase in productivity
) Intangible
; Customer good-ill ; Alignment bet-een business activities
'- d*ards Deming on Qualit"

) Meeting customer needs . -ants / *uality& ) 0uality improves products1services and processes& ) Improved products1services and processes / pro%itability&
A Qualit" Approach Aenefits - - 2mployees

3rgani(ations
Customers
Suppliers
Society
Aenefits to mplo"ees
Pride in products and services .o! satisfaction Improved communications Streamlined *or+ processes 7appier customers Strong customer relationships 9reater jo! securit"C!enefits
Aenefits to 4rgani6ations
Qualit" Studies and Standards

5eleased the Profit Impact of Mar+et Strateg" ?PIMS@ stud"Partnered *ith the Aaldrige recognition programAoth organi6ations support the lin+ !et*een ,ualit" and profita!ilit"-
xternal and Internal Customers
Pu!lication Department
Sales Department
Customer
Aenefits to Customers
Qualit" results in2 ) Increased choices) Improved goods and services) xpectations met or exceeded
Aenefits to Suppliers
) Achievement of performance re,uirements ) Streamlined processes ) fficient communication
) Increased customer satisfaction
Aenefits to Societ"
conomic gro*th and sta!ilit"
Increased emplo"ment opportunities
Product safet"
Process Management
Qualit" improvements are applied to single processes *ithin manufacturingQualit" improvements are applied to all organi6ational activities through process management-
4rgani6ational Process
Introduction to Six Sigma
'hat is #ean Six Sigma8

Introduction ) Six Sigma goal is process perfection through defect reduction) #ean goal is c"cle time reduction through elimination of *aste04nl" those companies that eliminate their defects *ill have *hat it ta+es to *in-1 0Area+through companies strive for DEE percent D F CT=F5 products and services- Larry Bossidy CEO of AlliedSignal
'hat is Six Sigma8

Methodolog" and Improvement Strateg"
) ) Six Sigma is an overall strateg" to accelerate improvements in processes3 products3 and services;create !rea+throughSix Sigma measures ho* effective strategies are in eliminating defects and variations from processes3 products3 and services-
F G f?xD3
H xJ3
H xB
H --I@
Process output ?G@ is a function of ?f@ the inputs ?Ks@(nderstanding and controlling this relationship is a major aspect of Six Sigma projects-
Focus on $ariation
) Sigma ?4@ refers to standard deviation3 a measure of process variation ?smaller is !etter@) Process Sigma is the num!er of units of standard deviations !et*een the process center and the closest specification limit ?larger is !etter@) A Six Sigma process has six standard deviations ?short term@ !et*een the target and the closest specification limit#o*er Spec Target (pper Spec
#o*er Process Control Center
(pper Control 6 Sigma
6 Standard Deviations
Sources of $ariation
Desig n Materi al
Process Capabil ity
Measurem ent System
7arvesting the Fruit of Six Sigma

'(eet )ruit
Design for Manufacturability N 'all = Must Address Designs N 'all = Must Address Designs
================================== M 'all = Must Improve Internall"
M 'all = Must Improve Internall"
Co nti
+o( ,anging )ruit

Seven Basic Tools
==================================
B 'all = Demand improvement B 'all = Demand improvement
"round )ruit
Logic and ntuition
The walls crumble faster when working W T! su""liers and CO#C$%%E#TL& addressing design and "rocess issues
nu ou s
Process Characterization and Optimization
D is c r e t e d a ta
da
Bul* of )ruit
ta
The Focus of Six Sigma
f ?K@
5 Dependent 3utput 2%%ect Symptom Monitor
G
KD
- - - K% Independent Input=Process Cause Pro!lem Control
'ould "ou control shooter or target to get the 9old Medal at 4l"mpics
Six Sigma Defines Pro!lems Statisticall"
GF f ?K@
The product is used to evaluate the process-
Should we focus on the process outputs (Y) or inputs ( )!
G = 4utputs uDependent u4utput u ffect uS"mptom uA!le to Monitor
KD - - - K % = Inputs uIndependent uInputs3 process uCause uPro!lem uControlla!le
The process is used to control the product-
RRS 9ood ?B-Q Sigma@

JE3EEE lost articles of mail per hour (nsafe drin+ing *ater for almost DN minutes each da" N3EEE incorrect surgical operations per *ee+ T*o short or long landings at most major airports each da" JEE3EEE *rong drug prescriptions each "ear %o electricit" for almost seven hours each month
RR-RRRPPS 9ood ?P Sigma@

Seven articles lost per hour 4ne unsafe minute ever" seven months D-O incorrect operations per *ee+ 4ne short or long landing ever" five "ears PQ *rong prescriptions per "ear 4ne hour *ithout electricit" ever" BM "ears
Define Phase
The Define Phase

The define phase module provides an overvie* of the follo*ing tools2
D-Project selection J-Project charter B-Supplier3 Input3 4utput3 Customer ?SIP4C@ diagram M-Collecting $oice of the Customer ?$4C@
Projects Must !e I Meaningful And Manageable!
Sorting Projects from Messes

) A mess is a morass of unsettling s"mptoms3 causes3 data3 pressures3 shortfalls3 opportunities3 etc) A "roblem is a *ell=defined situation that is capa!le of resolution ) Identif"ing a "roblem from *ithin the mess is fre,uentl" the first step in the process of project definition
Project Qualifications
) There is a gap !et*een current and desired performance) The cause of the pro!lem is not clearl" understood) The solution is not predetermined
Project Selection is Critical

) 7igh leverage projects lead to largest LL Savings ) #arge returns justif" the investment in time and effort ) Developing a #ean Six Sigma culture depends upon successful projects having significant !usiness impact ) Alac+ Aelt training depends on completion of a meaningful project *ithin T P months ) Future Projects are fre,uentl" identified through initial projects
#ean Six Sigma Project Criteria

) Aligned *ith !usiness o!jectives and plans ; $oice of CustomerC Critical to Satisfaction ?CTS@ ; Qualit" ?CTQ@ C Cost ?CTC@ C Deliver" ?CTD@ ) Consistent *ith principles of Six Sigma ; limination of process defects ; 5eduction of variation ) Concentrates on significant issuesCopportunities ----- not 0pro!lem of the da"1 ) .ustif" the investment
Project Selection
0The !est Six Sigma projects !egin not inside the !usiness !ut outside it3 focused on the ,uestion U ho* can *e ma+e the customer more competitive8 'hat is critical to the customer/s success8 I 4ne thing *e have discovered is that an"thing *e do to ma+e the customer more successful inevita!l" results in a financial return to us-1
.ac+ 'elch Address to the 9eneral lectric annual meeting April JB3 DRRO
'hat is Customer Satisfaction8

) A comparison of expectation to experience ) A matter of degree ) A result of a good match !et*een suppl" and demand ) A predictor of repeat !usiness
Customer %eeds Translated into 0Critical To1 ?CT@ Characteristics

Customer needs are translated into product3 service or delivera!le re,uirements in terms of ,ualit"3 deliver" and cost- The 0G1 of GFf?K@+ypical Critical to Characteristics C+0 Critical to Qualit" Critical to Deliver" Critical to Cost
C+D C+C
Area+do*n of ) Projects are identified !" the relationship !et*een the processes product3 service3 or delivera!le re,uirements and processesre,uired to produce the ) The process parameters that affect the re,uirements are later identified ?K 3 K 3 I K @ product3 D J n service3 or delivera!le- #everage processes are identified-
Project Selection Criteria

Pain area Faster Deplo"ment
Aligned *ith core o!jectives
Sta+eholder Satisfaction 7igh pro!a!ilit" of success S& 6o& D J B M N P O Q ase of implementation R DE Criteria
Aligned *ith core o!jectives 7igh pro!a!ilit" of success Data Availa!ilit" Pain area Process Improvement 7igher returns 5epeata!le Faster Deplo"ment Sta+eholder Satisfaction ase of implementation
5epeata!le
Weight DE DE Q Q O Q P N O P
7igher returns
Process Improvement Data Availa!ilit"
Project Selection Matrix Diagram

Pro7ect 1 S& 6o& Criteria
D J B M N P O Q R DE Aligned *ith core o!jectives 7igh pro!a!ilit" of success Data Availa!ilit" Pain area Process Improvement 7igher returns 5epeata!le Faster Deplo"ment Sta+eholder Satisfaction ase of implementation Moderate 5elationship 'ea+ 5elationship
Pro7ect !
Pro7ect "
Pro7ect #
Pro7ect $
Soot ,lo-ing Sy&3ptimi(ation
APC 'eduction in ,oiler 8Stage II9 stage I 3ptimi(ation
DM 8Stage I9 Make up 2SP 8Stage I9 Inlet 3ptimi(ation Duct 'eplacement
Strong 5elationship
" #
Selected Project
Factors that Improve Project Success

) Dedicated Alac+ Aelt ) Champion phase revie*s ) Alac+ Aelt has +no*ledge of processCproduct to !e improved ) 7istoric data ) Clearl" defined delivera!les ) Committed process o*ners ?s+in in the game@ *ith the authorit" to modif" the process
Project Metrics ; Success Criteria

Primar" Metric ) (sed to measure project success ) Consistent *ith the pro!lem description and o!jective ) Plotted on a time series graph and sho*s the goal and actual performance lines
Secondar" Metric?s@ ) Control unintended negative conse,uence ?assures the Primar" Metric is not achieved artificiall"@ ) Ma" !e used to measure project progress *hen the Primar" Metric responds slo*l" ) More than one ma" !e re,uired ) Plotted on a time series graph and sho*s the goal and actual performance lines
Project Charter
Traditionall" created !" #SS Champion Specifies details of a project including2
D-Scope J-5esponsi!ilities B-Aenefits M-Schedule N-Success criteria

A project charter template that ma" !e adopted to fit "our organi6ation is2 Project Charter Template-
SIP4C
Supplier3 Input3 Process3 4utput3 Customer diagrams are used to2 D-Define the scope of the project J-Identif" +e" sta+eholders B-9ain a 0BE3EEE foot1 vie* of the process targeted !" the project Tips2 D-The SIP4C is the first of several *a"s the process *ill !e documented- Therefore3 it should !e at a relativel" high level of a!stractionJ-It is a good *a" to assure agreement on the scope of the project
SIP4C Diagram ; Questions to Ans*er

) 'ho are the customers of this process8 ) 'hat are their re,uirements8 ) 7o* are those re,uirements reflected in the process parameters ?output measures@8 ) 'hat are the process outputs8 ) 'hat are the process inputs that cause the outputs8 ) 'hat controls are in place for the inputs8 ) 'ho are the suppliers of the material for this process8 ) 'hat are their re,uirements8
Purpose of $arious Diagrams C Maps

) SIP4C diagrams provide a 0forest vie*1 of the process ; maintains focus on the customer/s re,uirements) $alue stream 0trees level1 descri!es the time3 effort3 resources3 and information used in the process ; a fre,uent source of earl" *ins) #a"=out diagram ?spaghetti diagram@ documents the distance an item travels during its production ; illustrates unnecessar" movement) Process map 0ground level1 documents the inputs and outputs of each step in the process ; provides the ra* material for !uilding a model of the process
The SIP4C Diagram ; The Forest #evel

) ) ) ) ) ) ) ) ) Start *ith the end in mind'ho are the customers of this process8 'hat are their re,uirements8 7o* are those re,uirements reflected in the process parameters ?output measures@8 'hat are the process outputs8 'hat are the process inputs that cause the outputs8 'hat controls are in place for the inputs8 'ho are the suppliers of the material for this process8 'hat are their re,uirements8
Sample SIP4C
Supplier8s9 Inputs1'e*:ts Process 3utput8s91'e*:ts Customer8s9
Grocery store Utility Company Appliance store
Coffee machine Measuring cup Electricity Qualified operator Water Filters Cream/mil !"eetener Ground coffee
+nstall filter Measure coffee Add coffee Add Water ,urn on machine
#$%% !pecified num&er of cups 'o grounds ( one hour old )ar *ot Aromatic Fresh !trong
*us&and Wife Guest
SIP4C $alidation
5evie* the SIP4C *ith "our2 team3 Champion and process o*ner?s@ to assure agreement on the SIP4C content as *ell as the project scope and success criteria among all sta+eholders-
S
Suppliers
Coal Field 5ihand Dam Suppliers Coal 'ater Air Po*er
I
Inputs
P
Process
4
4utputs
Po*er ffluent Process Ash Steam C4J
C
Customers
Internal =Plant Mgmt =Corp Mgmt = mplo"ee xternal =Cent-9ov=Ministr" =State 9ov=Shareholders =Contractors =Suppliers =PAPs =#a!our nvironment
CTQs
Plant Availa!ilit" (I arned Plant #oad factor 7eat 5ate
A(KI##IA5G P4' = 5 C4%S(MPTI4%

Preferred mplo"er 5&5 Man 2 M' 5atio #ead Time Manpo*er (tili6ation Qualit" S"stem Social 5esponsi!ilit" Maintenance Cost 4verhead xpenses
Manpo*er #u!ricants Maintenance Services Spares and Consuma!les
Boiler 1ur3ine
-./+ /I0 W/1%0 '1%/2
% 59G FFICI %CG

DM 'ater (sed Safet" aspect Training of manpo*er C"cle Time In process idle stoc+ Air mission ,ualit" %oise level
"enerator
"0I4
Collecting $oC
) 'ho is a customer8 ) 'hat does the customer need8
; 9athering the $oice of the Customer ; Define customer re,uirements
) 7o* do the customers prioriti6e their needs8 ) 7o* are customer needs translated into CTQs8
u9enerall" uMore
a less intense exercise for DMAIC projects than for DFSS projects li+el" to !e !ased on information that is alread" availa!le internall" for DMAIC projects than for DFSS projects
'ho Is a Customer8
Process $oice of Customer
Step
Step
Step
ProductC Service Customer
A customer setsCaffects re,uirements for "our product or service xternal Au"ing customers nd=users 5egulator" agencies A customer is one *ho receives "our outputInternal
Customer Segmentation
9enerall"3 external customer needs are more important than internal customer needsAre all customers e,uall" important8 xternal vs- internal Customer segments 5egions T"pe of !usiness $olumes Profita!ilit" Strategic mar+et Future potential
$oice of the Customer

) The $oice of the Customer ?$4C@ is the starting point of an" project and data collection plan) The $oice of the Customer includes2
; xpectations ; 5e,uirements ; 4pinions
Questions to Find $oice of the Customer

) 'hat are the elements of "our !usiness that are the most critical3 from the perspective of "our top or !est customers8 'hat are their relevant needs8 ) 'hat data has !een collected to understand the customer re,uirements8 ) 7o* do "ou operationall" define the defect from the perspective of the customer8 (nder *hat conditions does it occur8
$oice of the Customer Concerns

) 5eal vs- stated needs ) Perceived needs ) Intended vs- actual usage ) Internal customers vs- external customers ) ffectiveness vs- efficienc" needs
) Change over time
Determining Customer 5e,uirements

) (se ver!atim comments from customers to help determine the +e" customer re,uirements
; 'e often receive man" ver!al comments from our customers; 'e need to loo+ for *a"s to pro!e for deeper meaning !ehind the comments in order to translate these comments into *hat the customer actuall" re,uires-
Tools for 9athering $oice of the Customer
Unsolicited data %rom customers ; Complaints ; Field reports ; Trade journals ; Aenchmar+ing ; Internal research
) ) ) ) 5e,uirements documents Contracts Customer o!servation Ae a customer
Tools for 9athering $4C
Solicited data %rom customers ; Intervie*s ; Focus groups ; Surve"s ; Informal customer discussions ; Mar+et research
Determining Customer 5e,uirements

) 5evie* customer ver!atim comments and comparative data ) If possi!le3 pro!e for deeper understanding ) Convert into terms of process performance ) Descri!e the actual customer re,uirement
; 'rite the re,uirement3 not the solution ; (se measura!le terms ; Identif" performance targets ; Ae concise
Define Phase 5evie*

) The purpose of the define phase is to identif" and launch a project) $4C should !e reflected in the project charter especiall" in the project success criteria) The Alac+ Aelt and Champion should revie* and agree on the details of the project charter and SIP4C) A common cause of project failure is poorl" defined or projects *ith excessive scope- This revie* is an opportunit" to mitigate that ris+-
$4C ? $oice of Customer @
D M A I C
Arrest Pro.Deviation
,.W 6
Improve Processes Systems
0an*ing 7 1-5 8
Replace with eff. pro
-o#&etitor
Worse 'a#e Better
W,/1 6
-usto#er5eed 5eeds -usto#er

Aux Power Consumption Generation & Dispatch
Environmental Pollution Conservation of Coal
5 5 4 4
5 5 4 4 4 5 4 4 5 4 4 5 5 5
0/1I5"
81 90 64 66
Better
D M A I C
Project Charter
D M A I C
Resource Plan
Champion2 Process 4*ner2 Process 4*ner2 Qualit" #eader Coach ?AA@2 9reen Aelt2
General Information
Pro ect Review Dates

Tollgate Date
?xxCxxCxx@
S. Sinha &!D '. Agarwal EE!G
!r. P.#.!ohapatra !r. $.$.!ishra !r. $.#.Sinha !r. S. !athew !r. S. Baner ee !ohit "adav Iswar B!D Ashish %ain E!D
Signoff
Aig G2 Reduction in APC of Stage-I units
Define Measure Anal"6e Improve Control
9reen Aelt Contact Information 2
!ohit "adav ()*+,*-*(( .++/
ENCEJC EO ENCEMC EO DNCEMC EO BDCENC EO BDCEQC EO
Ges Ges Ges Ges Ges
Pro ect &eam Rhythm and Review

Meeting Fre,uenc"2 Mandator" Attendees2 4ptional Attendees2
.+ Days !ohit "adav MrP-<-Mohapatra $- Agar*al Mr- %-%-Mishra Is*ar S- Sinha A- .ain
5h"thm and 5evie*2 4n=Trac+ 4ff=Trac+ %eed Attention
Mr- %-<-Sinha
Mr- S- Aanerjee
$4C ? $oice of Customer @
The Seven QC Tools
Chec+ sheet Stratification Pareto diagram C & Diagram 7istogram Scatter diagram 9raphs and Charts
v Pro!lem Solving v Continuous Improvement== ;ai(en v Dispersion Control== Si< Sigma v 'aste limination===ean v QMS3 MS3 TS DPRMR3 3>SAS?Process control@ CAP v Supplier Development v Project Management==+eam -orking
What is their role B
In problem solving
+ool
'ole they Play
Data gathering
Chec+ sheet Stratification
Quantif" current status or magnitude of the Pro!lem Facilitate data gathering
Identif" and segregate different sources of the pro!lem
+ool
'ole they Play
) Pareto diagram
) Prioriti(e the problem
) Arain storming
) Cenerate many ideas %or solving a speci%ic problem ) Identi%y possible causes o% a problem in a structured -ay
) Cause & effect diagram
Tool
5ole the" Pla"
) 7istogram
) Scatter diagram
) Study pattern o% variation in a set o% data& ) Study relationship bet-een ! types o% variable ) Disual display o% data
) 9raph & chart
4/1/ "/1,%0I5"
Data Collection
"&at is .ata <

.ata is a numerical e0pression of an activit
-onclusions 3ased on facts and data are necessary for any i#&ro$e#ent9 =, #s&i(a7a If you are not a3le to e:&ress a &heno#enon in nu#3ers; you do not (no7 a3out it adequately *ord =elvin
1y&es of 4ata
Quantitative
/easura8le e9g9 <+ength; 1e#&erature -ounta8le 0 e9g9 <5u#3er of defects
Qualitative
1u8>ective assessment 0e9g9 <'core in a 3eauty contest
Po&ulation; 'a#&le and 4ata
3ction
Population
;andom 1amplin!
1ampl e 3ction
/easurement ? 98servation
.ata
A Sa"ing
q 'hen "ou see the data3 dou!t it q 'hen "ou see the measuring instrument3 dou!t itq 'hen "ou see the chemical anal"sis3 dou!t itq Three Categories
q
D False Data J Mista+en Data B %o Data availa!le
7o* to Collect Data8

qDefine the purpose (1ollow +2 .3 approach). qDefine the period for data collection. qDefine the Stratification. qDesign the chec4 sheet and assess !easurement System Capa5ility.
Purpos e
Be Clear on 2hat, 2here, 2hen, 2hy, 2ho and 3ow the data should 5e generated
S+'A+IEICA+I36
1tratification
-ommon
v/et&od of !roupin! data 8 points or c&aracteristics
v3 @iltration Process for isolatin! t&e cause of a pro8lem, vPrevent mi0 up and &elps in eas A faster identification-
%asis for 1tratification

"or(ers /ac&ines .efect
/aterial :ime Environment
Product @older
;e!ion @iles
Machine Machine A A 6 / #$F 6 / #$F D / 1! D / 1! / !&H PG /PG !&H
Machine , Machine , 6 #$F / #$F 6/ D 111I / 11I D/ PG/ / !J&! !J&! PG
C3M,I62D 6 / KFF D / 1"F PG / 1#&#
,lister De%ect UC= C= Pinhole De%ect UC= C = All De%ects UC = C=
C>2C;S>22+
-hec* sheet
/ con$enient and co#&act for#at for data collection infor#ation (ith #ini#u# (riting efforts and easy to fill
v / 'i#&le rule> 2a:i#u#

Check Sheet Eor Machining 3peration

=ocation
M/C No. Comp. DRG. No To en No. !p.
Croup
"t. Prod. Material Defect No. )n*p. # $ % & ' (
Date
M/c Defect* A + C D,,,P " R
Shi%t
Total Remar
Material De%ect
1 ! " # $ J ,lo- holes Cracks >ard Metal 2ccentric 3thers +otal A , C D 2 E Dia&. Dia&? Ch. Ch? CDD. CDD? C > I L ; =
Machining De%ect
=ength. =ength ? Sp. Sp? DA + Si(e. D A + Si(e? M 3blong 6 +aper 3 >ole Shi%ted P PCDD 0 Poor Einish ' 3thers
9raphs & Charts

9raphs represent data pictoriall"- A picture can see *hat DEEE *ords can not tellv #ine chart v Aar chart v Multiple !ar Chart v Component !ar Chart v 5adar chart
v
v
v Pie chart v 9antt chart v Pareto diagram v Scatter diagram v Control chart
v
,ar Chart
,ar graphs are parallel !ars of identical *idth !ut differing length to compare si6e of different ,uantities C things-
=ine Chart
=ine graphs manifest the overall trend in time series data !" direction of their lines-
Pie Chart
Pie charts ma+es it eas" to grasp the !rea+do*n of the components of a ,uantit" over a certain period-
-omparison of /ac&ines 3 A % for 7ee(l ;e>ection

5 0 15 11 10 5 0 1 ? 4 5 6 ! 8 9 10 Wee* 5u#3er 5 0 1 0 15 1? 10 11 8 6 ? 4 ? 10 15 1? 11
@ 0eAection
14
Multiple bar chart
Comparison o% Machines A A , %or Units Produced

900 800 !00 Bnits Produced 600 500 400 ?00 00 100 0 1 ? 4 5 6 ! 8 9 10 W ee* 5 u#3 er 6? 85 44 1 4?5 01 !5 ?!5 1 ?? ? 1 ?0! 94 ?48 ??6 5! 4 81 ?1! 99 ?58
Component bar chart
Pie -&art for -ustomer returned 7atc&es

" 5@
3 > Glass %ro(en % - 1top - - /vt, :rou8le . - .efective .ial E - ;e!ulation @ - 1tem *oose G - 9t&ers
1 @
% 4@ 4 6@
) ?@
/ 4?@
B !@
Control Chart
K= Aar and 5 Chart

K!arC5 Chart for CD
JQNM (C#FJQNB
Sample Mean
JQMR
MeanFJQMR
JQMM Su!group E N DE DN JE JN
#C#FJQMM
Sample 5ange
DN DE
(C#FDP-MD
5FO-OP N E #C#FE
'adar Chart on IS3 KFF1?1KK# Implementation

D JE DR DQ QE OE PE NE ME DO BE JE DE DP E P
Series D
J B M
DN
DM DB DJ DD DE R
9antt Chart for Construction Activit"

1y&e of (or* )oundation (or* )ra#e (or* 4ry-(alling %:terior touch u& 'heetroc* (or* Plu#3ing %lectrical (iring )it )i:tures Paint interior (all Interior touch u& Ins&ection deli$ery
1 ? 4 5 6 ! 8 9
Weeks
10 11 1
Cantt Charts makes it easy to understand the details o% a plan and progress in its implementation schedule&
Pareto Diagram
$ital fe* from Trivial man"

RM-P
50
DEE
100 90 80 !0 60 -u# Percentage
QO-N OP-Q
40 5o9 of co#&laints
PE-O MD-O
Pareto Anal"sis of Customer Complaints
?0
50 40 ?0
10
0 10
Fish not $egeta!le fresh *ilted
Aread Stale
Cashier 5ude
Meat not Fresh
ggs rotten
Pareto Princi&le
v80@ of &ro3le#s are caused 3y less than 0@ of &ro3a3le causes v%sta3lishes &roof of the need vIdentifies $ital fe(
PA5 T4 A%A#GSIS2 4utstanding !ranch *ise

BRA$C3 A= +./ B= *.+ C = ./ . / D = */ . / E = )+ . / 1=..+ G=../ 3=..+ ,9 ./ 7th .= / . + BRA$C3 E D C A B 1 3 &7&A8 G 7&3ERS Rs . D6E )+./ */./ ././ +./ *.+ ..+ ..+ ../ ,9 ./ /.+ RA&I7 &7 &7&A8 +..9 *-./ ...+ +.9 *.( ..9 ..9 ..* .// /.: ./ C6! . % +..9 9).9 ,:.* (..( ()., (:.+ (,.* ((.) .//./
Pareto /nalysis on .utstanding

RP-N RQ-J RR-M DEE
RD-R
50
RM-Q
100 90 80 !0
QP-J OM-O
40 .utstanding Calue
?0
ND-O
60 50 40 ?0
10
0 10
Aranches
-u# Percentage
Pareto 4iagra# for Production 'to&&age

DEE
/92D- quality change B9Inter#ediate con$eyor -9Po(er failure

5o9 of sto&&ages
MN ME BN BE JN JE DN DE N E / B - 4 % ) " , I E F + 25 . P Q 0 '
RE QE OE PE NE ME BE JE DE E -u#ulati$e @
49,o&&erDduct line Aa##ing %94ryer dru# cou&ling &in )9B P full &ress &ro3le# "94ryer &re$enti$e ,95i& roller I90otary co#3 tri&&ed E9-o#3er Aa##ing F9/l con$eyor idle roller +9)ire 29/ccu#ulation 594ru# seal changing .9)an tri&&ing P9-hain &ro3le#
Pareto /nalysis of -o#&laints at a +aundry

00 180 160 5o9 of co#&laints 140 1 0 100 80 60 40 0 0
RB QN ON PE
RQ
DEE
100 90 80 !0 60 50
BN
40 ?0 0 10 0
#ate Missing or Fading deliver" *rong colours items
Stains
Creased
Auttons Stretched Missing or torn
Brain 'tor#ing
generating large num!er of ideas !" a group of people
Basic 0ules
v4efer e$aluation v)antasiHe freely v"enerate quantity vBuild on ideas
4efer %$aluation
vPut critical faculties in cold storage - e$en constructi$e criticis#9 v v%nsure a &ro&er cli#ate for acce&tance of all sorts of ideas9 v v5o idea should 3e treated as stu&id9
)antasiHe )reely
v4onIt o&erate (ith your 3ra*es on9
v
vPartici&ants are encouraged to generate ideas; no #atter ho( fanciful they are9
"enerate Quantity v"enerate as #any ideas as &ossi3le9

v
v/ &earl di$er (ill 3e #ore successful in finding &earls; (hen he 3rings u& 00 oysters than (hen he surfaces only 150 oysters9
Build on ideas
#dea of one participant is more effectivel 8uilt up 8 anot&er participant,
'te&s in Brainstor#ing
v Select the topic v v ach mem!er3 in rotation gives ideas v v Mem!er offers onl" one idea per turn3 regardless of ho* man" he or she has v v Continue till all ideas are exhausted v v Ideas are recorded and displa"ed
Benefits
v Indi$idual is li#ited in generating ideas and grou& &roduces #ore ideas v v Ideas are i#&ro$ed u&on 3y #e#3ers v v Presence of others increases creati$ity v v Pooling of ideas and resources is #ade &ossi3le 3y co#ing together as a grou&
CAUS2 A6D 2EE2C+ DIAC'AM
q9raphic tool to represent relationship !et*een an effect and influencing causes qThere can not !e an effect *ithout a causeq5educe incidence of su!jective decision ma+ingqIdentif" main causes K/s influencing G
Construction of C &
Diagram
v Define pro!lem v 9ather mem!ers for discussion v Conduct Arainstorming v 9roup causes into MM/s v Man3 Material3 Machine3 Method v For each cause3 as+3 0'hat goes *rong that produces the effect1v Identif" major causes
-ause and %ffect 4iagra# for high &etrol consu#&tion

Procedure
I#&atience -raHe
4ri$er
,ea$y Body
Cehicle
'&ar* &lugs -ontacts +ife 1echnical details )uel #i: -ar3uretor
Poor antici&ation Wrong gears Wrong culture Poor s*ill
Bad attitude
/l(ays late +ac* of a(areness 0iding on clutch
'ha&e Ine:&erience ,igh ,9P
%ngine -ylinders
,igh Petrol
I#&urities Incorrect .ctane no9
0estrictions .ne (ay -ircuitous 0oad 5o turn
-rossings 1raffic
'&ares
'&urious
-onsu#&tio n
0oad
1yres Inferior )requent Petrol )aulty sto&s 5egligence &ressure '&eed Brea*ers /dditi$es Ignorance Potholes Irregular Incorrect $iscosity +o( &ressure ser$icing Poor -logged .il condition filters )alse 'tee& 5ot changed econo#y +o( le$el 2aintenanc
2aterials
-ause J %ffect 4iagra#
P54C SS
C44<I%9 TIM
MAT 5IA#S
$A5IATI4%
T5AI% D
QTG 4F '44D Q#TG C44<I%9 'AT 5 ?4#DCF5 S7@ T MPI% P %T4SA%S I% FI%A# P(#P I%ST5(M %T (%T5AI% D ACC(5ACG
P 5S4%% #
Q(IPM %TS
(ses of C &
v
Diagram
vTrace out real root cause v7elp evolve countermeasures vMa+ing C & an education in itself v ver"one participating3 learn more a!out their *or+vIs a focus for discussionvSho*s level of expertise availa!levCan !e used for an" pro!lem
v
v
?0 5 0 )requency 15 10 5 0 1 9! ! 6 1 98 6 8 1 99 6 90 5 91 4 4 9 ?6 9? 8 94 95 1 - o n s u # & ti o n 7 F W h 8 9 6 ? 19 1! 1 4
11
>IS+3C'AM
,istogra#
vMethod of anal"6ing data
v
v
vData is condensed in a ta!le

v
v
vTa!ulation is +no*n as fre,uenc" distri!utionv
vPresented !" a 9raph displa"ing distri!ution of data v
7istogram
"ra&h is -haracteriHed 3y ? constituents K centre 7 #ean8 K (idth 7s&read-$ariation8 K o$er all sha&e
v
)requency
?0 5 0 15 10 5 0 1 9! ! 6 1 98 6 8 1 99 6 90 5 91 4 4 9 ?6 9? 8 94 95 1 - o n s u # & ti o n 7 F W h 8 9 6 ? 19 1! 1 4
11
,istogra# -onstruction
v'elect a sa#&le of #in9 50 v0ecord the #easure#ents9 v4eter#ine the range9 v4ecide the no9 of classes9 v4i$ide range into no9 of classes v4eter#ine 3oundary or class li#its9 vPre&are frequency distri3ution9 v-onstruct histogra# 7"0/P,89
4ata on 2etal Bloc* thic*ness 7in ##8

?956 ?946 ?948 ?950 ?94 ?94? ?95 ?949 ?944 ?950 ?948 ?956 ?950 ?95 ?94! ?948 ?946 ?950 ?956 ?9?8 ?941 ?9?! ?94! ?949 ?945 ?944 ?950 ?949 ?946 ?946 ?955 ?95 ?944 ?950 ?945 ?944 ?948 ?946 ?95 ?946 ?948 ?948 ?9? ?940 ?95 ?9?4 ?946 ?94? ?9?0 ?946 ?959 ?96? ?959 ?94! ?9?8 ?95 ?945 ?948 ?9?1 ?946 ?940 ?954 ?946 ?951 ?948 ?950 ?968 ?960 ?946 ?95 ?948 ?950 ?956 ?950 ?95 ?946 ?948 ?946 ?95 ?956 ?95 ?948 ?946 ?945 ?946 ?954 ?954 ?948 ?949 ?941 ?941 ?945 ?9?4 ?944 ?94! ?94! ?941 ?9?8 ?954 ?94!
5angeF Max- ; Min-FB-PQ=B-BEFE-BQ

%FDE E
%o- of classesF R Class *idthF E-N
)requency 1a3le
-lass no,
,istogra# for 2etal Bloc* 1hic*ness

45 40 ?5 ?0 )requency 5 0 15 10 5 0 ?9? ?9?5 ?94 ?945 ?95 1hic*ness 7in ##8 ?955 ?96 ?965 ?9! ? ? 9 10 ? ?? ?!
,istogra# for Bearing 1hic*ness
45 40 ?5 ?0 )requency 5 0 15 10 5 0 59 4 59 8 59? 59?6 5 1 18 9
41
?1
1! 9 ?
594
5944
5948
595
5956
596
1hic*ness 7in ##8
,istogra# for %nergy -onsu#&tion
?0 5 0 )requency 15 10 5 0 19!!6 19868 1996 905 9144 -onsu#&tion7FWh8 9 ?6 9? 8 94 951 9 6 ? 19 1! 1 4
11
1y&es of ,istogra#s
Bell sha&ed
'y##etrical sha&e (ith a &ea* in #iddle re&resenting a nor#al histogra#
?0 5 0 )requency 15 10 5 0 1 9 ! ! 6 1 98 6 8 1 99 6 90 5 91 4 4 9 ?6 9? 8 94 95 1 - o n s u # & tio n7 F W h8 9 6 ? 19 1! 1
4 11
S+e*ed to #eft & 5ight
S+e*ed to #eft Caused !" centering the process to*ard high end of the tolerance S+e*ed to 5ight Caused !" centering the process to*ard lo* end of the tolerance--
Aimodal & Truncated
Aimodal 2 T*o com!ined populations== t*o shifts3 operators3 inspectors3 suppliers3 machine settings3 gages3 tools3 machines3 measurement locations3 etcTruncated2 This can happen *hen a process is not capa!le of meeting the specifications3 parts are sorted from !oth ends3 or too fe* classes are chosen-
Missing Centre Spi+eVs at Tail ?s@
Missing Centre 2 Centre of the distri!ution has !een sorted from the rest- Portion ma" have !een delivered to a customer *ith tighter specificationsSpi+eVs@ at the Tail ?s@ 2 Parts in outer ends of distri!ution are pro!a!l" !eing re*or+ed to !ring characteristic just *ithin specifications-
&he $eed for &ransformations--Prediction from a $D is possi5le S4ew distri5ution &ransformation

Before
KTrans F DCK0a(
After
,istogra# Bses
v1o *no(--(hether $ariation in data is due to chance or assigna3le causes9 v1o tell a3out Process Beha$ior v --a3out its ca&a3ility to &roduce defect free out&ut v
v
Sources 4f $ariation
vCommon Causes ===Chance Causes 4f $ariation vSpecial Causes === Assigna!le Cause 4f $ariation
Common Cause
vConsists of combined effect of several sources of uncontrollable variation inherent to a process. vCollective influence of common cause variation defines natural process fluctuation and is known as Chance causes of variation. vProcess output is predictable vProcess is said to be in Statistical Control
v
v
Special Cause
vVariation has a large impact on performance. vDetermination of source of impact makes cause "assignable." and is termed as assignable cause of variation. vIf they exist, process or key characteristic is said to be "out-of-control". vOut-of-control process is not predictable
Process Beha$ior
vIt tells (hether &rocess is under control6 vIs it &roducing defect free out&ut6
v
#S #
?0 5 0 )requency 15 10 5 0 1 9! ! 6 1 98 6 8 1 99 6 90 5 91 4 4 9 ?6 9? 8 94 95 1 - o n s u # & ti o n 7 F W h 8 9 6 ? 19 1! 1 4
====process under control or
Process out 4f control

(S # #S# (S#
11
process varia!ilit"
process varia!ilit"
"ood Process Beha$ior

v'ha&e close to nor#al cur$e v2ean at target $alue v'&read (ithin '&ecification li#its
v
v v
v
v-& is greater than 196! and -&* is greater than 19?? v
SCA++2' DIAC'AM
'hat2 To stud" the possi!le relationship

!et*een t*o varia!les-
'h"2
Diagram ma+e it clear *hether a
relationship exists3 and sho*s the strength of relationship-
'hen2To test a theor" that the J

varia!les are related-
%:a#&les<
v-utting s&eed and tool life vBrea*do(n and equi&#ent age v1e#&erature J li&stic* hardness v1e#&erature and &ercent foa# in soft drin*s v,ardness and tensile strength
4ifferent 'catter diagra# Patterns
'catter 4iagra# on -on$eyor '&eed $s9 'e$ered +ength
1050 1045
'e$ered +ength 7##8
1040 10?5 10?0 10 5 10 0 1015 1010 1005 1000 5 595 6 695 ! !95 8 895 9
-on$eyor '&eed 7c#Dsec8
Bses<
v-ontrol Purpose v;eplacin! a destructive test 8 a non-destructive test v1tud of -ause A Effect relations&ip vProcess 9ptimi5ation
v
Process Mapping
A!out This ModuleI

Process Mapping is a tool used to2
NClearl" define processes N NIdentif" areas *here data
collection should ta+e
place
N N$isuali6e
activities involved in a process at the earl" stages of project development
Si! Sigma" # $uest for Process Perfection #ttac% &ariation and Meet 'oals
(Data)ile(ProcessT*ppt
'hat 'e 'ill #earn I

D-The importance of process maps and the character of the product and process parametersJ-The *hen3 *h" and *here to use process mapsBM-The x/s & "/s and K/s & G/s and GFf?x3K@NP-'hat to measure and controlOQ-The need for process maps prior to FM A/s3 9age Studies3 D4 /s and SPCRDE-'hen the process map is completedDDDJ-7o* to use the tool for "our process-
xtending xtending Flo* Flo*Charts Chartsto to Process ProcessMapping Mapping
Process Management
Aasic Process Model
1 stem
;esources 4 4 4 4 4 People E'uipment /aterial /one :ime
#nput #nput
Process Process
-
9utput 9utput
P
)eed3ac*
C"cle Time
) Time it ta+es to complete a process from !eginning to end
Question
07o* does a reduction in c"cle time !enefit an organi6ation81
Fundamentals of Process Mapping

A Aprocess processmap mapshould shoulddescri!e2 descri!e2 N Major N MajoractivitiesCtas+s activitiesCtas+s N Su! N Su!processes processes N Process N Process!oundaries !oundaries N Inputs N Inputs N 4utputs N 4utputs N Process N Process& &Product ProductParameters Parameters N Customers N Customers& &Suppliers Suppliers N Process N Processo*ners o*ners A Aprocess processmap mapshould should!e !e revie*ed revie*edfre,uentl" fre,uentl"and andis is never neverdonedoneA Aprocess processmap mapshould should document documentho* ho*the theprocess process actuall" actuall"operates3 operates3notho* notho*it it is issupposed supposedto tooperateoperate?0As ?0Asis31 is31not not0To 0ToAe1@ Ae1@ A Aprocess processmap map*ill *illidentif" identif" opportunities opportunitiesfor for,ualit" ,ualit" improvementsimprovements-
Principles of Process Management

) sta!lish o*nership) $erif" and descri!e the purpose of the process) Define the process3 !oundaries3 and interfaces) 4rgani6e and train the process improvement team-
Principles of Process Management

) Define and document the process) Define points of control) sta!lish process measurements) Improve process-
Process Mapping Steps

D-ID %TIFG I%P(TS A%D 4(TP(TS Identif" Inputs ?ra* material3 e,uipment3 energ"3 PM/s3 etc-@ Identif" 4utputs ?measura!leCassessa!le end product parameters@ B-S74' A## ST PS $alue adding steps have the follo*ing characteristics2
uSomething uTransforms uDone
the customer *ould !e *illing to pa" for the product or service
right the first time
%on value=added steps in the process are presented graphicall"2

u
valuation points points
u5e*or+ uScrap
points
uInventor"
Steps ?cont-@
B-S74' 4(TP(TS 4F AC7 ST P Sho* after each process step the characteristics that can impact the follo*ing step?s@u
M-S74' A## P54C SS PA5AM T 5S AT AC7 ST P #ist under each step the parameters that can change a product characteristic at that step ?i-e-3 parameters that can !e controlled at that step@u
More Steps
N-C#ASSIFG T7 PA5AM T 5S Classif" the process parameters identified ?in WM a!ove@ into the follo*ing categories2
u
% F %oise Factors = (ncontrolla!le = Ma" !e controlla!le3 !ut are not controlled !" decisionC F Controlla!le factors = Process factors that can !e changed to see the effect on product characteristicsS F Standard 4perating Procedures = A procedure is used to define and run those factors- Tooling3 Fixtures C5 F Critical Factors = Determined through FM A3 D4 3 etc-
Inputs & 4utputs

7igh #evel Process Map K/s I%P(T S P54C SS G/s 4(TP (TS
Processes Come in 7ierarchies

Process = #evel WD
Step WD Step WJ Step WB
Process = #evel WJ
Process = #evel WB
Select Selectthe theappropriate appropriateprocess processlevellevel-
Product and Process Parameters

ST P 4F P54C SS
Process Parameters3 x/s
Inputs
4utputs
" F f?x@
5emem!er the P M/s

NMan
?People@ ? ,uipment@
< G for ?x/s@ Process Parameters % C %oise Parameters Controlla!le Process Parameters S S4P Parameters C5 Critical Parameters
NMachine NMethod NMaterial
?Procedures@
Product Parameters3 "/s
NMeasurement NMother
%ature ? nvironment@
'h" #ist the Parameters8

T4 5 D(C D F CTSX
NThe N$ariation NThe NThe NThe
Defects
x/s and K/s are the sources of variation in "our processcauses defectsx/s must !e under control to prevent defectsroot cause of a defect is variation of the x/sX
"/s and G/s are the measured results of the process and include the failure modes of the processare also outputs of a process step-
NDefects
'hat Are 'e Measuring8

Measure the x/s3 not the G/s X
K/s x/s "/s G/s
Inputs YYYYYYYYY 8 Process Parameters YYYYYYYYY 8
YYYYYYYYY 8 YYYYYYYYY 8
Process 4utputs
Process Step 4utputs
'e 'ecannot cannot control control*hat *hat *e *edon/t don/t measureX measureX
Is 'or+manship an x8
K/s I%P(T S P54C SS
Process Parameters3 x/s '45<MA%S7IP 88 4P 5AT45 88
P M/s reminders2
NMan
G/s 4(TP (TS
" F f?x@
?People@ ? ,uipment@
Product Parameters3 "/s

ND ND
?Procedures@
F CT F5
4(TP(T
F CTS I% P54C SS 4(TP(T
'hat Are the x/s and "/s8

K/s Inputs
P M/s reminders2
NMan
G/s 4utputs
Process Parameters3 x/s % % % % C C C C S S Product Parameters3 "/s
?People@ ? ,uipment@ ?Procedures@
" F f?x@
Identif"ing Product & Process Parameters

The follo*ing ma" !e helpful to identif" process parameters that have a potential effect on the product parameters and process output2
NArainstorming N#iterature N4perators N'or+
5emem!er the P M/s
NMan
revie* manuals
Instructions experience C supplier input
N4perator
NCustomer N
ngineering +no*ledge theor"
NScientific
Completeness Chec+s
Inputs Step of the Process
Process Parameters3 x/s Are there "/s for ever" x in this step8 Is there a 0good1 t"pe of " for ever" x 8 Is there a 0!ad1 t"pe of " for ever" x 8 Are x/s here that impact do*nstream "/s8 Does the map have input of extended team8
N9ood NAad
4utputs
" F f?x@
Product Parameters 3 "/s
5emem!er the P M/s

NMan
F Defects N9ood 5ejected F Defect NAad Accepted F Defect N7orror Stories2 'hat has happened in the past that caused disasters8 NSuccess Stories2 'hat outputs of this step thrilled the customer?s@8 NAre there x/s for each " in this step8 NAre upstream x/s changing "/s of this step8 7o*8
'ave Solder Process

I%P(T K/s Aoards *ith components Solder Flux lectricit" Machine Setup #4AD I% FIKT(5 KMas+ing 4perator Aoard thic+ness M=Corner support Ad- Spacing in Fixture Ad ,uantit" in fixture Ad- position *ithin fixture Front or rear Side FixturesCconve"or *idth Ad 4rientation Frame si6eCAoard si6e Fixture massFixture dimensions $ert- location !d-in fixComponent 4rientation Component Densit" #4AD 4% C4%$ G45 KConve"or speed 5ail position over pot Conve"or angle Time re,/d in Solder M=Corner Fixture support Conve"or drive smoothness Finger condition 5ail Straightness F#(K Pressure Flux Arand Flux T"pe Thinner Cleanliness of flux Tritration level Stone T"pe Cleanliness of Stone 7eight over Stone Temperature 7umidit" Am!ient Temp F#(K AI5 <%IF KPressure of air +nife 4rientation Distance to !oard Air <nife used Angle of air +nife Temperature T4 P5 7 AT
Current Setup2 Specifications

Flux2 <ester JS Solid EAC+3' S-9-2 %A 'A6C2 63MI6A=
Ad 4rientation Ad Spacing Ad Alignment Ad to 5ail Position Critical areas mas+ed
Immersion depth 5aised Carrier Frame corner Poor *etting Partial filled holes S+ip Soldering Aridging xcessive solder speed Insufficient solder Aridging C"cle Time
#ead Soldera!ilit" Amount on AdFlux through holes to top side Distri!ution on Ad< G for ?x/s@ xcessive Solder Process Parameters Aridging %oise Parameters Icicles Controlla!le Process Parameters S4P Parameters Partiall" filled 7oles qCritical Parameters Cleanliness of Aoard
Solder in all holes Solder Aridging Solder Insufficients Fire Qt" of xcess Flux
Pre 7eat WD Temperature = MEE
Pre 7eat WJ Pre 7eat WB Solder Temperature Temperature Temperature = = = MMN MOE MRE
mmersion Conve"er Depth Speed = J-ON=M-JN DCMZ B-ON
Angle = %CA
7ot Air <nife Angle MN=PN PE
Pressure DJ=JE DOPSI
'ave Solder Process ?cont-@

P5 7 A T WD P5 7 A T WJ P5 7 A T WB KTemp = >one B Temp = >one D Temp = >one J Time in Preheat B 5esponse time of heater Sta!ilit" of temp Distance to !oard Temp- Distri!ution Fixture *arp 5ail *arp S4#D 5 P4T KSolder Temp 7t of PotCPosition Solder Pot Angle xit Point Amount of dross Solder Pump Speed Solder 7eight Solder Pump Pressure Cho+e !ar settingCadj Solder contact Solder t"pe Solder tin content Solder Contam Aaffle ,ual3 hole *ear3 !ent condition3 clean Aoard Deflection 74T AI5 <%IF Position Dist to !oard Temp Angle Air Pressure Temp Dist KIT 'A$ S4#D 5
'or+manship Stds Samples #ighting Magnification ProdCDmndCSch CensusCInspection Staffing Inspection Sample ?S@ "esight Inspector $ariation
I % S P C T I 5 4 % ' 4 5 <
4(TP (T
G/s Aoards2 Accepted 5ejected Scrap
KTemp = >one J KTemp =>one D Temp = >one D Temp = >one J Temp =>one B Temp = >one B Time in Preheat J Time in Preheat D 5esp time of heater 5esp time of heater Sta!ilit" of temp Sta!ilit" of temp Dist to !oard Distance to !oard Temp distri!ution Temp distri!ution Fixture *arp Fixture *arp 5ail *arp 5ail *arp
Solder Distri!ution Coverage on !oard Solder Appearance Solder joint Qualit"
Aoard Temp Flux Condition Flux Activated Flux Solvent drive off Thermal Shoc+ Aoard 'arping xcess 7eat xcess Solder Poor Fillets xcess flo* thru Inade,uate 7eat Solder splatter Trapped 9as Solder Aridges Poor flo* thru ?plated thru holes@
"/s Product Paramenters Solder Appearance Solder .oint Qualit"CDefects 4pen Solder .oints 9ood Aoards Dirt" Aoards Scrap Aoards 7ot Aoards Damaged Components #ifted Components
< G for ?x/s@ Process Parameters %oise Parameters Controlla!le Process Parameters S4P Parameters qCritical Parameters
xcess Pressure Solder Aridging Insufficient solderC4pens Insufficient Pressure Solder Aridging Temp Too #o* Solder Aridging Temp Too 7igh Insufficient SolderC4pens Damage Aoard Angle Too Steep Insufficient SolderC4pens Angle Too Shallo* Solder Aridging
The Importance of Questions

#oise 'arameters(
N'hat
are the"8 NAre the" impossi!le or impractical to control8 N7o* ro!ust is the s"stem to the noise8
N
Controllable 'arameters(
N7o*
are the" monitored8 N7o* often are the" verified8 NAre optimum target values +no*n8 N7o* much variation is there around the target values8 N7o* consistent are the"8
N
Standard O"erating 'rocedures(

NDo
the" exist8 NAre the" understood8 NAre the" !eing follo*ed8 NAre the" current8 NIs operator certification performed8 NIs there an audit schedule8
Process Parameter Questions

'rocess 'arameters(
N'hat
causes variation of the process parameter8 N7o* is the process parameter controlled8 N7o* often is the parameter out of control8 NIs there data on the parameter8 N'hich of "our process parameters should have control charts on them8 N'hen should "ou place a control chart on a process parameter8 N'hich of "our process parameters have control charts on them8 N7o* are the control charts used8 N7o* do "ou +no* *hich process parameters to monitor8 NShould *e focus on parameters of non value=added steps8
Product Parameter Questions

'roduct 'arameters(
u'hat uIs
is the goal of the improvement effort8 An attri!ute or a varia!le8
the product parameter ,ualitative or ,uantitative8 the product parameters2 Is larger !etter8 Is nominal !est8 Is smaller !etter8 Is it d"namic in nature8
uFor
uIs
the concern for--Process centering8 Process variation8 Aoth8
u'hat
is the process !aseline for the product parameter8 'hat is the mean and sigma8
More Product Parameter Questions

'roduct 'arameters(
NIs
the product parameter currentl" in statistical control8 NIs the product parameter affected !" time8 N7o* much of a change in the product parameter do "ou needC*ish to detect8 NDo "ou +no* the expected distri!ution of the product parameter8 NIs the measurement s"stem ade,uate8 NAre there multiple responses of concern8 'hat are the priorities for optimi6ation8 N'hat measurements are ta+en on product parameters8 N7o* do "ou +no* *hich product parameters to monitor8 N'hich product parameters need control charts on them8 N'hich product parameters have control charts on them8 N7o* are the control charts used8
Process Map=Flo* Charting ?Step=!"=Step@

Identif" the inputs and outputs of the process Document entire flo* of the process selected Identif" all value and %on value=added operations Identif"Cclassif" the scope of the process
Identif"Cclassif" measurements ta+en on product & process parameters
Continue to update & classif" process mapX
Classif"C characteri6e process parameters into B main factors
Develop initial list of process parameters along *ith current operating conditions
Identif"Cclassif" upstream in=process product parameters
r%oise factors rStandard operating procedures rControlla!le process parameters
Consider ver" Process Step

2* What 'rocess
NAll
*or+ is a process NAll processes have o*ners NAll processes can !e descri!ed as a ver! and noun NAll processes can !e anal"6ed and improved
feedback
3* What Customer
%e+uirements
NSpecifications
0* What Out"ut 1* What n"ut /* Who are Customers 4* Who are Su""liers
feedback
NProcess
Function 5elia!ilit" Format

NCost NSchedule
,* 'rocess Controls-.e"endencies
ProceduresCPolicies An" *ritten document that controlsCimpacts a process TrainingC ducation
NPerformance
o*ner/s re,uirements on the supplier NSpecifications NCost NSchedule ,uipmentCFacilities

NSpace
)* What Su""lier %e+uirements
Qualit" Attitudes Personal attitude *hich is less than the re,uirement
s+ills NCertifications
re,uired
NProcessing
e,uipment
Causes of Process Map Failure

C4%C 5%
5e,uires extra effort = 0'e +no* the process = lets just move for*ard1
5 SP4%S
Initial pa"off is team understands process == team mem!ers are not *or+ing on different set of assumptions- (se experts to help "ou through the mapping process If necessar"3 adjust process !oundaries == initiate another improvement team Thin+ a!out approaching the pro!lem hierarchicall"
The process appears straight for*ard = then !ecomes difficult as "ou reali6e "ou do not understand process as *ell as "ou thought The process map just seems to gro* and gro* and gro*
Process Mapping Summar"

'hat 'hatis isthe thetool8 tool8 N 9raphical N 9raphicalmethod methodto toillustrate illustrate the details of a process the details of a process
N N
'hat 'hat*ill *illthe thetool toolidentif"Csho*8 identif"Csho*8 N All N Allprocess processsteps3 steps3value=added value=added & non value=added & non value=added Input @ N i3in Inputparameters parameters?K ?K N i3in @ N i@ @ ndproduct productparameters parameters?G ?G N nd
N In=process N In=processparameters parameters?x/s ?x/s& & i
'hen 'hendo do"ou "ouappl" appl"this thistool8 tool8 N Al*a"s2 N Al*a"s2to tofull" full"understand understand process & process process & processflo* flo* N Find N Find*hereC*henCho* *hereC*henCho*defects defects are !eing created are !eing created N Define elements of c"cle time N Define elements of c"cle time N
N
"/s@ "/s@ N Characteri6ation N Characteri6ationof ofall all parameters parameters N DefectCdata N DefectCdatacollection collectionpoints points N Steps N Stepsneeding needingFM FM A/s A/s N Sources N Sourcesof ofvariation variationidentified identified
'hat 'hatresults resultscan can"ou "ouexpect8 expect8 N S"stems N S"stemsneeding needingMS MS /s /s N #ist N #istof ofFactors Factorsfor forD4 D4 /s /s N Find N Findthe thehidden hiddenfactor" factor" N 4pportunities N 4pportunitiesfor forprocess process step elimination ?i-estep elimination ?i-e-flo* flo* improvement@ improvement@ N 'a"s N 'a"sto tore=la"out re=la"outthe theprocess process N Sources N Sourcesof ofvariation variationreduced reduced
Map Gour Process

Eor the ne<t session ) Map and characteri6e a critical part of "our process) Identif"2 ; The Inputs ?K/s@ ; The 4utputs ?G/s@ ; The Process Parameters ?x/s@ ; The Product Parameters ?"/s@ ; The process o*ner3 supplier3 & customer ; Classif" the parameters at each step ; The next step to reduce defects in "our process
'hat 'e 7ave #earned I

D-The importance of process maps and the character of the product and process parametersJ-The *hen3 *h" and *here to use process mapsBM-The x/s & "/s and K/s & G/s and GFf?x3K@NP-'hat to measure and controlOQ-The need for process maps prior to FM A/s3 9age Studies3 D4 /s and SPCRDE-'hen the process map is completedDDDJ-7o* to use the tool for "our process-
xtending xtending Flo* Flo*Charts Chartsto to Process ProcessMapping Mapping
Failure Modes and ffects Anal"sis FM A
A!out This ModuleI

Failure Modes and ffects Anal"sis
An FM A is a s"stematic method for identif"ing3 anal"6ing3 prioriti6ing and documenting potential failure modes3 their effects on s"stem3 product3 process performance and the possi!le causes of failureSi! Sigma" # $uest for Process Perfection Meet 'oals and #ttac% &ariation
(Data)ile()M+#form*!ls (Data)ile(Copy)M+#*!ls ,Datafile,causeeffecte*ig! (Datafile(catapultflo-*ig! (Datafile(catapultC.+*ig!
'hat 'e 'ill #earnI

Failure Modes and ffects Anal"sis D- As a Team3 ho* to construct an FM A and associated Action Plan J- 7o* the FM A process ties to process mapping B- The relationship !et*een Failure Mode3 Cause and ffect M- The different t"pes of FM As
Sample FM A
S Eailure 2 2%%ects D Causes Must redo cop" P Paper .am Must redo cop" Must redo cop" 3 D ' C 2 P C Controls + 6 Action 'ecommended 'esp& Schedule Person Date Periodic Maint- O JRM Periodic preventive maintence Place sign over copier outlining xisting standard si6e enlargeCreduce or (ser notes on relia!le mach to clearl" indicate P misset si6e P copier N DQE standard reduceCenlarge (ser xisting misset notes on Place sign to encourage user to P control N copier M DJE utili6e auto settings (sed landscape instead of portrait or Tra" Place note on ruler re tra" P vice versa O Selection J QM selection O align mar+ing not clear Doc moved *hen lid closed (se Auto Feeder C M align ruler B OJ Action +aken PM Schedule created and implemented Actual Compl& p p p Date S O D p r p n ' i 'isk s M k prpn
<e" 4pr
BCD
JCDN
P B O DJP B BOQ
<e" 4pr
JCJE
Place sign over mach Place sign over mach
JCDN
P B J
BP D
BP
<e" 4pr
JCJE
JCDN
P J J
JM D
JM
Must redo cop" Must redo cop" Must redo cop" Must redo cop" Must redo cop"
<e" 4pr
DCJE
Placed %ote
DCDN
P J D
DJ D
DJ
nlarge mar+s for Q-N Z paper on ruler <e" 4pr
DCDN
nlarged mar+s
DCDM
P J D
DJ D
DJ
(se Auto Place sign over copier re Feeder C Z nsure align prior to cop"ing or N align ruler J PE use auto FeederZ <e" 4pr Place sign over copier to encourage user to use auto tra" select Place cleaning material near copier
DCDN
Displa"ed Sign
DCDM
P D D
Auto select P function B NM Periodic Cleaning P Dirt" 9lass P S4P D BP
(ser selected *rong tra" B
<e" 4pr
JCJN
Place sign over mach Placed Cleaning Matl
JCJE
P J B
BP D
BP
Maint-
DCDN
DCDN
P D D
.atafile-Co"y56EA*7ls
'h" (se FM As8

'hat is an FM A8 Identif" critical product characteristics and process varia!les Prioriti6e product and process deficiencies in support of do*nstream improvement actions 7elp focus on prevention of product and process pro!lems
Aenefits of FM A/s
'hat is an FM A8
% u 7elps increase customer satisfactionM u 5educes product development timing and cost& u 5educes the amount of re*or+3 repair and scrapu Documents and trac+s actions ta+en'
u u
Improves the ,ualit"3 relia!ilit" and safet" of products-
Prioriti6es deficiencies to focus improvement efforts-
Process and FM A 7ierarchies

'hat is an FM A8
Process = #evel WD
FM A = #evel WD
Process = #evel WJ
FM A = #evel WJ
Process = #evel WB
FM A = #evel WB
Process FM A Steps
'hat is an FM A8
uSteps
Completed Prior to FM A2
uFM
A Steps2
DJ-
Charter Team Develop and Characteri6e Process Map D-Identif" 07eav" 7itter1 Process Step D-Identif" Associated "/s ?Product Parameters@ D-Identif" Failure Mode D-Identif" Failure ffectsC5ate Severit" D-Identif" CausesC5ate 4ccurrence D-Identif" Controls ?if an"@C5ate Detection D-Calculate 5P% D-Prioriti6e !" 5P% 4rder D-Determine ActionsCPlan D-5ecalculate 5P% Aased on Plan D-Ta+e Action
BMNPOQRDE-
FM A Form
'hat is an FM A8
7eader Accessi!le from $ie* 7eaderCFooter in xcel
'or+!oo+ in xcel
(Data)ile()M+#)orm*!ls
Cause =Failure Mode = ffect Continuum

'hat is an FM A8
ffect ?"/s@ Cause ?x/s@ Failure Mode
The Cause and ffect Diagram xample

'hat is an FM A8
AdminCService xample
Measurements Materials Manpo*er
First produced in DRNE !" Professor <aoru Ishi+a*a = Also called the2 uIshi+a*a Diagram uFish Aone Diagram Developed to represent the relationship !et*een some 0effect and all possi!le 0causes1 influencing itCreate using Igrafx2
Failure Mode ?Defect@
Mother %ature
Methods
Machines
Failur e ffec t
The Cause and ffect Diagram xample

Cause and ffects Diagram
Measurement Manpo*er
Inade,uate training #ac+ of experience Distractions
Material
#ate 'rong ,uantit" Defective
5eproduci!ilit"
5epeata!ilit" #inearit" Sta!ilit" Cali!ration
Defects
Too hot Too humid %ot maintained Too cold Inade,uate capa!ilit" $ague 4ut of date Complex
Mother nature Machines
Methods
.atafile-Causeeffecte*ig7
Cop" Machine xample

'hat is an FM A8
) 4ur process is cop"ing documents on a Kerox model KCDEMN cop" machine) First *e *ill construct a process map ) Then *e *ill construct a cause and effect diagram ) Finall" *e *ill complete an FM A
Process2 Ma+ing A Cop"

'hat is an FM A8
Place Document in Copier

% 7inges % 9lass clean
Set num!er of copies

C Copies re,uired Cr %um!er !utton
nter si6e re,uired

C Si6e desired Cr Si6e !utton
Set lightCdar+ settings

C Dar+ness desired
Select paper source

C Si6e desired x Paper
Press !utton
5etrieve copies
Cr Autton
Ma+e Copies
Document set correctl" 9lass clean
%um!er of copies selected correctl"
Si6e selected correctl"
Dar+ness set directl"
Correct paper tra" selected
Copies
Copies 5ight num!er 5ight contrast 5ight orientation 5ight si6e 5ight paper
#egend C Controlla!le Cr Critical % %oise P Procedure x Input
Step D2 Identif" 07eav" 7itter1 Process Step

The FM A Process
uFrom
the Process Map3 identif" the process step *ith the most li+elihood of having failure modes *ith significant effects
defect data andCor team +no*ledge a!out failure % modes *hen selecting process steps impact to the !usiness8 ?C4PQ3 c"cle time3 fill M uSignificant rate3 ---@ & u(se a Cause and ffect Diagram to capture !rainstorming results' uAfter completing FM A Steps WJ=O for all failure modes
u(se
associated *ith this process step3 return to this step and select the next most li+el" 07eav" 7itter1 process step all process steps *ill need to !e anal"6ed !" the FM A
u%ot
Step J2 Identif" Associated "/s

The FM A Process
uFrom uAs
the Process Map3 identif" the "/s that are associated *ith the process step !eing investigated the "/s are the indications of a successful completion of the process step3 the" are crucial as a !asis for determining failure modes
Step B2 Identif" Failure Mode

The FM A Process ) Arainstorm failure modes for the selected process step 2 ; Identif" the *a"s in *hich the process could fail to generate each of the expected 0"/s1 ) liminate 0duplicates1 from !rainstorm list ) Are the failure modes from the same level of the process8 ) Are the failure modes specific8 ) Are the failure modes the most li+el"8 ) Do the failure modes provide good coverage of the process step8 ) 7ave all "/s !een considered8
Step M2 Identif" Failure ffectC5ate Severit"

The FM A Process
uPic+
the most li+el" failure mode and !rainstorm the most important ffects2 FAI#(5 FF CTS are the outcome of the occurrence of the failure modeon the processThe impact on the customer === 'hat does the customer experience as a result of the Failure Mode8
uIdentif"
each effect as !eing 0Attri!ute1 or 0$aria!le1 doesn/t change unless the design changes-
uSeverit"
Step N2 Identif" CausesC5ate 4ccurrence

The FM A Process uIdentif" the most li+el" causes for each failure mode using a Cause and ffect Diagram2 CA(S S are the conditions that !ring a!out the Failure Mode
uTransfer uAssign
the resulting information to the FM A form
an occurrence value ?D=DE@ to the li+elihood that each particular cause *ill happen and result in the failure mode occurrence score for each cause should !e related to the li+elihood of that cause resulting in the failure mode and producing the specific associated effect
uThe
4rgani6e Arainstorming Ideas

The FM A Process
Measurement
'rong Si6e Selected
Materials
'rong Paper Si6e
Manpo-er
Selected *rong orientation
Cop" Misaligned
Too 7umid Document Moved 'hen #id *as Closed Alignment Mar+ing (nclear
Mother 6ature
Method
Machine
'hat *ould "ou add8
Step P2 Identif" ControlsC5ate Detection

The FM A Process
uIdentif"
the current mechanisms in place *hich prevent the causefrom occurring3 or detect it !efore the product reaches the customer- Some examples of controls are SPC3 training3 maintenance3 inspection3 S4P etca detection value ?D=DE@ !ased on an assessment of the li+elihood that the current control mechanisms *ill detect the cause of the failure mode !efore it reaches the customeragoni6e over detecta!ilit"-
uAssign
uDon/t
Step O2 Calculate 5is+ Priorit" %um!er

The FM A Process The product of the estimates of severit" occurrence and detectionThe 5P% provides a relative priorit" for ta+ing action the !igger the 5P%3 the more important to address-
5P% F S $ 5ITG x 4CC(55 %C x D T CTI4%
Steps Q and R
The FM A Process Q2 Prioriti6e !" 5P% 4rder (se the 0Sort1 command in xcel to order the spreadsheet in descending order of 5is+ Priorit" %um!er ?5P%@R2 Determine ActionsCPlan Aased on the causes found3 determine actions that *ill minimi6e the effect of each cause3 in priorit" order-
Steps DE and DD
The FM A Process
DE2 5ecalculate 5P% Aased on Plan uAssuming the actions are carried out successfull"3 reassign severit"3 occurrence and detecta!ilit"u
uPlace
u
these ne* ratings in the 0predicted1 columns ?ps3 po & pd@-
uAssign
a rating from D to N for each action that *ill sho* the 0ris+1 associated *ith each action ?N !eing the greatest ris+@- Place the rating in the 0ris+1 column-
DD2 Ta+e Action uAased on the ris+ mitigation column ?5is+ [ prpn@3 ta+e the actions indicated or reassign actions- ThenIu
uComplete
the actions indicated !" the times statedX
The FM A Process
Steps D=DD2
Step D ID Process Steps
S Eailure 2 2%%ects D Causes Must redo cop" P Paper .am Must redo cop" Must redo cop"
Step J ID "/s Step M

3 D ' C 2 P C Controls + 6 Periodic Maint- O JRM
Step P
Step R
(Data)ile(Copy)M+#*!ls
Step DD
Schedule Date Action +aken PM Schedule created and implemented Actual Compl& p p p Date S O D p r p n ' i 'isk s M k prpn
FM A
Action 'ecommended
'esp& Person
Must redo cop" Must redo cop"
Periodic preventive maintence Place sign over copier outlining xisting standard si6e enlargeCreduce or (ser notes on relia!le mach to clearl" indicate P misset si6e P copier N DQE standard reduceCenlarge (ser xisting misset notes on Place sign to encourage user to P control N copier M DJE utili6e auto settings (sed landscape instead of portrait or Tra" Place note on ruler re tra" P vice versa O Selection J QM selection (se Auto Feeder C align ruler B OJ
<e" 4pr
BCD
JCDN
P B O DJP B BOQ
<e" 4pr
JCJE
Place sign over mach Place sign over mach
JCDN
P B J
BP D
BP
<e" 4pr
JCJE
JCDN
P J J
JM D
JM
<e" 4pr
DCJE
Placed %ote
DCDN
P J D
DJ D
DJ
align mar+ing P not clear M Doc Must moved Step redo B *hen lid cop" P closed ID Failure Modes N
nlarge mar+s for Q-N Z paper on ruler <e" 4pr
DCDN
nlarged mar+s
DCDM
P J D
DJ D
DJ
(se Auto Place sign over copier re Step Feeder C Z nsureStep align prior O to &cop"ing or align use auto FeederZ N ruler J PE
Step DE
<e" 4pr DCDN Displa"ed Sign DCDM P D D P D P
Process FM A
T"pes of FM A ) 7elps anal"6e manufacturing and assem!l" processes to reduce the occurrence and % improve detection of defects) Assists in the development of process control M plans& ) sta!lishes a priorit" for improvement activities) Documents the rationale !ehind process ' changes and helps guide future process improvement plans) IS P54ACTI$ X Should !e started *hen ne* processes are designed or *hen old processes are changed-
Process FM A Scoring Definition

T"pes of FM A
Score 1F K I H J $ # " ! 1 S2D2'I+5 C'I+2'IA 7a6ardous 'ithout 'arning 7a6ardous 'ith 'arning $er" 7igh 7igh Moderate #o* $er" #o* Minor $er" Minor %one 3CCU''26C2 D in J $er" 7igh D in B $er" 7igh D in Q 7igh D in JE 7igh D in QE Moderate D D in in MEE Moderate J3EEE Moderate D in DN3EEE #o* D in DNE3EEE #o* \D in D3NEE3EE 5emote D2+2C+I36 A!solute (ncertaint" $er" 5emote 5emote $er" #o* #o* Moderate Moderatel" 7igh 7igh $er" 7igh Almost Certain
%ote2 'hen completing a Process FM A3 first assume the material is good and the process is !ad- Then assume that the process is good and the material is !ad- #astl"3 revie* the process for safet" considerations-
DesignCProduct FM A
T"pes of FM A
% M & '
) 7elps to identif" potential product failure modes earl" in the product development c"cle) Increases the li+elihood that all potential failure modes and their effects on assem!lies *ill !e considered) Assists in evaluating product design re,uirements and test methods) sta!lishes a priorit" for design improvement) Documents the rationale !ehind design changes and helps guide future development projects) IS P54ACTI$ X Should !e done *hen ne* products are designed or existing products are changed-
Defect FM A
T"pes of FM A
) 7elps identif" the root causes of defects% ) sta!lishes a priorit" for improvement activitiesM ) Documents plan of action) Provides methodolog" to !attle initial ground s*ell & of defects' ) Focuses effort on defects *ith highest L impact) IS %4T P54ACTI$ X
Scoring Criteria
T"pes of FM A
Score DE R Q O P N M B J D
(se actual defect ,uantities
S $ 5ITG C5IT 5IA

7a6ardous 'ithout 'arning 7a6ardous 'ith 'arning $er" 7igh 7igh Moderate #o* $er" #o* Minor $er" Minor %one
4CC(55 %C
$er" 7igh $er" 7igh 7igh 7igh Moderate Moderate Moderate #o* #o* 5emote
D T CTI4%
A!solute (ncertaint" $er" 5emote 5emote $er" #o* #o* Moderate Moderatel" 7igh 7igh $er" 7igh Almost Certain
%ote2 To change header information3 clic+ on Z$ie*Z then Z7eaderZ5IS<2 4ptional field used to reflect the pro!a!ilit" of completing actions-
The Catapult
FM A xercise Anal"6e the Catapult process using the FM A tool?5emem!er *e *ant to get the 0most !ang for the !uc+1-@
) ) ) ) ) )
Area+ into the Catapult teams 'e have alread" constructed a process map First3 *e *ill construct a cause and effect diagram Then *e *ill complete at least t*o failure modes for the most critical step?s@ of our process Appoint a spo+esman for "our team to de!rief the class on "our progress3 ,uestions3 etcComplete the FM A ?FM Aform-xls@ for the Catapult process !efore the third session ?'e *ill use this information for our D4 competition@
JN minutesX
Catapult Process Map

FM A xercise
Start Assem!le Catapult Pins ?J@ Arm 5u!!er Aand Aall Arm moves smoothl" Secure to ta!le Select Catapult Settings Plan or Prediction e,uation Computer Feasi!le settings Set Catapult Pins Positions Designated Clamp Tape Measure Tape
Aligned *ith tape ] B inches
Correct settings
Pull Arm to Proper Angle
Shoot
Measure distance
5ecord distance
Stop
4perator Consistenc" %o Parallax
4perator #ateral movement Correct angle
Tape Measure 4!servers positioned properl"
5ecorder Computer
Aall fl"s straight
Accurate measurement ] J inches
Correct distance recorded
Datafile/Catapultflo-*ig!
Complete the Diagram Aelo*

FM A xercise
Men Method Calculation procedure Material 5elease consistenc" 5u!!er Aand Aall Angle measurement
Distance Air Conditioner Mother nature Machine Arm moves freel" 5epeata!ilit" 5eproduci!ilit" Measure
Datafile/CatapultC.+* '0
'hen To (pdate an FM A8
FM A Summar" An FM A should !e updated *henever a change is !eing considered to a product/s2
u u u u u
design application environment material product/s manufacturing or assem!l" process
Summar" of ProductCProcess FM A/s

FM A Summar" u 'hat is the tool8
Spreadsheet
u'hat u'hen
do "ou appl" this tool8

'hen evaluating product for ro!ustness ?functionalit"3 producea!ilit"3 relia!ilit"@ During earl" stages of defect reduction efforts to identif" causes 'hen identif"ing +e" processCproduct parameters and evaluating methods for controlling them
*ill the tool identif"Csho*8

All productCprocess failure modes3 related effects3 causes3 & methods of controlling them 5is+ Priorit" %um!er ?5P%@ for action !ased on failure severit"3 pro!a!ilit" of occurrence and detection capa!ilit" ActionsCplans to reduce elements of 5P%
u'hat
results can "ou expect8

#earn to identif" critical productC process parameters Achieve consensus on solutions and methods of implementation Detailed productCprocess understanding
<e"s to Success
FM A Summar" uIdentif" purpose---A
SP CIFICX
u u(nderstand effects---I%$4#$ C(ST4M 5S & S(PP#I 5SX u u#in+ to the process mapu u(se to prioriti6e efforts3 allocate resourcesu u(se as a ris+ assessmentCprioriti6ation tool !ased on predicted u u(se to !uild consensus on prioriti6ationu u ncourage creativit"---T AM'45<X u uP#A%X u uAS< Q( STI4%SX
impact-
Process FM A Steps
'hat is an FM A8
uSteps
Completed Prior to FM A2
uFM
A Steps2
DJ-
Charter Team Develop and Characteri6e Process Map D-Identif" 07eav" 7itter1 Process Step D-Identif" Associated "/s ?Product Parameters@ D-Identif" Failure Mode D-Identif" Failure ffectsC5ate Severit" D-Identif" CausesC5ate 4ccurrence D-Identif" Controls ?if an"@C5ate Detection D-Calculate 5P% D-Prioriti6e !" 5P% 4rder D-Determine ActionsCPlan D-5ecalculate 5P% Aased on Plan D-Ta+e Action
BMNPOQRDE-
FM A Appendix
<e" Definitions for FM A
Severit" is an assessment of ho* serious the effect of the potential failure mode is on the customerThe customer in this case could !e the next operation3 su!se,uent operations3 or the end user-
4ccurrence is an assessment of the li+elihood that a particular cause *ill happen and result in the
failure mode-
Detection is an assessment of the li+elihood that the current controls ?design and process@ *ill detect
the cause of the failure mode3 should it occur3 thus preventing it from reaching "our customerThe customer in this case could !e the next operation3 su!se,uent operations3 or the end user-
Current Controls ?for !oth design and process@ are the mechanisms *hich prevent the cause of the
failure mode from occurring3 or detect the failure mode3 should it occur3 !efore the product reaches "our 0customer-1 For example3 current controls include SPC3 inspections3 *ritten procedures3 training3 preventive maintenance and all other activities that ensure a smooth running process-
Critical Characteristics are those items *hich affect customer safet" andCor could result in non=
compliance to regulations and thus re,uire controls to ensure DEES compliance- These are usuall" process0settings1 such as temperature3 time3 speed3 etcaccepta!le levels of capa!ilit"-
Significant Characteristics are those items *hich re,uire SPC and ,ualit" planning to ensure
FM A Appendix
Terminolog"
AACDF97I.<#M%4PQ5STProcess or Product %ame ; Description of Process or Product !eing anal"6ed5esponsi!le ; %ame of Process 4*nerPrepared A" = %ame of Agent coordinating FM A stud"FM A Date ; Dates of Initial and su!se,uent FM A 5evisionsProcess StepCPart %um!er ; Description of individual item !eing anal"6edPotential Failure Mode ; Description of ho* the process could potentiall" fail to meet the process re,uirements andCor design intent3 i-e- a description of a non=conformance at that Potential Failure ffects ; Description of the effects of the Failure Mode upon the customer3 i-e- *hat the next user of the process or product *ould experience or noticeS $ ?Severit"@ ; An assessment of the seriousness of the effect of the potential failure mode Potential Causes ; Description of ho* the failure could occur3 descri!ed in terms of something 4CC ?4ccurrence@ ; Description of ho* fre,uentl" the specific failure cause is expected to Current Controls ; Description of process controls that either prevent3 to the extent possi!le3 D T ?Detection@ ; An assessment of the pro!a!ilit" that the current controls *ill detect the potential cause3 or the su!se,uent failure mode5P% ?5is+ Priorit" %um!er@ ; The product of the Severit"3 4ccurrence3 and Detection 5an+ings i-e-3 5P% F S $ [ 4CC [ D TActions 5ecommended ; Actions to reduce an" or all of the 4ccurrence3 Severit" or Detection ran+ings5esponsi!ilit" ; Person or group responsi!le for the 5ecommended ActionActions Ta+en ; Arief description of actual action and effective date%e* S $ 5ITG 5ating after corrective action%e* 4CC(5 %C 5ating after corrective action%e* D T CTI4% 5ating after corrective action5esulting ne* 5P% after corrective action-
Introduction to Process Capa!ilit"
A!out this Module

) Process capa!ilit" ena!les the prediction of the a!ilit" of an" process to produce products and services that meet their desired specifications) This module focuses on t"pical manufacturing processes- Transactional and other manufacturing processes are not discussed here) The principles of process capa!ilit" *ill !e introduced and Minita! *ill !e used to calculate process capa!ilities
#earning 4!jectives
At the conclusion of this module participants *ill !e a!le to2
D-5ecogni6e the value of and uses for process capa!ilit"J-Calculate and explain the capa!ilit" of processes *hose output is normall" distri!utedB-Predict the pro!a!ilit" that the output of a process *ill !e *ithin its specification limits-
'e #ive in a Statistical 'orld

Aasic Statistics
) Statistics have a pervasive influence on our lives

; ver" da" there is another poll ; Sampling is !eing used to perform man" aspects of the census ; All major economic indicators are !ased on samples ; T$ ratings are !ased on samples ; Statistics determine insurance rates
) Quantum ph"sics has demonstrated that pro!a!ilit" determines the structure and
T"pes of Statistics
Definitions Descriptive statistics is the process of descri!ing the information *e have'e summari6e information from a sample or population give a clear understanding3 or description3 of the dataInferential statistics is the process of using information from a smaller set of data ?sample@ to reach conclusions or inferences a!out a larger group ?population@- (suall"3 *e have onl" sample information3 not the entire population3 and must infer understanding of the population !ased on our sample- 'e *ant these conclusions to !e mathematicall" correct-
Data T"pes
Definitions Attri!ute Ges = no 9ood = !ad Accept = reject Discrete Multiples of *hole units Can not !e meaningfull" divided Count or classification Continuous Can !e meaningfull" divided into finer and finer increments of precision *eight3 length3 voltage3 time
Measures of Central Tendenc" = #ocation

Definitions Mean = the sum of all mem!ers divided !" the population si6e ?average@ 2
= Population mean= 0 = Sample mean
0 ; 0 ; 0 999 0
1 1 ?
Mode = the most fre,uentl" occurring or most li+el" value Median = the fiftieth percentile ?half the values are a!ove and half !elo* the median@
Population $ersus Sample

Definitions Population Si6e % n #ocation Average ?Mean@ Dispersion2 $ariation $ariance Std dev 5ange
J
Samples x
sJ s 5 F K7i =K#o
Statistics infer information a!out the parameters of the population-
Quantif"ing Dispersion = Spread

Definitions xD xn xJ
'e could add the differences !et*een each value x and the average of the values x ho*ever that *ould al*a"s "ield 6ero- Therefore *e s,uare the difference !et*een each x and x3 to eliminate the negatives and emphasi6e the 2 outliers3 then ta+e the average of the results- This is defined as the variance or J- 4!viousl"3 =
Attri!utes of the 7istogram = #ocation & Spread

Definitions
%um!er of Cases F NE Mean & Median F ON Standard Deviation F Q-BJRR 5ange F ME $ariance F PR-BQQ Minimum F NN Maximum F RN
s= =
2 ( X X ) i i=1
DM
n 1
F r DE e , Q u e P n c M "
J E NE PE OE QE RE DEE DDE
DJ
X i X= = i=1 n
$aria!le K measurements2 ON QN PE QE OE QE PN ON OE ON QE OE QE ON ON NN OE QN ON RN ON OE QE ON ON OE QE RE ON RE PN QN QE OE ON OE ON QE QE PE OE OE PN QN QN QN PN PN QE PN
$alues of K
Measures of $aria!ilit" = $ariance F Sigma S,uared

Definitions
uSigma
S,uared is a measure of dispersion of the population a!out the mean u$ariances are not in the units of interest^ standard deviations are in the units of interest u$ariances are additive^ standard deviations are not additive---
Iso
H H
H H
B B
is 4<3
!ut3
is %4T 4<
Measures of $aria!ilit" ; S- D- F Sigma

Definitions Standard Deviation is a measure of dispersion of the population a!out the mean F the units of interest and is population standard deviation F population mean % F total population s F estimate of standard deviation + 0 + 0? n F sample si6e 01 999 0 1 (0 )= 1
) (
) (
) (
) )
M LsL
(0
) + (0
) + (0
n34
) 999(0
%ormal Distri!ution
ach curve sho*n here has2 u An area of one u A mean of 6ero u A standard deviation of Therefore3 the same S of the population is under each of the curves for n a!out the mean-
Quantif"ing the %ormal Distri!ution

Definitions
3 +3
1
689 6@ 95946@ 999!?@
+1
+2
Pro!a!ilit" Densit" Function ; Standard %ormal Distri!ution

Definitions Area under the curve F D L1 L0 An" normal distri!ution can !e converted to a standard normal distri!ution The formula for the pro!a!ilit" densit" function is z
f 7 z8 =
1 e
The Standard %ormal Transform

Definitions
0 5=
Permits conversion of an" data point ?K@ into a > value- This value allo*s us to loo+ up the percentage of the population that is a!ove and !elo* the data point-
Sample QuestionsIII
Q-
A ne* iron ore mine is discovered- DE <g ore is collected from each of JE spots-
D-
!- This procedure is called as YYYYYYYYYY c- Average is calculated from iron content of each of JE
spots- This value is called as YYYYYYYYYYYYYYYY d- If *e calculate average3 range and standard deviation from the iron content values of each lot3 this data is called as YYYYYYYYYYYYYYYY e- Predictions are made regarding average iron content of the mine3 total iron that can !e extracted3 impurities present etc- the anal"sis is called as YYYYYYYYYYY f- The a!ove calculations *ill give ans*ers *hich *ill !eDEES correct- True C False
Ans*ers
a.The procedure is called as SAMPLING. b. c.The value is called as STATISTIC. d. e.The values are called as STATISTICS. f. g.The analysis is called as STATISTICAL ANALYSIS. h. i.The S a e!en is "ALS#.
Aasic Principle
) All measures of process capa!ilit" are !ased on the concept of calculating the num!er of standard deviations !et*een the process center and the specification limits) A Six Sigma process has six units of standard deviation !et*een the process center and !oth specification limits-
#S#
(S#
$isuali6ing Process Capa!ilit"
Process *idth Specification *idth
Quantif"ing Process Capa!ilit"

-ield of a one sigma process %./$0
If *e assume the process is centered on the target and does not shift or drift the "ields *ould !eProcess Sigma . * ) + : "ield /.:,*:,(+ /.(+))((9 /.((9-//* /.((((-:9 /.(((((() .
#S #
(S #
The Standard Deviation

D Sigma = PQS J Sigma = RNS B Sigma = RR-OB S
?K ; K@J n
1
B&&er '&ecification +i#it 7B'+8 1arget '&ecification 718 +o(er '&ecification +i#it 7+'+8 2ean of the distri3ution 7 8 'tandard 4e$iation of the distri3ution 7 8
&7d8 B'+
Calculating Gield
'e +no* that > is the num!er of units of standard deviation on a standard normal curve *hich has a mean of 6ero and a standard deviation of one- 'e also +no* an" normal distri!ution can !e converted to the standard normal using the > e,uation-
x Z =
X X Z = 1
If the specification limits are su!stituted for x *e can determine the num!er of units of standard deviation !et*een the process center and the specification limits on the standard normal curve- Then *e can use the ta!les to loo+ up the pro!a!ilit" a value *ill !e less than that num!er-
In most cases *e do not +no* or so *e su!stitute the sample statistics for the population parameters as sho*n-
Calculating Gield
The mean time ta+en for completing an operation is NEE hrs-
and this is normall" distri!uted *ith a standard deviation of DEE hrsD- 'hat is the pro!a!ilit" that an operator ta+en at random *ill ta+e !et*een NEE to PNE hrs to complete the operation8 J- 'hat is the pro!a!ilit" that he *ill ta+e _ OEE hrs8
xpress in graphical form also-
Calculating Gield ; xample

D- 'hat is the pro!a!ilit" that an operator ta+en at random *ill ta+e !et*een NEE to PNE hrs to complete the operation8
Z =

F ?PNE ; NEE@ C DEE F D-N
#oo+ing up 6 ta!le3 the corresponding value under the Standard %ormal Distri!ution is E-MBBJ- i-e- MBS-
Calculating Gield ; xample

J- 'hat is the pro!a!ilit" that he *ill ta+e _ OEE hrs8 B-
x F ?OEE ; NEE@ C DEE F J #oo+ing up 6 ta!le3 the corresponding value under the
Z =
Standard %ormal Distri!ution is E-MOOJThus the pro!a!ilit" of an operator ta+ing more than OEE hrs is ?E-N ; E-MOOJ@ F E-EJJQ3 i-e- slightl" over JS-
Calculating Gield xample

Consider a process that has the follo*ing specification limits2 #o*er Specification #imit ?#S#@ of =D and a (pper Specification #imit of D- Data indicates the process is centered on E *ith a standard deviation of D- 'hat is the "ield8 (sing ta!les or soft*are to loo+ up the area under the curve *hen >FD *e U!234 5U!26 find -QMDB- This means that QM-DBS of 1 7 the product has a value less than the #3% upper specification limit5U!26 =# # Again using ta!les or soft*are to loo+
2!234 52!26 1 7 3#3% 52!26 = # #
up the area under the curve *hen >F=D *e find -DNQP- This means that DN-QPS of the product has a value less than the lo*er specification limit-
'hat if the Process Shifts8

9enerall" spea+ing3 processes have !een o!served to shift andCor drift D-N standard deviations over time- 7o* *ould that effect the "ield of a one sigma process8
2!2 U!2
39
30
38
3#
Calculating Shifted Process Gield

(sing ta!les or soft*are to loo+ up the U!234 5U!26 area under the curve *hen >F=-N *e 1 7 find -BEQN #3%3#.: 5U!26 = .: # 2!234 Again using ta!les or soft*are to loo+ 52!2 6 1 7 up the area under the curve *hen >F= J-N *e find -EEPJ3#3%3#.: 52!2 6 = 8.: # Su!tracting the t*o *e o!tain -BEJB
(sing These Principles

) Process capa!ilit" and process capa!ilit" indices are unam!iguous and *ill !e addressed first) Process sigma is some*hat am!iguous and *ill !e addressed second-
Process Capa!ilit" Terms

Measures of the a!ilit" of a process to produce compliant productsCservices2
Cp ? Short?term process capability
uFor
a limited period of time ?not including shifts and drifts@ not consider process centering +no*n as process entitlement
uDoes uAlso
Cpk ? Short?term process capability inde<

uFor
a limited period of time ?not including shifts and drifts@ consider process centering
uDoes
Pp ? =ong?term process capability

uFor
an extended period of time ?including shifts and drifts@ not consider process centering
uDoes
Ppk ? =ong?term process capability inde<

uFor
an extended period of time ?including shifts and drifts@ See formulae on ne!t page* uDoes consider process centering
Capa!ilit" Formulae
U!232!2 Specification 'idth ?s@ Cp F Short=Term Process 'idth F ; !, /7
Pp F Cp+ F Pp+ F
Specification 'idth ?s@ F #ong=Term Process 'idth #esser of2 #esser of2
U!234 ; !, 07
U!232!2 ; 2, /7
432!2 ; 07
!,
or or
U!234 ; 2, 07
432!2 ; 07
2,
(sing Minita!
The data is continuous so test for normalit" Stat_Aasic Statistics_%ormalit" Test
The %ormalit" Test

Probability Plot of Caps
'ormal <<.<< << <: Percent $% :% 8% : # %.%# %.<<8 %.<<9 %.<</ %.<<$ #.%%% #.%%8 #.%%9 #.%%/ #.%%$ Caps
Mean #.%%% !t)e= %.%%#<$/ ' >:% A) %./#< ?3@alue %.#%>
Wor sheetA Bottle Caps.M,W
The P value is _ -EN therefore do not reject the assumption of normalit"-
(sing Minita! to Calculate Process Capa!ilit"
Minita! 5esults
Process Capability of Caps
2!2
?rocess )ata 2!2 %.<<: ,arget # U!2 #.%%: !am ple Mean #.%%%%/ !am ple ' >:% !t)e=CWithinD %.%%#<$90# !t)e=CE=erallD %.%%#<$/0/
,arget
U!2
Within E=erall
?otential CWithinD Capa&ility 5 .Bench 8.8/ 5 .2!2 8.:: 5 .U!2 8.9< Cp %.$0 E=erall Capa&ility 5 .Bench 5 .2!2 5 .U!2 ?p Cpm 8.8/ 8.:: 8.9$ %.$0 %.$9
: % : % : % : % <9 <</ <<> <<< %%% %%8 %%0 %%: < %. %. %. %. #. #. #. #.

E&ser=ed ?erform ance F ( 2!2 %.9% F G U!2 %./> F ,otal #.%> EHp. Within ?erform ance F ( 2!2 %.:9 F G U!2 %./9 F ,otal #.#$ EHp. E=erall ?erform ance F ( 2!2 %.:9 F G U!2 %./: F ,otal #.#<
W or sheetA Bottle Caps.M,W
#et/s examine this in detail
Process Data
2!2
?rocess )ata 2!2 %.<<: ,arget # U!2 #.%%: !ample Mean #.%%%%/ !ample ' >:% !t)e=CWithinD %.%%#<$90# !t)e=CE=erallD %.%%#<$/0/
,arget
U!2
ly %rom the data& +he -ithin standard deviation E=erall is the pooled standard
?otential CWithinD Capa&ility 5.Bench 8.8/ 5.2!2 8.:: 5.U!2 8.9< Cp %.$0 E=erall Capa&ility 5.Bench 5.2!2 5.U!2 ?p Cpm 8.8/ 8.:: 8.9$ %.$0 %.$9
Within
9: /% >: <% %: 8% 0: :% << .<< .<< .<< .%% .%% .%% .%% . % % % % # # # #
E&ser=ed ?erformance F ( 2!2 %.9% F G U!2 %./> F ,otal #.%> EHp. Within ?erformance F ( 2!2 %.:9 F G U!2 %./9 F ,otal #.#$ EHp. E=erall ?erformance F ( 2!2 %.:9 F G U!2 %./: F ,otal #.#<
4!served Performance
2!2 ,arget U!2
ent o% product that -as outside o% the upper and lo-er speci%ication lim Within
E=erall
9: /% >: <% %: 8% 0: :% << .<< .<< .<< .%% .%% .%% .%% . % % % % # # # #
xpected 'ithin Performance

,arget 2!2
utside o% the upper and lo-er speci%ication limits on an short term basis& +hi
U!2
Within E=erall
9: /% >: <% %: 8% 0: :% << .<< .<< .<< .%% .%% .%% .%% . % % % % # # # #
xpected 4verall Performance

,arget 2!2
o-er speci%ication limits on an long term basis& +his pro7ection is based on t

U!2
Within E=erall
9: /% >: <% %: 8% 0: :% << .<< .<< .<< .%% .%% .%% .%% . % % % % # # # #
Potential 'ithin Capa!ilit"

,arget 2!2 U!2
rocess mean@ the respective speci%ication limits and the -ithin standa ted -ithin yield 8&KI!9 le%t hand tail o% a standard normal then looking Within
E=erall
or N =S= &
9: /% >: <% %: 8% 0: :% << .<< .<< .<< .%% .%% .%% .%% . % % % % # # # #
%ote2 Minita! uses *ithin to descri!e short term variation and overall to descri!e long term variation-
4verall Capa!ilit"
2!2 ,arget U!2
Process Capability of Caps process mean@ the respective speci%ication limits and the overall stand cted -ithin yield 8&KI19 le%t hand tail o% a standard normal then looking Within
E=erall
US=
or N =S= &
: % : % : % : % <9 <</ <<> <<< %%% %%8 %%0 %%: < %. %. %. %. #. #. #. #.

'ithin = Aet*een
2!2
,arget
U!2
e as indicated by the Within and ,et-een lines being very close

Within E=erall
: % : % : % : % <9 <</ <<> <<< %%% %%8 %%0 %%: < %. %. %. %. #. #. #. #.

Calculating Cp+ and Pp+

2!2
e process mean@ the respective speci%ication limits and the -ithin standard d Process Capability of Caps cted -ithin yield 8&KI!9 le%t hand tail o% a standard normal then looking up the
,arget U!2
o% CP= or CPU&
Within E=erall
?otential CWithinD Capa&ility Cp %.$9 C?2 %.$: C?U %.$0 Cp %.$0 E=erall Capa&ility ?p ??2 ??U ?p Cpm %.$9 %.$: %.$0 %.$0 %.$9
9: /% >: <% %: 8% 0: :% << .<< .<< .<< .%% .%% .%% .%% . % % % % # # # #
E&ser=ed ?erformance ??M ( 2!2 9%%%.%% ??M G U!2 ////./> ??M ,otal #%///./> EHp. Within ?erformance ??M ( 2!2 :0::.%$ ??M G U!2 /908.%9 ??M ,otal ##>$>.#8 EHp. E=erall ?erformance ??M ( 2!2 :0<:.:< ??M G U!2 /9>$.9$ ??M ,otal ##$>9.%>

,arget 2!2 U!2
e process mean@ the respective speci%ication limits and the overall standard d cted -ithin yield 8&KI1"9 le%t hand tail o% a standard normal then looking up the
Within E=erall
?otential CWithinD Capa&ility Cp %.$9 C?2 %.$: C?U %.$0 Cp %.$0 E=erall Capa&ility ?p ??2 ??U ?p Cpm %.$9 %.$: %.$0 %.$0 %.$9
o% PP= or PPU&
9: /% >: <% %: 8% 0: :% << .<< .<< .<< .%% .%% .%% .%% . % % % % # # # #
E&ser=ed ?erformance ??M ( 2!2 9%%%.%% ??M G U!2 ////./> ??M ,otal #%///./> EHp. Within ?erformance ??M ( 2!2 :0::.%$ ??M G U!2 /908.%9 ??M ,otal ##>$>.#8 EHp. E=erall ?erformance ??M ( 2!2 :0<:.:< ??M G U!2 /9>$.9$ ??M ,otal ##$>9.%>
A Process that Drifts

Consider a similar process that does change over timeThe data is in Caps Drift-MT'First test for normalit"
%ormalit" Test
Probability Plot of Caps Drift
'orm al <<.<< << <: Percent $% :% 8% : # %.%# %.<<:%
Mean #.%%% !t)e= %.%%#8<# ' >8% A) %.08/ ?3@alue %.:8%
%.<<>:
#.%%%% Caps Drift
#.%%8:
#.%%:%
Wor sheetA Caps )rift.M,W
Do not reject the assumption of normalit"-
Capa!ilit" Anal"sis
Capa!ilit" Anal"sis
Process Capability of Caps Drift
2!2 ,arget U!2
?rocess )ata 2!2 %.<<: ,arget # U!2 #.%%: !ample Mean %.<<<<<$ !ample ' >8% !t)e=CWithinD %.%%#%8%// !t)e=CE=erallD %.%%#8<%::
time as indicated by the di%%erence bet-een the WithinWithin and

E=erall
?otential CWithinD 5.Bench 5.2!2 5.U!2 Cp 5.Bench 5.2!2 5.U!2 ?p Cpm Capa&ility 9.>/ 9.<% 9.<% #./0 0.>% 0.$> 0.$$ #.8< #.8<
hart&
E=erall Capa&ility
%.<</% %.<<>: %.<<<% #.%%%: #.%%8% #.%%0: #.%%:%

E&ser=ed ?erformance F ( 2!2 %.%% F G U!2 %.%% F ,otal %.%% EHp. Within ?erformance F ( 2!2 %.%% F G U!2 %.%% F ,otal %.%% EHp. E=erall ?erformance F ( 2!2 %.%# F G U!2 %.%# F ,otal %.%#
Wor sheetA Caps )rift.M,W
Process Capa!ilit" xercise D
D ] -EP Inches
Data indicate the process is centered *ith a standard deviation of -EJCalculate the "ield-
xercise D $isuali6ing the Ans*er

=-EP =-EM =-EJ H-EJ H-EM H-EP
Process average
Design *idth ?-DJ Inches@
There are B units of standard deviation !et*een the process average and the specification limits therefore this is a B sigma process ?short term@-
xercise D Calculating the Gield

(sing ta!les or soft*are to loo+ up the U!234 5U!26 area under the curve *hen >FB *e 1 7 find -RRQP #.%/3# 5U!26 =0 .%8 Again using ta!les or soft*are to loo+ 2!234 52!2 6 up the area under the curve *hen >F=B 1 7 *e find -EEDBN.<93# 52!2 6 = 0 Su!tracting the t*o *e find a "ield of .%8 RROB
xercise J
Assume a process o*ner has as+ed "ou to anal"6e the data in Process Capa!ilit" xercise D-MT'- Parts A and A are made on different machines in lots ?su!groups@ of NThe customer has esta!lished specification limits of DE ] -D and re,uires a Pp+ of D-BBPrepare a !rief presentation to descri!e "our anal"sis and recommendations8 5emem!er to present data practicall"3 graphicall" and anal"ticall"-
xercise J %ormalit" Tests

Probability Plot of Part A
'ormal <<.<< << <: Percent $% :% 8% : # %.%#
Mean #%.%% !t)e= %.%:%9< ' >:% A) %.98% ?3@alue %.089
<.$
<.<
#%.% Part A
#%.#
#%.8
Wor sheetA ?rocess Capa&ility EHercise #.M,W
Probability Plot of Part B

'ormal <<.<< << <: Percent $% :% 8% : # %.%#
Mean #%.%% !t)e= %.%:/9> ' >:% A) %.89$ ?3@alue %.>:8
%o reason to reject the assumption of normalit" for either part-
<.$
<.<
#%.% Part B
#%.#
#%.8
xercise J Process Capa!ilities
xercise J Part A Process Capa!ilit"

Process Capability of Part A
2!2
?rocess )ata 2!2 <.< ,arget #% U!2 #%.# !ample Mean #%.%%%< !ample ' >:% !t)e=CWithinD %.%9</009 !t)e=CE=erallD %.%:%9<88
,arget
U!2
Within E=erall
?otential CWithinD Capa&ility Cp %./> C?2 %./$ C?U %./> Cp %./> E=erall Capa&ility ?p ??2 ??U ?p Cpm %.// %./> %./: %./: %.//
<.$$ <.<8
E&ser=ed ?erformance F ( 2!2 #.$> F G U!2 8.$% F ,otal 9./> EHp. Within ?erformance F ( 2!2 8.#% F G U!2 8.0% F ,otal 9.9%
<.</ #%.%% #%.%9 #%.%$ #%.#8

EHp. E=erall ?erformance F ( 2!2 8.8$ F G U!2 8.9< F ,otal 9.>>
xercise J Prepare the Control Charts
Control Chart Sho*s Process Sta!ilit"

Xbar R C!art of Part A
UC26#%.%/>: #%.%: Sample Mean
#%.%%
I I 46#%.%%%<
<.<: 2C26<.<090 # #/ 0# 9/ /# >/ Sample <# #%/ #8# #0/
UC26%.899# Sample Range %.8
9
%.#
9 9 9
I J6%.##:9
%.% # #/ 0# 9/ /# >/ Sample <# #%/ #8# #0/
2C26%
Process Capa!ilit" Part A

Process Capability of Part B
2!2
?rocess )ata 2!2 <.< ,arget #% U!2 #%.# !ample Mean <.<<</# !ample ' >:% !t)e=CWithinD %.%9</009 !t)e=CE=erallD %.%:/9>8>
,arget
U!2
Within E=erall
?otential CWithinD Capa&ility 5.Bench #.># 5.2!2 8.%# 5.U!2 8.%8 Cp %./> E=erall Capa&ility 5.Bench 5.2!2 5.U!2 ?p Cpm #.90 #.>/ #.>$ %.:< %.:<
<.$9
E&ser=ed ?erformance F ( 2!2 9.#0 F G U!2 9.%% F ,otal $.#0
<.<%
<.</
#%.%8 #%.%$ #%.#9
EHp. Within ?erformance F ( 2!2 8.89 F G U!2 8.#/ F ,otal 9.0<
EHp. E=erall ?erformance F ( 2!2 0.$< F G U!2 0.>> F ,otal >.//
Control Chart Sho*s Time Aased $ariation

Xbar R C!art of Part B
# #
#%.%: Sample Mean #%.%% <.<: <.<% #
##
# :
UC26#%.%//8
I I 46<.<<</ : #
#/ 0# 9/
#
/# >/ Sample <# #%/
: #
#8#
2C26<.<00%
#
#0/
UC26%.899# Sample Range %.8
9
%.#
9 9 9
I J6%.##:9
%.% # #/ 0# 9/ /# >/ Sample <# #%/ #8# #0/
2C26%
Confidence Intervals
A!out This ModuleI

Confidence Intervals ?CI@ permit us to state that *e are KS confident that the population parameter of interest is at most a specified interval from the sample statistic-
Si! Sigma" # $uest for Process Perfection Meet 'oals and #ttac% &ariation
(Data)ile(PurchOrd*mt(Data)ile(P-rSuply*mt(Data)ile(Conf3 nt*mt(Data)ile(O+ac%*mt-
'hat 'e 'ill #earn---
D-Significance of confidence intervals 9 B-7o* to calculate confidence intervals for2 Means Standard deviations or $ariation
Confidence in the Midst of (ncertaint"8

uMean
and Standard Deviation statistics are2 estimates of the population Mu/s ? @ and Sigma/s ? @ !ased on one sample
%inet"=five Percent Certain
u$aria!ilit" exists from sample to sample

u uA"
using statisticall" !ased confidence inter8als9 uncertaint" can !e ,uantified
u u (suall"3 RNS confidence intervals The chances are approximatel" RN
are out of DEE calculated that the calculated confidence inter8al contains the population parameter3 orI 'ith RNS certaint"3 the population parameter is inside the confidence interval-
Population vs- Sample

Population is the entire area of interestSample is a su!set of the population'hat is the relationship !et*een the population and the sample8 7o* representative is this sample8
Population
Sample
Confidence Interval S"m!ols and Definitions

Measure
Mean
Population Sample Parameter Statistic
(se
n ?0 > is +no*n n < ?0 t
J Cp p
$ariance Standard Deviation Process Capa!ilit" Proportion Alpha 5is+
sJ s Cp
` p `

F or > ?approx@ T"picall" -EN
Confidence Intervals ?CI@

CI ta+e the general form 2 C-I-FStatistic HC= < [ ?Standard Deviation@ StatisticF Mean3 $ariance3 CP3 etc< F Constant !ased on a statistical distri!ution CI reflect the sample to sample variation of our point estimates 'e *ill loo+ at CI for2 3
K
3 and C P
'hat is the Student t=distri!ution8

The Student t distri!ution is a famil" of !ell shaped ?%ormal li+e@ distri!utions that var" !" degrees of freedom ?sample si6e@ = the fe*er degrees of freedom3 the *ider and flatter the distri!utionE-MN E-M E-BN E-B E-JN E-J E-DN E-D E-EN E
N N N N N N N E D J B N B=M =B =J =J =D =E ED=B =D JM
DFJ
DFDE
DFBE
t=Distri!utions3 %ormal Approximation3 5is+

To give an idea of the values of t compared to > for RNS ? F E-EN@3 loo+ at the ta!le !elo*2 Sample
N DE JE BE DEE DEEE
t=value
J-OQ J-JP J-ER J-EN D-RQ D-RP
>=value
D-RP D-RP D-RP D-RP D-RP D-RP F ris+
D L090 5
D L 090 5
'e can use > to estimate t if n BE and is +no*n

'1/1I'1I-' ).0 %=P%0I2%51%0'-- B,,
7"pothesis Testing
A!out This ModuleI

7"pothesis Testing helps2 uDetermine if there is a statisticall" significant difference !et*een t*o relativel" small samples uQuantif" the ris+s of ma+ing an incorrect decision
'hat 'e 'ill #earn

D-7o* to test varia!le data a-(se a t=test to compare t*o means !-Alpha ? @ and Aeta ? @ 5is+s c-(se a paired t=test to compare paired treatments d-(se test for e,ual variances J-7o* to test discrete data a-Compare proportions !-Compare discrete data
5eal 'orld Scenario

Design2 Determine if t*o alternate design changes are significantl" differentManufacturing2 Determine if t*o different t"pes of material *ear differentl"AdministrativeCTransactional C Service2 Determine if the change to a process affected the c"cle time-
Purposes of 7"pothesis Testing

uDetermine
if there is a real difference !et*een 8 and 8 relativel" small samples to ans*er ,uestions a!out the populationthe associated ris+s-
u(se
uQuantif"
xample
.ld 4esign u7ard dis+ transfer speed 899! ?mega!"tes per second@ is 8194 marginal- A ne* design is 8498 proposed8!9? !99! uAn ngineering Change 8591 %otice ? C%@ is incorporated 819! 8?9! uIs the ne* design !etter8 8495 5e( 4esign 849! 8691 9199 869? !99? 869 8991 8?9! 8895
To ans*er this3 *e need some fundamentals of significance testing firstX
Steps in 7"pothesis Testing

D-Define the pro!lem J-Determine the o!jectives B- sta!lish the 7"pothesis 'rite the %ull 7"pothesis ?7E@ 'rite the Alternative 7"pothesis ?7a@ M-Determine the appropriate statistical test ?assume distri!ution >3 t3 F@ N-State the alpha ris+ ?usuall" N S@ P-State the !eta ris+ ?usuall" DE=JE S@ O- sta!lish the effect si6e ?delta@ Q-Compute the sample si6e R-Develop a sampling plan DE-Select the samples DD-Conduct the test and collect data DJ-Calculate the test statistic ?>3 t3 or F@ from the data DB-Determine the pro!a!ilit" that the calculated test statistic has occurred !" chance DM-If that pro!a!ilit" is less than alpha3 reject 7E DN-If that pro!a!ilit" is greater than alpha3 do not reject 7E DP-Translate the statistical conclusion into a practical solution
Significance Tests
u(sed
to assess evidence provided !" sample data to reject3 or fail to reject a claim a!out a population parameterh"pothesis ?7o@ is the statement *e assessis usuall" stated as3 0there is no difference1h"pothesis ?7a@ is usuall" stated as 0there is a difference1fail to reject 7o unless there is convincing evidence to reject it-
T"pical xamples
u%ull
u7o
uAlternate
u'e
!o ( a F b !a ( a b !o ( a F b !a ( a b !o ( "a F "b !a ( "a "b
'hat does NS #evel of Significance Mean8

This means that *e *ill reject the null h"pothesis if the difference !et*een the sample statistic and the h"pothesi6ed population parameter is so large that it *ould occur3 on an average3 onl" N or fe*er times in ever" DEE samples *hen the h"pothesi6ed population parameter is correct-
T"pe I and II rrors2 Associated 5is+s

T"pe I errors are made *hen *e reject the null h"pothesis *hen in fact it is trueAlpha ? @ ris+ is the pro!a!ilit" of ma+ing a t"pe I errorT"pe II errors are made *hen *e fail to reject the null h"pothesis *hen in fact it is falseAeta ? @ ris+ is the pro!a!ilit" of ma+ing a t"pe II error-
This ris+ is t"picall" set at NS-
This ris+ is t"picall" set at DES
5is+s are set !efore the test or experiment is conducted
The 5is+ Truth Ta!le

is the ris+ of finding a difference *hen there reall" isn/t one is the ris+ of not finding a difference *hen there reall" is oneAction State of %ature 7o should not !e rejected ?7o is true@ 7o should !e rejected ?7o is false@
7o not rejected 7o rejected
Correct Decision
T"pe I or producer/s ris+ F P?T"pe I@
T"pe II or consumer/s ris+ = P?T"pe II@
Correct Decision
Another #oo+ at 5is+s

5emem!er2 is the ris+ of finding a difference *hen there reall" isn/t one is the ris+ of not finding a difference *hen there reall" is one,o ,a
%o* *e can determine if the C% improved performance
%ormal Distri!ution and t Distri!ution

When Po&ulation When Po&ulation '4 is Fno(n '4 is 5ot Fno(n 'a#&le siHe n is N 5or#al 5or#al ?0 distri3ution; H ta3le t distri3ution; H tta3le 'a#&le siHe O ?0; 5or#al distri3ution; and (e can assu#e distri3ution; H ta3le ta3le &o&ulation is a&&ro: nor#al
P $alue
P value is the smallest level of significance that *ould lead to rejection of the null h"pothesis 7oeg- supposing 7o *ere true3 *hat is the pro!a!ilit" of getting a value of x=!ar this far from the population mean8 This pro!a!ilit" is called a pro! value or p=value-
Manual Test %ull 7"pothesis
8 8 s# s8 )ifference Upper &ound = X# X 8 + t Kdf L + n# n8 )ifference Upper &ound = 3#.<< + ( #.>9:<L #.:>)
)ifference Upper &ound = .>:/

The upper !ound for the difference in the means indicates the difference !et*een the means of these populations could !e as great as -ONP ?at the RNS confidence level@Therefore3 the evidence is not statisticall" significant to conclude that the difference !et*een the ne* design and the old design is less than E- Fail to reject 7o-
Manual Test %ull 7"pothesis
t7 ;df 8 =
01 0 s1 s + n1 n
= -19 6
P = 911
P_ -EN therefore fail to reject 7o-
7"pothesis Testing Decision Tree

7"pothesis Testing
7o2 PDFPJ 7a2 PD0 PJ Minita!2 Stat=Aasic Stat=Dor J proportions
Proportions Testing ?J factors onl"@ Continuous Data ?4ne factor onl"@ Attri!ute Data
7o2 J factors are independent 7a2 J factors are dependent Minita!2 Stat=ta!les=Chi s,uare test
#on:normal
%ormalit" test
7o2 Data is %ormal 7a2 Data is %4T %ormal Minita!2 Stat=Aasic Stat=%ormalit" Test (se Anderson=Darling
#ormal
Contingenc" Ta!le
0 or more sam"les 7o2 DF JF BI 7a2 At least D is different #evene/s Test Minita!2 Stat=A%4$A=Test for ,ual $ariances For onl" J s3 this is similar to an F=test2 FF?SD@JC?SJ@J If Fcalc_Fcrit 3 reject null ?(se Chi=S,uared for D sample@
2 Sam"le
0 or 6ore Sam"les
Chi=S,uare
7o2 DF Target 7a2 D0 Target Minita!2 Stat=Aasic Stat=9raphical Summar" If target falls *ithin CD2 then fail to reject 7o
Aartlett/s TestCF=Test
7o2 DF JF BI 7a2 D At least J are different Minita!2 Stat=A%4$A=Test for ,ual $ariances For onl" J s3 this is same as F=test2 Stat_AasicStat_J $ariances FF?SD@JC?SJ@J If Fcalc_Fcrit 3 reject 7o
7o2 MD F Target 7a2 MD Target Minita!2 Stat=%onparametric=D Sample = sign ?or@ D Sample Stat=%onparametric=D Sample 'ilcoxon ?Also used for paired comparisons 7o2 MD=MJFE@ MDFMedian or sample D M target F Target Median
D Sample T Test
7o2 DF Target 7a2 D0 Target Minita!2 Stat=Aasic Stat=D Sample=T ?Also used for paired comparisons27o2 DF JFE@
4ne 'a" A%4$A
D Sample > Test

7o2 DF Target 7a2 D 0 Target Minita!2 Stat=Aasic Stat=D Sample=> ?Also used for paired comparisons2 7o2 DF JFE@ Sample Si6e _FBE is +no*n
7o2 DF JF BFI 7a2 D at least J are different Minita!2 Stat=A%4$A=4ne 'a" ?Caution Aartlett/s paE-EN^ assumesFvariances@ 0 Sam"les 7o2 D= JFE 7a2 D=#J 0 E Minita!2 Stat=Aasic Stat=Paired T Paired T Test ?$ariance F@ ?Compares Means *hen o!servations are paired or dependent in a pair*ise manner@
7o2 MDFMJFMBI 7a2 At least are different JJ or More Samples Minita!2 Stat=%onparametric=Mann='hitne" ?or@ Stat=%onparametric=<rus+al='allis ?or@ Stat=%onparametric=Freidmans MDFMedian sample D3 etc-
J Sample T Test
7o2 DF J 7a2 D0 J Minita!2 Stat=Aasic Stat=J Sample=T ?Compares Means using pooled Std Dev@ Chec+ !ox to assume e,ual variances or Chec+ !ox to assume une,ual variances
5egression and Correlation
A!out This ModuleI

Correlation Anal"sis is used to ,uantif"2 the degree of linear association !et*een varia!les 5egression Anal"sis is used to ,uantif"2 the functional relationship !et*een varia!les
(Data)ile(Correlat*mt(Data)ile(6arn*mt(Data)ile(Cases*mt(Data)ile(7ater*mt(Data)ile(8Mystery*mt(Data)ile(83+!ampl*mt(Data)ile(Call9ue*mt(Data)ile(Pizza*mt(Data)ile(Cases*!ls (Data)ile(C+O3COMP+1S#T O1*!ls (Data)ile(8ealestate*!ls (Data)ile(Oilcons*mt-
'hat 'e 'ill #earn

D-Correlation 7o* to measure the linear relationship !et*een t*o varia!les The correlation coefficient Implication of the correlation coefficient 0r1 B-5egression Definition of the regression line and ho* it is developed 7o* to calculate and anal"6e a regression e,uation 7o* to anal"6e relationships !et*een an independent varia!le and one or more dependent varia!les using regression 7o* to interpret rJ (nderstanding and anal"6ing residuals 7o* to use the regression A%4$A ta!le
u
5eal 'orld xamples

ADMI%IST5ATI$ A soft*are compan" *ants to +no* the relationship !et*een calls in ,ueue and service timeMA%(FACT(5I%9 A customer and supplier disagree on the ,uantit" received !" customer versus several months/ ,uantit" ordered for a given lead timeD SI9% A chemical engineer3 designing a ne* process3 *ants to investigate the relationship !et*een +e" input varia!les and stac+ loss of ammonia-
5egression and Correlation

u5egression
and Correlation anal"ses sho* us ho* to determine !oth the nature and the strength of a relationship !et*een t*o varia!lesregression anal"sis *e develop an estimating e,uation relating the dependent and independent varia!les- i-e- ho* much percent of variation can !e explained !" the regression e,uationcorrelation anal"sis *e determine the degree to *hich the varia!les are related-
uIn
uIn
5egression Anal"sis
u(sed uThe
to fit lines and curves to data
fitted lines Quantif" the relationship !et*een the process varia!les ?K/s@ and process performance ?G@ 7elp identif" the vital fe* K/s na!le predictions to !e made Identif" the impact of controlling the process varia!les ?K/s@
uProduces
an e,uation to match the line
Terms
Correlation A measure of linear association (sed *hen !oth K and G are continuous r values range from2 Perfect positive relationship F D %o relationship F E Perfect negative relationship F =D 5egression Provides the !asis for predicting the values of a varia!le from the values of one or more other varia!les (sed *ith a continuous G and continuous Ks3 or continuous G and categorical Ks rJ = proportion of variation of G explained !" the prediction e,uation
u
A 'ord of Caution
It
is important that *e consider the relationships found !" regression to !e relationships of association !ut not necessaril" of cause & effectis unless *e have specific reasons for !elieving that the values of the dependent varia!le are caused !" the values of the independent varia!le?s@3 do not infer causalit" from the relationships *e find !" regression-
That
Correlation Illustrated
JE DN DE N E E J M P Q DE
E E =N =DE =DN =JE =JN =BE
DE
Correlation F D
DJ DE Q P M J E E J M
Correlation F = D
DE
Correlation F E
Correlation xample
(Data)ile(Correlat*mt-
T*o test stations are used to measure po*er suppl" voltageIs there a correlation8 Minita!2 Stat_!asic stat_correlation Correlation of Station D and Station J F E-RNR3 P=$alue F E-EEE The t*o are highl" correlated ?-RNR@
Is this reasona!le8 Are "ou comforta!le *ith -RNR8 'hat does it mean to "ou8 7o* does the data actuall" loo+8 7o* *ould "ou find out8
Plot the Data

9raph_ScatterPlots
The Data
Scatterplot of Station 1 "s Station 2
<.9 <.0 <.8
If all of the <.# data points <.%diagonal *ere on the $.< line3 *ould *e have perfect correlation8 $.$
Station 1 $.> $./ $.: $./ $.$ <.% <.8 Station 2 <.9 <./
Wor sheetA Correlat.M,W
#et/s tr" regression
5egression
Fitted line plot is used *hen there is onl" one predictor-
xample D ?cont-@
#itte$ %ine Plot
!tation # 6 #.%8% O %.$>8< !tation 8 <.: <.9 <.0 <.8 Station 1 <.# <.% $.< $.$ $.> $./ $./ $.$ <.% <.8 Station 2 <.9 <./
! J3!M J3!MCadND %.%::>8$$ <8.%F <#.:F
In *hat *a"s is this graph different from the preceding one8 'hat are the implications8 'hat action *ould "ou ta+e8
Wor sheetA Correlat.M,W
Slide JJN *as actual line ; this is a fitted line- Can !e used for prediction-
Aest Fit #ine

#itte$ %ine Plot
Cases 6 88.9> O %.>:9/ ,est ?iece #%% <% $% Cases >% /% :% 9% 9% :% /% >% &est Piece $% <% #%%
! J3!M J3!MCadND ##.:#0# 9<.9F 9>.>F
A" minimi6ing the residual sum of s,uares3 *e get a !est fit line of the form2
Pi = a + 3= i
Wor sheetA cases.M,W
a F coefficient of the constant term or intercept ! F coefficient of the predictor3 K
Statistical Process Control
A!out This ModuleI

Control charts portra" process performance and separate causes of variation2 5andom Assigna!le Control Chart S"stems are2 A proven techni,ue for improving productivit" ffective in defect prevention Prevent unnecessar" process adjustments Provide diagnostic information Provide information a!out process capa!ilit"
(Data)ile(#ttribut mt(Data)ile(&ariable*mt-
'hat 'e 'ill #earn-
D- Control charts are a po*erful tool to hold the gainsJ- 7o* control charts discriminate !et*een common cause and assigna!le cause variationB- 'h" control charts must !e designed to fit the data t"pe and the control purpose-
Process $ariation
Process variation is the result of2 ) Common causes) Special ?assigna!le@ causes
Common Causes
) 5esult in normal process variation) Are specific to each process) Can !e reduced !" changing the process-
Special ?Assigna!le@ Causes

) Are attri!uted to something outside of the process) 5esult in a!normal process variation) Do not result in process improvement if eliminated-
(ses of Control Charts

19 Attain a state o% statistical control All su!group averages and ranges *ithin control limits = no assigna!le causes of variation present !9 Monitor a process "9 Determine process capability
:uran;s $uality Control <andboo%" =th edition" page >=*?
What happens after an out-of-control situation occurs at the core of a successful SPC program?
9eneral Concepts
* F some characteristic of interest F mean of each sample
=W
S* F standard deviation of * (pper Control #imit UCL = X + 0S w Centerline F
X
LCL = X 0Sw
#o*er Control #imit
Therefore RR-OBS of points *ill !e *ithin the control limits unless there is an assigna!le cause
Control Chart Selection Tree

T"pe of data
Discrete $aria!le
Count
Count or Classification
Attri!ute
%o
Su!group _D8
Ges
Fixed
Fixed or varia!le opportunit"8
$aria!le
Fixed
Fixed or varia!le opportunit"8
$aria!le
IM5 Chart
K Aar and 5 or K Aar and S
C Chart
( Chart
%P Chart
P Chart
Supplement *ith 'MA if CTQ is sensitive to small process shifts
K and 5 Control Chart Formulae & Constants

= -ontrol +i#its L= Q / 0 0 B&&er -ontrol +i#it L 4 0 4 0 +o(er -ontrol +i#it L 4 0 ?
'a#&le 'iHe ? 4 5 6 ! 8 9 10 / 19880 190 ? 9! 9 95!! 948? 9419 9?!? 9??! 9?08 4? 90!6 91?6 9184 9 ? 44 ?9 6! 95!4 9 8 9114 9004 199 4 19864 19816 19!!! d 191 8 1969? 9059 9? 6 95?4 9!04 984! 99!0 ?90!8
Creating Control Charts for $aria!les

(Data)ile(&ariable*mt-
Creating an K=!ar and 5 Chart
An K=Aar and 5 Chart

Xbar R C!art of meas're1( )))( meas're*
# # # :
9# Sample Mean 9% 0<

# / # # / #
UC269#.8<9 I I 469%.%%%
2C260$.>%/
0$ : #%
#:
8%
8: Sample
0%
0:
9%
9:
9.: Sample Range
UC269.>90
0.%
I J68.890
#.:
%.% : #% #: 8% 8: Sample 0% 0: 9% 9:
2C26%
Wor sheetA @aria&le.M,W
The num!ers sho* violations of the assumption of control- The nature of the violation is given in the session *indo*-
K=Aar and 5 Chart Session 'indo*

Test Results for Xbar Chart of measure1, ..., measure5 TEST 1. One point more than 3.00 standard deviations from Test !ailed at points" #, 10, 13, 1$, 1#, %&, '$ enter line (on enter line.
TEST 5. % out of 3 points more than % standard deviations from one side of C)*. Test !ailed at points" 10, 1#, '#
TEST $. ' out of 5 points more than 1 standard deviation from one side of C)*. Test !ailed at points" #, 3%, 3'
enter line (on
+ ,-R./.0 + /f 1raph is updated 2ith ne2 data, the results above ma3 no + lon1er be orre t.
Stat9uide Interprets the Tests
Comparing the Suppliers

Sample Mean 9# 9% 0< 8
#

Sample Mean UC269#.8%> I I 460<.$<9 2C260$.:$% 98
# # # : :
9%
: / : #
UC269#.0$9 I 469%.#%/ 2C260$.$8<
0$ 8 9
#%
#8 #9 Sample
#/
#$
8%
88
89
#%
#8 #9 Sample
#/
#$
8%
88
89
Sample Range
Sample Range
UC269.$#/ 9 8 % 8 9 / $ #% #8 #9 #/ #$ 8% 88 89 Sample I J68.8>$ 2C26%
UC269./$0 9 8 % 8 9 / $ #% #8 #9 #/ #$ 8% 88 89 Sample I J68.8#9 2C26%
Wor sheetA @aria&le.M,WC!upplier 6 #D
Wor sheetA @aria&le.M,WC!upplier 6 8D
Supplier J/s process is less sta!le than Supplier D/s process-
Add Dates to Control Charts
Adding dates to the control charts may help identi%y potential sources o% shi%ts&
Adding Dates Indicates 'hen the Shift 4ccurred

Xbar R C!art of Stac+e$ Data 1s
#8
: / 8 8
Sam ple Mean
##
#%
/ /
< >/0#/8%%/ $/9/8%%/ $/#%/8%%/ $/#//8%%/ $/88/8%%/ $/8$/8%%/ </#/8%%/ Date
</>/8%%/ </#0/8%%/ </#<
9.: Sam ple Range
0.%
#.:
Detecting 9radual Process Drift

#8
: / 8 8 8 8
UC26##.$:%
8 8
Sample Mean
##
I I 46#%.:/$
#%
/ /
2C26<.8$/ < 9 $ #8 #/ 8% Sample 89 8$ 08 0/ 9%
9.: Sample Range
UC269./<<
0.%
I J68.888
#.:
%.% 9 $ #8 #/ 8% Sample 89 8$ 08 0/ 9%
2C26%
The tests detect the process shift !ut it *ould have !een detected earlier if the control limits had !een !ased on the first JE data points-
Detecting 9radual Process Drift

Minitab calculates control limits based on the entire data set be de%ault& Eree(ing control limits detects shi%ts earlier than using the entire data set to calculate limits&
#8
: # / / # 8 8 8 8 8 8 8 8 8
UC26##.0/%
Sample Mean
##
#%
I I 46#%.##8
< 9 $ #8 #/ 8% Sample 89 8$ 08 0/ 9%
2C26$.$/9
9.: Sample Range
UC269.:>:
0.% I J68.#/9 #.:
%.% 9 $ #8 #/ 8% Sample 89 8$ 08 0/ 9%
2C26%
5ational Su!grouping
(suall" consecutive units Must come from a single distinct population 'ithin su!group3 variation should !e *hite noise onl" Aet*een su!groups should capture variation due to !lac+ noise D@ Su!group needs to represent the distinct population J@ sta!lish minimum su!group si6e to reflect the *ithin variation B@ sta!lish sample fre,uenc" to capture the !et*een variation M@ Collect data maintaining the se,uential information
1ever %no-ingly subgroup unli%e things together Minimize variation -ithin each subgroup Ma!imize opportunity for variation bet-een subgroups #verage across noise not across signals Treat charts in accordance -ith the use of the data +stablish standard sampling procedures
.onald ;* Wheeler has si7 guiding "rinci"les for subgrou"ing in a rational manner*
'hen to Increase Sample Si6e

If the sample si6e is too small3 assigna!le causes ma" produce real effects that are relativel" small and unimportant- In this case3 it ma" not !e economical to ta+e actionTo minimi6e the num!er of these 0nuisance1 causes3 the sample si6e should !e increased using the follo*ing techni,ues2 Identif" characteristics Determine logical nature of su!groups Identif" sampling se,uence Measurement methods proven to !e accurate
'hat 'e 7ave #earnedD- Control charts are a po*erful tool to hold the gains J- 7o* control charts discriminate !et*een common cause and assigna!le cause variation B- 'h" control charts must !e designed to fit the data t"pe and the control purpose
#ean nterprise
A 5ormula for Organi<ational Success
Agenda
D- Introduction to #ean3 'astes J- NS 'P4 & $isual Management B- Standard 'or+ M- $alue Stream Mapping N- Quic+ changeover P- Po+a="o+e O- Continuous Improvement Q- <ai6en Alit6 R- Starting #ean
Introduction to #ean and O 'astes
) Identif" and liminate the O 'astes
Arief 7istor" of #ean

) Craftsman to mass production ) Mass production to lean production ) #ean production2 Ford to To"ota
; Frederic+ Ta"lor ; Taiichi 4hno ; Shigeo Shingo ; .ames 'omac+
) From shop floor to office and support functions then to service industr"
Craft Manufacturing
=ate 1IFFOs Car !uilt on !loc+s as *or+ers *al+ed around Auilt !" craftsmen *ith pride Components hand=crafted3 hand=fitted xcellent ,ualit" $er" expensive Fe* produced
G G G G G G
Mass Manufacturing
) Assembly line ? >enry Eord 1K!Fs
G G G G G
#o* s+illed la!or3 simplistic jo!s3 no pride in *or+ Interchangea!le parts #o*er ,ualit" Afforda!l" priced for the average famil" Aillions produced ; identical Model P+O
#ean 7istor" ; .apan

Sa+ichi To"oda at his textile mills ?JE/s;BE/s@ To"ota Motor Compan"/s <iichiro To"oda and Taiichi 4hno made innovations ?ME/s@ in assem!l" lines that provided efficient3 customer=focused3 streamlined processes *ith flo*3 variet"3 and short lead timeU To"ota Production S"stem ?TPS@^ to face competition of 9M and Ford after ''II *hen +e" need *as flexi!ilit" Taiichi 4hno and Shigeo Shingo developed #ean !ased on TPS
#ean Manufacturing
G G
Cells or flexi!le assem!l" lines Aroader jo!s3 highl" s+illed *or+ers3 proud of product Interchangea!le parts3 even more variet" xcellent ,ualit" mandator" Costs !eing decreased through process improvement 9lo!al mar+ets and competition-
G G G
'hat is #ean8
0A !usiness s"stem for organi6ing and managing product development3 operations3 suppliers3 and customer relations that re,uires less human effort3 less space3 less capital and less time to ma+e products *ith fe*er defects to precise customer desires compared *ith the previous s"stem of mass production-1
Lean Le7icon3 #ean nterprise Institute3 JEEB
De%inition
#ean manufactuirng is aimed at the elimination of *aste in ever" area of production including customer relations3 product design3 supplier net*or+s and factor" managementThe goal of #ean Manufacturing is to incorporate less human effort3 less inventor"3 less time to develop products3 and less space to !ecome highl" responsive to customer demand3 *hile at the same time producing top ,ualit" products in the most efficient and economical manner-
'h" Call it 0#ean18

) +he term QleanR is used because lean uses QlessRS
#a!or Space Capital investment Materials Time !et*een the customer order and the product shipment
Definition of #ean

#ean has !een defined in man" different *a"s ; 0A systematic a""roach to identifying and eliminating waste ?non=value=added activities@ through continuous im"ro8ement !" flowing the "roduct at the "ull of the customer in "ursuit of "erfection-1
Definition !" the M P #ean %et*or+
9ive the customers *hat the" *ant3 *hen the" *ant it3 and do not *aste an"thing-
Definition of $alue Added

'aste is an" activit" that does not add value to the final product for the customer) Dalue?added is an activit" that transforms or shapes ra* material or information to meet customer re,uirements) 6on?value added is an activit" that ta+es time3 resources or space3 !ut does not add to the value of the product or service itself) 6on?value?adding@ but necessary ; does not add value to the product or service !ut is re,uired ?e-g-3 accounting3 governmental regulations3 etc-@
'aste
0An"thing that adds Cost to the product *ithout adding $alue1
To"ota 'a" ; the DM Principles !" .effre" <- #i+er

D- Aase management decisions on long=term philosoph"3 even JBMNPOQ-
at expense of short=term financial goalsCreate continuous process flo* to !ring pro!lems to surface(se pull s"stems to avoid overproduction#evel out *or+load ?heijun+a@^ *or+ li+e a tortoise3 not a hareAuild culture of stopping to fix pro!lems3 to get ,ualit" right first timeStandardi6ed tas+s are the foundation for continuous improvement and emplo"ee empo*erment(se visual controls so no pro!lems are hidden(se onl" relia!le3 thoroughl" tested technolog" that serves people and processes-
To"ota 'a" = DM Principles ?cont-@

R- 9ro* leaders *ho thoroughl" understand *or+3 live philosoph"3
and teach othersDE-Develop exceptional people and teams that follo* compan"/s
philosoph"DD-5espect extended net*or+ of partners and suppliers !"
challenging them and helping them improveDJ-9o and see to thoroughl" understand situation ?genchi
gen!utsu@DB-Ma+e decisions slo*l" !" consensus3 thoroughl" considering
all options^ implement decisions rapidl"DM-Aecome a learning organi6ation through relentless reflection
?hansei@ and continuous improvement ?+ai6en@-
#i+er Model
To"ota Terms 9enchi 9en!utsu Pro!lem Solving People and Partners 5espect and Team*or+ <ai6en
Process Challenge
Philosoph"
#i+er Model = M Ps
D- Philosoph"
#ong=term thin+ing even at expense of short=term financial goals J- Process liminate *aste !" focusing on flo*3 pull3 *or+load !alance3 error reduction3 standardi6ation3 visual controls3 and jido+a or use of relia!le3 tested technolog"Cautomation *ith mista+e proofing and human touch B- People and Partners 5espect3 develop3 and challenge people M- Pro!lem Solving Fix and prevent pro!lems !" continual learning3 going to place *here pro!lem occurs3 getting hands dirt"3 and ma+ing good decisions !ased on fact
'hat Is #ean8
#ean is a methodolog"
that allo*s organi6ations to drasticall" improve !ottom line !" improving processes and monitoring ever"da" !usiness activities to reduce errors in *a"s that increase value and minimi6e *or+3 non= value=add tas+s3 and *aste *hile increasing customer satisfaction
Aased on idea that faster processes "ield less *aste3 less cost3 less *or+ in process3 less complexit"3 higher ,ualit"3 and happier customers
'omac+ and .ones Model

D-Define value from customers/ perspective J-Document value stream B-Improve flo* of value stream M-Drive for pull versus push N-Continuousl" improve
#ean Principles
Create value for customer (nderstand2 'ho is customer ; person or entit" *ho is recipient of product or service^ one *ho places value on output^ catal"stCtrigger in value chain
Another !usiness Someone inside o*n !usiness Specific individual3 group3 or team ConsumersUultimate customer
'hat customer considers valua!le Ma+e value flo* !ased on customer/s needs
#ean Principles ?cont-@

Consider life c"cle of information3 materials3 processes3 products3 and services #oo+ for process pro!lems that prevent people from performing !est *or+ liminate *aste
%on=value=add steps 'IP Cost

Standardi6e *or+ Do not !ecome distracted !" other sta+eholders
$alue=Add Qui6
In *hich categor" should the follo*ing !e placed8

Activit" $alue Add Attending *ee+l" team coordination meeting Filtering through dail" e=mail list 5eporting status to upper management 9aining multiple approvals on documents 9aining management approval for routine actions xpediting document through approval list 'riting formal policies and procedures 'riting !rief *or+=method instructions 9aining regulator" or agenc" approvals Creating IS4 REEE documentation 7unting for needed information to do "our jo! Auilding 0!est practices1 data!ase 7olding lessons learned meeting Spending time on process improvements T"pe D T"pe J
#ean vs- Traditional
=ean
Simple and visual signals Demand driven Inventor" as needed 5educe non=value added Small lot si6e Minimal lead time Qualit" !uilt $alue stream managers
+raditional
) ) ) ) ) ) ) )
) Complex ) Forecast driven ) xcessive inventor" ) Speed up value=added *or+ ) Aatch production ) #ong lead time ) Inspected=in ) Functional departments
Aenefits of #ean Manufacturing

7elps in ; ) Cost reduction #ead Time 5eduction ) C"cle time reduction Productivit" Increase ) 0'aste1 minimi6ation 'IP 5eduction ) limination of non= Qualit" value=added Improvement activities Space (tili6ation ) 5esulting in a more 0lean31 competitive3 agile3 and mar+et= responsive
5eal 5esults E NE DEE
'h" the mphasis on #ean %o*8

) ) ) ) ) ) 9lo!al econom" Pressure from customers for price reduction Fast=paced technological changes ?e-g- Internet auctions@ Continued focus on ,ualit"3 cost3 deliver" 7igher and higher expectations of customers Qualit" standards3 such as QS=REEE ?or TS DPRMR@3 the ne* IS4 REEE2JEEE ) 7olding on to 0Core Competencies31outsourcing the rest ) Mar+et=driven pricing2 Customers expect !etter performance at lo*er prices "ear after "ear
volution of #ean Across Mar+ets

Proven glo!al concept since DRQEs Transformed !usiness processes across man" industries2 Automotive Aerospace 4ther industries !eginning to em!race #ean concepts *ith excellent results2 Construction 7ospitals Pharmaceutical Manufacturing Service 4rgani6ations
Pricing Model
4ld 'a"
Costs Increase
Maintain Profits
Increase Price
Cost H Profit F Price

Customers Demand #o*er Prices If Costs Sta" the Same Profits Decrease
What can -e doB
%e* 'a"
Price = Cost F Profit
Pricing Model
4ld 'a"
Costs Increase Maintain Profits Increase Price
Cost H Profit F Price

Customers Demand #o*er Prices Implement #ean Increased Profits
%e* 'a"
Price = Cost F Profit
Core Concepts of #ean

) Creativit" !efore Capital ) A solution that is not=so=perfect implemented toda"3 is !etter than a perfect solution that is late- 0.ust do it-1 ) Inventor" is not an asset3 !ut a *asteCcost) T"picall"3 RNS of lead=time is not value added) #ean implementation using the Plan=Do=Chec+= Act methodolog" ) Continuous Improvement environment2 !oth incremental and !rea+through) #ean is a never=ending philosoph"-
Aluminum Can xample

From aluminum ore to usa!le cans3 it t"picall" ta+es a!out BEE da"s
Aauxite Cr"olite Aluminum Cast Product
B hours
Cans Sheet 5olled Plate
9uess *hat the total value=added time is8
$ideo
) Introduction to #ean
O 'astes of #ean
04MIT 'hat ( D41
) ) ) ) ) ) ) )
4verproduction Motion Inventor" Transportation ?Movement@ 'aiting Defects ?Correction@ 4ver=processing $nderutili<ed 'eo"le
C 3 M M WI P
4verproduction
Ma+ing more=earlier=faster than the next process needs it

.ust in case logic (n!alanced *or+load (nleveled scheduling False sense of efficienc" Printing JE copies of a report that onl" B people loo+ at All=staff e=mails *hen it pertains to onl" a fe* 'aiting to 0!atch1 *or+
Motion An" movement of people that does not add an" value to the product or service
Poor la"out Inefficient 'or+place 4rgani6ation #ac+ of Standardi6ation3 inconsistent *or+ methods People3 Material and Machine Ineffectiveness
'here is are copier3 printer3 files and coffee=ma+er located8 7o* far does the paper*or+ travel8
Inventor"
An" suppl" in excess of one=piece flo*

.ust in case logic (n!alanced *or+load (nleveled scheduling (nrelia!le suppliers 5e*ard s"stem 0Pac+ rat1 mentalit" Printed forms or tags that !ecome o!solete Q *ee+s of paper located !" the copier 7o* man" pens do "ou have in "our des+ dra*er8
Transportation
Moving people3 materials and information around the organi6ation

Poor la"out Inefficient 0flo*1 Carr"ing large ,uantities Moving 0!an+er/s !oxes to a storage area Mail carts Messenger services
'aiting
'aiting forI man3 machine3 materials3 information etc
.ust in case logic (n!alanced *or+load (nleveled scheduling (nplanned do*ntime %eeds not understood
'aiting for files or information %eed a signature Customer repl" to a voice=mail or e= mail Someone is printing NE copies of a OE page report
Defects
Information3 products and service that need correction

%ot using .ido+a or Po+a="o+e #ac+ of Standardi6ation3 inconsistent *or+ methods Ineffective communication #ittle investment in training 7ave to fix paper*or+ that is not completel" filled in or trac+ do*n the right person to get the information An entr" error causes the *rong actions li+e shipping too man"3 or too fe* to the *rong address3 etc-
4ver=processing ffort that adds no value to the product or service from the customer/s standpoint
.ust in case logic Inconsistent *or+ methods Ineffective communication 5edundant approvals xcessive information3 extra copies
Multiple sign=offs or chec+s
(nderutili6ed People
%ot utili6ing people/s experience3 s+ills3 +no*ledge3 creativit"

%ot utili6ing Teams 4rgani6ation structure Poor hiring practices #ittle investment in training #ac+ of suggestions 0That/s not m" jo!1 attitude 'aiting for lead from management
Mura and Muri

Mura ?unevenness@ ; variation in operation3 *asted resources *hen ,ualit"3 cost3 or deliver" cannot !e predicted
TestingCinspection3 Containment3 5e*or+3 5eturns3 4vertime3 (nscheduled travel

Muri ?overdoing@ ; unnecessar" or unreasona!le over!urdening of people3 e,uipment3 or s"stems *hen demand exceeds capacit" or tas+s are not designed properl" including harmful3 *asteful3 or unnecessar" tas+s
Auilding Aloc+s of #ean

Continuous Improvement & <ai6en Alit6 Cellular & Pull S"stem & TPM .IT Flo* <an!an Po+a="o+e P4(S #a"out Self Inspection Aatch Si6e 5eduction Standard 'or+ Autonomation $ S M
Quic+ Changeover $isual NS Teams
Change Management

Change Management
N 'ords that !egin *ith 0S1

.apanese Seiri Seiton Seison Sei+etsu Shitsu+e Translation 4rgani6ation %eatness Conversion Sort Set in order [4ther Sorting Simplif"ing access S*eeping Standardi6e Self= discipline
Cleaning Shine Standardi6ation Standardi6e Discipline Sustain
[ There are several other conversions
5 Steps to Workplace Organization

1 Clear1Sort A" red tagging
$ Continuous Improvement Sustain Discipline
"aste
! 3rgani(e1Straighten A place for ever"thing
# Maintain1Standardi(e sta!lish standards
" Clean1S-eep 7ouse+eeping C Inspection
'or+place Scan
=$nderstand your Current State
) Start *ith a *or+place Scan ) Team Aased ) Define the !oundaries ) Complete a Diagnostic Chec+list ) Dra* a Spaghetti Diagram ) Ta+e 0Aefore1 Photos ) Starts the NS Program
'or+place Scan Displa" Area Details Chec+li st Score
Spaghetti Diagram
Aefore Photos
After Photos
Sort
=When in doubt9 mo8e it out
) Move unneeded items out of the area ) (se the 5ed Tag Techni,ue ) (se a Temporar" 5ed Tag 7olding Area ) Criteria for unneeded items
; 0BE=da" 5ule1
6ameTTTT DateTTT Item TTTTTTTTTTTTT 'eason TTTTTTTTT
) <eep onl" *hat "ou need in the area
Set in 4rder
=A "lace for e8erything and
e8erything in its "lace*
) Ma+e it eas" for an"one to find

; 0BE=second 5ule1
) Ma+e it o!vious if an item is out of place ) Decide *here to +eep items3 ho* man" items to +eep3 ho* and *hen to replenish items ) Ma+e it $isual
Shine
=Clean and ns"ect
) 9et items to a li+e=ne* condition

; 0DE Second 5ule1
) Must plan Shine ; assignments & supplies ) Perform as a Team ) Prevent dirt3 grime3 or contamination ) 5epair as needed
Standardi6e
=Create the rules and follow them
) Determine ho* the first BS conditions are met ) (se 04ne=Point #essons1 ) Maintain and monitor the conditions ) (se $isual techni,ues
Sustain
=6ake 3S a habit
) NS is not something additional3 it is part of ever"one/s dail" jo! ) Supports discipline ) Train ) Communicate ) Support from Management ) 5e*ard and recognition
NS is Fundamental to #ean
) NS is directl" related to other Auilding Aloc+s
; Teams ; $isual ; P4(S ; Standard 'or+ ; TPM
Point=4f=(se Storage
) 5a* material and 'IP are stored at *or+station *here used3 *hich reduces the inventor" that can !e carried) 'or+s !est if vendor relationship permits fre,uent3 on=time3 small shipments ?.IT@) Simplifies ph"sical inventor" trac+ing3 storage3 and handling-
Carpenter Stor"
) Does the carpenter *al+ !ac+ to the tool!ox ever" time a tool is needed8 ) 'hich *aste is this8 ) 'hat does the carpenter do8
Proximit"
) T"picall"3 up to PES of time is spent on findingCcollecting items needed) Minimi6e non=value activities ) Store as close as possi!le and *ithin reach ) #a"out and *or+station design should accommodate re,uired materials ) Tr" to use the pac+aging from the supplier or have the supplier change pac+aging
P4(S Components
) 7ave the2
; Information ; Parts & materials ; Tools & e,uipment
that "ou need to perform "ou tas+s *ithin reach
P4(S 'or+place >ones

) Items used most often ?i-e-3 dail"@ should !e +ept *ithin reach ) Items used less often ?i-e-3 *ee+l"@ should !e +ept close=!" ) Items used rarel" ?i-e-3 monthl"@ should !e +ept in the vicinit"
Dail" 'ee+l" Monthl"
#ocation of items
) (se hori6ontal transfers and gravit" feeds *hen possi!le ) Support heav" o!jects ) Set items ergonomicall" ) Must !e comforta!le for a da"/s *or+
Aenefits
) Supports NS & $isual and other Auilding Aloc+s ) Simplifies inventor" trac+ing and accurac" ) 5educes *aiting3 inventor"3 motion and transportation *aste
Aenefits of NS
) Improved e,uipment relia!ilit" ) Superior ,ualit" ) Increased productivit" ) Aetter *or+flo* ) nhanced Safet" ) 5educed inventor" ) More pleasant place to *or+ ) Impress customers
$ideo
) Introduction to NS

Change Management
$isual
Signs3 lines3 la!els and color coding
'h" $isual8
) 'hat "ou need to +no* ) Coc+pit vie* ) Information sharing

7o* do "ou +no* *here to par+ *hen "ou drive to a shopping mall8 Does someone have to tell "ou *here to par+8
7o* to Appl" $isual

) (se Signs3 #ines3 #a!els and Color= coding ) Charts3 pictures3 lights3 score!oards ) <an!an3 Andon lights ) Inventor" #evels
xamples
) Productivit" 9oals ) Qualit" 9oals ) Deliver" schedules ) Set=up specification ) Safet" Initiatives ) Attendance 9oals ) Team 4!jectives
More xamples of $isual

) Signs ) Charts ) 9oals ) Pictures ) Color coding ) #ights ) Score!oards ) $SM Current & Future State ) Standard 'or+ instructions ) ) ) ) ) ) ) ) Tags Forms Training hours mplo"ees/ suggestions Cross trained s+ills mplo"ee a*ards A!senteeism Critical maintenance points ) Customer satisfaction goals ) Performance targets
$isual Controls as Communication Tools

) $isual controls expose *aste so *e can reduce or eliminate it ) $isual controls help in2
; Improving motivation & morale ; Focus on safet" ; Pride of *or+place & *or+manship
) $isuals and performance metrics

; 'hat gets measured3 gets done ; Policies drive !ehaviors
'orld Class $isual Controls

) An"one +no*s *hat/s going on !" loo+ing around ) Gou do not have to *ait for information to do "our jo! )
ver"one passes the BE second test
$isual xamples
Andon #ights Displa" Panels
5ange Mar+ings 4n 9auges Shado* Aoards
Standard 'or+
5educe tas+ varia!ilit"
Standard 'or+
Aest se,uence of operations3 using the most productive com!ination of resources2
) Man3 machine3 materials3 changeovers3 etc) Details an" special s+illC+nac+ needed ) Safet"3 ergonomics are integrated ) Tool for perfect ,ualit" and efficienc"
Standard 'or+ Properties

) Specific ) Measura!le ) 5epeata!le ) Documented
Standard 'or+
Identifies value added versus non=value added activities ) 5educe or eliminate non=value added activities ) Convert Internal Time to xternal Time3 *herever possi!le ) Continuous Improvement2 once Standard 'or+ is esta!lished as a !ase and displa"ed at *or+stations3 operators monitor and implement improvements ) (se as a training tool for ne* emplo"ees ) Created !" input from the people *ho actuall" *or+ in the process
Philosoph" of Standard 'or+

Ae specific a!out2 xample2 Installing a Car Seat ) Content ) Aolts are installed and tightened in the same ) Se,uence exact order ) Timing ) The time re,uired to tighten ) 4utcome !olts is stated and follo*ed ) Specified tor,ue is applied and chec+ed (sed as the !asis from *here the next level of improvement is made
T"pes of Standardi6ed 'or+ Forms

) Process Capacit" Ta!le ) 'or+ Com!ination Sheet ) Standard 'or+ Sheet ) 4thers forms ma" !e used !ased on "our organi6ation/s needs
Process Capacit" Ta!le

Times for2 ) lapsed time ) lement time ) Internal Time ) xternal Time ) Manual Time Step Identification ) Transportation ) Process ) Inspection ) Storage (se to identif" !ottlenec+s (se to calculate the capacit" of machines3 and identif" !ottlenec+s
Changeover Analisys Chart

Date2 Process2 Se, %olement S"m!ol lapsed Time %ame2 lement Internal Time Time xternal Time Comments
Review work order for die number and material Locate the die Get tools Locate material a!out machine
600
600
300 (onvert to -.ternal 300 (onvert to -.ternal 300 (onvert to -.ternal 300 (onvert to -.ternal /se 0uick connects /se 0uick connects )osition die cart
2 3 4 5 6 # 8 "
1800
1200 1200 600 "00 480 300 300 600 300 480 480 300 600 600 60 600 60
3000 3600 4500 4"80 5280 5580 6480 6"60 #440 ##40 8340 8"40 "000 "600
60 60 60
)re%are a!s before
Loosen bolts
$isconnect hoses
Remove die from %ress Load die into %ress &li!n die i!hten bolts
120 "0 60 60 60 60
Return die to ool Room
6180
300 (onvert to -.ternal /se %ositive sto%s 0 -liminate
10 11 12 13 14 15 16 1# 18 1"
/se 0uick connects /se 0uick connects
(onnect hoses (lear ta!out
)osition material
300 (onvert to -.ternal
*ake sam%le %iece
ake first %iece sam%le to +(
120
1ave +(
ech read2
&d,ust die and %ress *ake sam%le %iece
10800 10860 11460
1200 600
0 )re3ad,ust die 0 -liminate
20 Return tools
Transport
600 (onvert to -.ternal
Process
Inspection
Storage
Total Time
11460
#50
2400
Page YYYY of YYYYY
%$A F K
Shift2
'
' '
'or+ Com!ination Sheet

Se,uence of2 ) Manual *or+ time ) Machine operations time ) 'al+ing

Pr oduct Eam ily A = 9enerators Proces s A ssem!l" Cell WJB De s cription Final Assem!l" Cyclical Work 2le m e nt ID D J B M N P O Q R DE DD DJ DB DM DN DP DO DQ DR JE JD JJ JB JM JN JP JO JQ JR BE Description of 'or+ lement Manuall" load 'C=D Start 'C=D3 go to 'C=J (nload part3 #oad Part3 Start Machine3 go to 'C=B (nload part3 f ile c orner3 inspect to print #oad nex t part3 s tart machine3 go to 'C=D Fas ten parts A=D3 and A=D together3 go to A S=J Fas ten 7ousing and Aase Pac +age generator #oad on c art +im e Available JM3PEE seconds piec es seconds per piece Date .anuar" BD3 JEEK 6am e ileen %- Terpris e D S M gr I-C- Flo* 9raph Data Manual 4peration Machine 4peration 'al+ing DJ BJ B E D DP MN DJE J DN JB MD MJ E BN E N E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E BEE MEE NEE De m and JEE +ak t +im e DJB lement Time Manual Machine 4peration 4peration 'al+ing DJ BJ E B E J D DP J MN DJE E J DN B JB MD J MJ E E BN E E N E M E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E DPQ JJM DB BS MEN 5ate Total Time ?C"clical H %on=c"clical@2 Throughput per shif t F A vaila!le Time C Total 4perating Time2 'or+ Flo* Diagram MEN MDS NNS Total C"clic Time2 Seconds Time PCS
3peration +imes
Start D J B M N P O Q R DE DD DJ DB DM DN DP DO DQ DR JE JD JJ JB JM JN JP JO JQ JR BE E MM N DR DPN JE PP MJ BN R E E E E E E E E E E E E E E E E E E E E E DEE E MM MR PQ JBB JNB BDR BPD BRP MEN MEN MEN MEN MEN MEN MEN MEN MEN MEN MEN MEN MEN MEN MEN MEN MEN MEN MEN MEN MEN E MR PQ JBB JNB BDR BPD BRP MEN MEN MEN MEN MEN MEN MEN MEN MEN MEN MEN MEN MEN MEN MEN MEN MEN MEN MEN MEN MEN MEN JEE
Sho*s the interactions !et*een machines and operators Allo*s to recalculate operator *or+ content as ta+t time changes
E J J E B J E E M E E E E E E E E E E E E E E E E E E E E E

D J B M N
Total Time2 Percent 4perator Time2 6on?cyclical Work 2le m e nts
'C=D
'C=J
'C=B
AS=D
Total %on=c"c lic Time2
AS=D
AS=D
Standard 'or+ Sheet

Se,uence of processing steps ) 'or+er ) Machine ) Tools ) #a"out ) Material location ?Standard stoc+@

Standard Wo
Product E
Displa"ed at 'or+stations Continuousl" revie*ed and updated
A = 9ene
Proce
(ses
) The Process Capacity Table can !e used to reduce changeover times ) The Work Combination Sheet can !e used for line !alancing *hen creating a cell ) The Standard Work Sheet can !e used for training and team development
Standard 'or+ xamples
Ado&e Acro&at )ocument
Standard 'or+ and Training

Questions to as+ 4perators a!out Standard 'or+ ) 7o* do "ou do this *or+8 ) 7o* do "ou +no* "ou are doing this *or+ correctl"8 ) 7o* do "ou +no* there are no defects8 ) 'hat do "ou do if "ou have a pro!lem8 If these ,uestions cannot !e ans*ered satisfactoril"3 then either the 4perator needs additional training or the Standard 'or+ is unclear
(ses of Standard 'or+

) Consistent performance of tas+s F !etter ,ualit" ) Trac+ performance F actual versus standard for continuous improvement ) as" to Train F reduced learning c"cle time
Old Learning Cur8e #ew Learning Cur8e
Productivit"
Time
Productivit"
Time
Aenefits of Standard 'or+

) Standard documentation for all shifts ) 5eductions in injuries and strain ) mplo"ee o*nership of process ) More pleasant *or+ing conditions^ higher morale ) Aetter than traditional time and motion studies ) 5educed varia!ilit"
Po+a="o+e
) ) ) ) ) ) rror proof ?mista+e proof@ ta+es a*a" the possi!ilit" of human error The term Po+a="o+e *as made popular !" Shigeo Shingo Fail=safe devices #o* cost3 highl" relia!le mechanisms Detects a!normal situations !efore the" occur3 or 4nce the" occur3 *ill stop the e,uipment from further production- The machine stoppage ma+es the pro!lem visi!le-
4ther Po+a="o+e xamples

) (SA ports on computers ) 5e=t"ping pass*ords to verif" ) Computer prompts !efore deleting file ) Aar codes & scanning ) ATM s*ipe card or !eep
Aenefits of Po+a="o+e
) 9ives immediate feed!ac+ for root cause anal"sis & correction ?and prevention for the future@) Failure Mode and ffect Anal"sis ?FM A@ solution can !e Po+a="o+e ) Some examples of Po+a="o+e devices are2 sensors3 counters3 feelers3 limit s*itches3 electric e"es3 pro!es3 automatic stops-

Change Management
$alue Stream Mapping
See the Flo*
$alue Stream Anal"sis
$alue stream anal"sis encompasses all activities compan" must do to design3 order3 produce3 and deliver its products and services to customers Flo* of tas+s3 from re,uest for service ?trigger event@ to service complete3 from receipt of materials or information from suppliers to deliver" of finished product or service to customers From vie*point of customer3 service3 or transaction Flo* of information that supports and directs !oth flo* of materials and transformation of ra* materials or information into finished goods or services
Purchasing 7uman 5esources 4perations Finance Sales
$alue Stream $alue Stream
Purpose of $alue Stream Map

7as customers/ perspective and focuses on meeting customers/ *ants and needs Starts *ith immediate customer and maps !ac+ to receiving inputs from suppliers and sho*s ho* fits into overall value stream Provides single vie* that is a complete3 fact=!ased3 and time=!ased representation of stream of activities Provides common language and vie* for anal"sis Sho*s ho* information triggers and supports activities Sho*s time for activities and *hether the" add value
lements of $alue Stream Map

Process steps $alue=add classification for each step Information flo* such as orders3 re,uirements3 schedules3 messages3 approvals3 specifications3 +an!an signals3 shipping information3 standard procedures Aox score of +e" operational metrics including c"cle time3 *aiting time3 *or+ing time3 conve"ance time3 distance traveled3 items per shift3 items processed per hour3setup time3 !ac+logC*or+= in=process3 amount of inventor" !et*een last step and consumer3 defects3 cost information3 resource availa!ilit" and active time3 process variations #ead time is amount of time for one item to flo* completel" through process3 noted along !ottom of flo* Ta+t time sho*ing customer demand rate3 in upper right corner of flo*
$alue Stream Anal"sis Steps

Identif" delivera!le3 value stream3 and sponsor *ho has authorit" and responsi!ilit" to allocate resources and ma+e changes across organi6ation Identif" customer and value from customer/s perspective as *ell as regulator"3 legal3 and compliance re,uirements Dra* visual representation of process current state3 generall" dra* steps starting at consumer/s vie* *or+ing !ac+ through steps to sources of material and la!or^ flo*s from left to right *ith time3 *ith steps in order of occurrence Add metrics and o!servations li+e Ta+t timeCthroughput3 c"cle times3 defect rates3 and inventor"C*or+=in=process and information flo*s to identif" magnitude and fre,uenc" of *aste (se lean principles to reduce or eliminate *aste and reduce c"cle time Develop future state map3 document steps of process that need to happen3 and prioriti6e and implement action plans to achieve future state
$alue Anal"sis Matrix Steps

Structure value anal"sis matrix %um!er process steps on su! process map 7ave column for each process step stimate time for each process step Place chec+ in categor" for each process step3 either value=add or one of non=value=add categories Total num!er of hours or num!er of chec+s for each ro* 5eport percentages of value=add and non= customer re,uired
$alue Anal"sis Matrix xample

Process Step Time ?7ours@ $alue = Add Added %on Customer 5e,uired Internal Failure xternal Failure ControlCInspection Dela" PrepCSet PrepCSet=(p = (p Move
D D
JJ
B B
MM
NN
PP
OO
QQ
RR
DE DE
Total
S Total
DJ DE DE D D DE DE JE JE P P DE DE D D DE DE JE JE DJ
DEE DEE JJ
DEES DEES JS JS
DE DE PP NJ NJ
DES DES PS PS NJS NJS
BE BE
BES BES
Total
DEE DEE
DEES DEES
5FQ Creation $alue Stream Map

Customer Meetings CCT F DM da"s 'CT F J da"s $ACT F D da"
Iterate
Assign Au"er CCT F B da"s 'CT F M hours $ACT F T E 9ather 5e,uirements CCT F DM da"s 'CT F J da"s $ACT F D da" $erif" Customer 5e,uirements CCT F DM da"s 'CT F J da"s $ACT F D da"s Consult *ith Manu= facturing ngineer CCT F N da"s 'CT F J da"s $ACT F M hours Create Preliminar" 5FQ CCT F N da"s 'CT F J da"s $ACT F D da"
5evise Triggering vent

5evie* and Approval C"cle CCT F N da"s 'CT F D da" $ACT F T E
Measura!le Delivera!le Continue

5evie* and Approval C"cle CCT F N da"s 'CT F D da" $ACT F TE 5elease 5FQ CCT F J da"s 'CT F D da" $ACT F J hours
Continue
CCT F Calendar Time 'CT F 'or+ Time $ACT F $alue=Add Time
Create Final 5FQ CCT F N da"s 'CT F J da"s $ACT F D da"
As=Is Process C"cle Time[2 CCT F NQ da"s 'CT F TDM da"s $ACT F N da"s
5evise
Assumes no revisionsX
Sales 4rder Processing $alue Stream Map

Time Customer is 4n Telephone
Initial Phone Contact CCT F E 'CT F E $ACT F E 'ait for Availa!le Sales Person CCT F N minutes 'CT F E $ACT F E Sales Pitch CCT F DE minutes 'CT F DE minutes $ACT F DE minutes Configure S"stem CCT F BE minutes 'CT F BE minutes $ACT F N minutes
CCT F Calendar Time 'CT F 'or+ Time $ACT F $alue=Add Time
Fill 4ut 4rder Form CCT F DE minutes 'CT F DE minutes $ACT F N minutes
Promise to Ship CCT F N minutes 'CT F N minutes $ACT F E
Triggering vent Measura!le Delivera!le

Issue 'or+ 4rder to Factor" Floor CCT F D Da" 'CT F D hour $ACT F E
Change Ship Date Mtl%o Av ail a!l e 8

Chec+ Availa!ilit" of Materials CCT F B Da"s 'CT F D hour $ACT F E Aatch Together Similar S"stems CCT F P Da"s 'CT F D Da" $ACT F E Pending 4rder 0FIF41 Queue CCT F O Da"s 'CT F E $ACT F E
Ges
'hile customer is on telephone2
CCT F PE min'CT F NN min$ACT F JE min-
From Contact to 4rder #aunch2
CCT F DO da"s 'CT F T D da" $ACT F E
Au"ing a Salad Process Flo*

Customer !u"s salad *ith salmon Processed salmon is shipped to fish mar+ets 9rocer offers premade salads Salmon is processed at fisher" Salad compan" ma+es salad and delivers to grocer Fishers catch salmon Salad compan" !u"s supplies Mother nature ma+es salmon
5estaurant suppl" compan" distri!utes food
Produce is pac+age and shipped Meat companies process animals Manufacturers ma+e containers
Farmers gro* and harvest produce Farmers raise meat animals
Container compan" sells containers
Chemical producers ma+e plastics from petroleum
4il refined for petroleum products
C"cle Time Definition
04ne of the most note*orth" accomplishments in +eeping the price of products lo* is the gradual shortening of the c"cle time- The longer an article is in the process and the more it is moved a!out3 the greater is its ultimate cost-1 7enr" Ford3 DRJP
'or+
rrors3 *aiting3 transportation3 movement etcI-Total C"cle Time
Time that elapses from !eginning to end of process (ltimate o!jective or goal of #ean processes is to reduce c"cle time !" eliminating *aste
Aenefits of $SM
) 7elps "ou visuali6e more than the single process level ) #in+s the material and information flo*s ) Provides a common language ) Provides a !lueprint for implementation ) More useful than ,uantitative tools ) Ties together lean concepts and techni,ues
M Steps for $SM

D-Determine the Product Famil" J-Dra* "our Current State Map B-Create the Future State Map M-Develop "our plan to get there
Current State Map

) (nderstanding ho* the floor currentl" operates
; Material and Information flo*s ; Dra* using s"m!ols ; Start *ith the 0door to door1 flo* ; 7ave to *al+ the flo* and get actuals ) %o standard times ) Dra* !" hand3 *ith pencil and eraser ; Foundation for the future state
Current State Icons

Cus o!ers Suppliers
*on45 6ed45 7ri4 =
Process &o'
)aintin!
Go See
Ship!en $ Truc% Inven ory Push Sys e!
+pera or (a a &o'
(8 91 sec (8:9 1 hr Rel49 "8; 7)< 9 "5;
Cycle Ti!e Changeover )eliabili y *uali y
More Current State Icons

,ardcopy #lec ronic Cell &oa Person Plane
Train
Fun Current State Icons

> > >
Fa x
AUVWGX
Current State Map Setup

+ips (se DD1 x DO1 paper3 landscape (se pencil and eraser Dra* !" hand Don/t *aste time putting it on a computer just to ma+e it loo+ nice ?non=value added time@ Practice3 practice3 practice Steps Customer Supplier Process Information flo* Calculate process time and lead=time
Current State Map Setup

Supplier informati on Custome r informati on
Information Flo* Area
Process Flo* Area
Process Time and #ead=time Area Title Aloc+
Current State Map

2 da2s 308608"0 )rod (trl 7orecast *R) 6eekl2 6eekl2 6eekl2 Achedule Atam%in! = 2 6eeks
Ahared
20 min :rder -ntr2 *R) L8 9 2 da2s )8 9 20 min $ail2 $eburr = = 91 (8 91# sec (8:9 0 min Rel49 80; 7)< 9 100; 15225 92 (8 948 sec (8:9 5 min Rel49 100; 7)< 9 "8; 445 da2s 48 sec 40 da2s 105 sec *onthl2 6eekl2
55300 %cs8mo4 265 %cs8da2
)hl2e3 @ikni!ht
$ewe25 (heatem ? 1owe $ail2
A%ot 6eld = 55425 91 (8 93" sec (8:9 11 min Rel49 ""; 7)< 9 "0; 2045 da2s 3" sec 5 da2s
&ssemble
91 (8 91 sec (8:9 4 hrs Rel49 "8; 7)< 9 "5;
15400
10 da2s
1 sec
1# sec
Future State Questions

D- 'hat is the Ta+t Time8 J- 'ill *e !uild to shipping or to a supermar+et8 B- 'here can *e use continuous flo*8 M- 'here do *e have to use supermar+et pull s"stem8 N- At *hat single point in the production chain do *e trigger production8 P- 7o* do *e level the production mix at the pacema+er process8 O- 'hat increment of *or+ *ill *e release and ta+e a*a" at the pacema+er process8 ?#eveling the volume@ Q- 'hat process improvements *ill !e necessar"8 ?e-guptime3 changeover3 training@
D- 'hat is the Ta+t Time8

) Ta+t means drumbeat ) A!ilit" to meet customers/ demand
) Formula
Ta+t Time F
Time Availa!le Demand
Ta+t Time Calculation

Time availa!le Shift ?Q hours@ F MQE mins Area+s ?J xDE@ = JE mins #unch = BE mins Meetings = N mins CC4 = N mins Total Time F MJE mins F JN3JEE sec Demand F JPN parts Ta+t Time F RN secCpart

Ta+t Time F Demand 5ate 9oal2 Produce to demand *ith no excess capacit" Ta+t Time F *or+ time availa!le num!er of units sold Assume N people *or+3 sell NEE unitsC*ee+2 Ta+tTime F ?N x ME x PE@ C NEE F JM minCunit Set c"cle time to match personnelCoperation
xercise
) Ta+t time calculation

4ld re,uirements2 QMOCda" [ JME *or+da"sC"r F JEB3EEEC"r DES gro*th F JJB3PEEC"r %e* re,uirements2 JJB3PEC"rCJME *or+da"sC"r F RBDCda" Time availa!le2 Q-N hrsCda" = -N hrs ?lunch@ = -BB hrs ?!rea+s@ F O-PO hrsCda" B3PEE secsChr [ O-PO hrsCda" F JO3PDJ secondsCda" JO3PDJ secondsCda" divided !" RBD unitsCda" F JR-B secs per unit C"cle time ?actuall" DDP secsCunit@ divided !" Ta+t time ?JR-B secsCunit@ F B-RN F M operators re,uired
Quic+ Changeover
Changeovers in less than DE minutes
Quic+ Changeover
) Factor" definition ; the time from the last good piece of previous run to the next good piece of ne* run ) 4ffice definition ; the time it ta+es to s*itch from one tas+ to a ne* tas+
; T"picall"3 in the office the time savings is not as significant as in manufacturing
J 7our CC4 ; #arge Aatch Si6e

Q O P N M B J D E
A
Mon Tue
A
' ed
C
Thu Fri
D 7our CC4 ; 7alf Aatch Si6e

Q O P N M B J D E
A
Mon Tue
C
' ed Thu Fri
DE Minute CC4 ; Small Aatch Si6e

Q O P N M B J D E
D C A
Mon
D C A
Tue
D C A
Thu
D C A
Fri
D C A A
'ed
DE Minute CC4 ; Man" Different Small Aatches

Q O P N M B J D E
D C A
Mon
D C
A D F A C 7 9 A I
A
Tue
A
Thu
7
Fri
'ed
Changeover Summar"
Aased on per *ee+ CCE Time %um!er of ME hours %um!er of Production Changeovers Production Time 5uns Availa!le
J 7our D 7our
N DE
N DE DN JN
BE hours BE hours BJ-N hours BN-Q hours
BE Minute DN DE Minute JN
SM D
) Single Minute xchange of Dies ?SM D@ ) Shigeo Shingo ?DROE@ D-Separate internal steps and external steps J-Convert internal steps to external *here ever possi!le B-Streamline all steps
M Categories of SM D time
D- Preparation3 after= process adjustments3 chec+ing of material and tools ?BES@ J- Mounting3 removing tools and parts ?NS@ B- Measurements3 settings and cali!rations ?DNS@ M- Trial runs and
BES
NS DNS
NES
T"pical proportions
Step D- Separate Internal and xternal Times

) Chec+lists ) Functional chec+s ) Transportation of parts and tools
Step J- Convert Internal Time to xternal Time

) Preparation conditions
; Pre=heat3 correct air pressure3 stage materials
) Standardi6e
; Centering3 gripping3 securing3 replace fe*est parts3 standardi6e heights3 standardi6e !olts or fasteners
) Intermediar" jigs
; Mounting plates
Step B- Streamline
) Parallel operations
; More than one person *or+ing at the same time
liminate adjustments
; Mar+ings ; Scales
) Functional clamps ) Mechani6ation
Changeover Cart xample
Aefore ) 'aste of time to find correct tools ) Tools can !ecome damaged ) 'aste of mone" for extra tools
After ) Saves time ) Do not have to replace tools as often ) 7ave *hat "ou need *here its needed
QC4 and 4ther Auilding Aloc+s

) NS3 $isual3 P4(S3 Teams and Standard 'or+ ) $SM can discover opportunities for QC4 ) As Aatch Si6e 5eduction continues3 QC4 !ecomes more important ) <ai6en Alit6 is a great method to implement QC4 ) Must sustain the gains
Aenefits of Quic+ Changeover

) Shorter lead time ) #ess material *aste ) Fe*er defects ) #ess inventor" ) #o*er space re,uirements ) 7igher productivit" ) 9reater flexi!ilit" ) Aetter Team*or+

Change Management
Continuous Improvement
<ai6en vs- <ai6en Alit63 or Incremental vs- Area+through Improvements <ai6en Incremental Improvements2 ) Are continuous3 since there is al*a"s room for improvement in an" process ) It is never=ending ) Man" small improvements throughout the enterprise ) Done !" individuals or small teams ) Could !e functional3 departmental3 or tas+=oriented ) Part of the 0useful man"1 ) A little time spent on an ongoing !asis ) Standardi6ation of processes ?i-e-3 process improvement oriented@ ) Plan=Do=Chec+=Act methodolog"

Continuous Improvement
Ideas for continuous improvement could come from2

; ; ; ; ; ; ; ; ; mplo"ee suggestions Corrective & Preventive actions %on=conformities3 defects Customer complaints3 returns Aenchmar+s The #ean 0*astes1 $ariations from the standard Assessments3 audits & competitive anal"ses 5esearch & Development activities
Continuous Improvement & Continuous #earning
) Continuous #earning goes hand in hand *ith continuous improvement ) Management should have training given to emplo"ees in #ean and Qualit" tools3 pro!lem solving and root cause anal"sis3 the process model3 concepts of Theor" of Constraints3 !asic statistical techni,ues3 graphical tools3 etc) (nderstanding of Plan=Do=Chec+=Act and Standardi6e=Do=Chec+=Act ?SDCA@ *ill !e !eneficial
Continuous Improvement & Continuous #earning
) Continuous improvement and learning2

; Aecomes part of dail" *or+ life ; Is practiced at !oth personal3 functional and organi6ational levels ; Is result oriented ; Is shared *ithin the enterprise ; Aecomes part of institutional memor" and +no*ledge3 even after emplo"ees retire3 move upClaterall" or leave
'h" C- I-8
) Standing still is not an option2
; ; ; ; ; ; ; ; ; ; Competitors *ill overta+e us 9lo!ali6ed econom" 7igher customer expectations Technical and !rea+through changes Tapping into human potential and creativit" Cost and c"cle time reduction 0'aste1 minimi6ation Defect prevention nhancing customer satisfactionCdelight Attaining competitive advantage
) Improvements !ased on2
<ai6en Alit6
) Area+through strategies for lasting results

Change Management
<ai6en Alit6
) <ai6en Alit6 is a com!ination of the .apanese *ord <ai6en for 0continuous improvement1 and the 9erman *ord Alit6 for 0lightning-1 It is a focused3 week:long *or+shop *here a cross=functional team revie*s a process3 identifies and eliminates *aste3 there!" achieving dramatic and tangi!le breakthrough ?rather than incremental@ improvement results) <ai6en Alit6 no* stands to mean the improvement activit" itself) It is treated more as a 0Project1 ?rather than a 0Process1@-
'h" <ai6en Alit68

) Major !enefits in a flash) Can use !enchmar+ing for setting goals) Innovation has !ecome indispensa!le in toda"/s competitive *orld econom") C"cle Time 5eduction translates directl" to cost savings) <ai6en Alit6es t"picall" attac+ *asted time) Positive impact on organi6ational culture through 0!rea+through1 t"pe improvements-
'ho 'ill !e Involved8

) <ai6en Alit6 teams that come together for one *ee+ to implement improvements in a pre=selected !ottlenec+ project or process) Cross=functional teams ?seven to ten persons@ ) 7ourl" and salaried personnel ) 4perators3 engineers3 supervisors3 maintenance persons3 managers3 technical experts3 material handlers3 ,ualit" personnel3 !usiness support personnel3 participants from the outside ) T"picall"3 the Team #eader is person *ith clout and is the highest sta+eholder in the process *ho possess leadership s+ills3 open minds3 strong desire to succeed and some prior #ean experience-
<ai6en and C"cle Time 5eduction

) ) ) )
Focus on 'rocessFocus on Elimination of WasteFocus on S"eedTime improvement translates directl" to cost savings and customer satisfaction-
<ai6en Alit6 Steps

D-Select specific area for improvementJ-Define current situation in measura!le terms B-Set aggressive goals ?stretch goals@M-Identif" team mem!ersN-Conduct training on the first da" of the projectP-Do it ?in three to five da"s@-
Step D ; Select Project

) ) ) ) ) ) ) $alue Stream Map Aottlenec+s Customer or ,ualit" related issues Interdepartmental #ong lead=times or setup times Competitive advantage Cost reduction or avoidance
Step J ; Define Current State

) ) ) ) ) $ideo tape Time stud" Flo*chart 7istorical information 4!servations and intervie*s
Step B ; Identif" Team Mem!ers

) Cross=functional teams ?seven to ten persons@ ) 7ourl" and salaried personnel ) 4perators3 engineers3 supervisors3 maintenance persons3 managers3 technical experts3 material handlers3 ,ualit" personnel3 !usiness support personnel3 participants from the outside ) T"picall"3 Team #eader is person *ith clout and is the highest sta+eholder in the process- Ideall"3 the team leader must possess leadership s+ills3 open minds3 strong desire to succeed and some prior 0#ean1 experience-
Step M ; Set Aggressive 9oals

) Define the purpose or o!jective and set stretch goals) 9oals should !e clearl" defined and ,uantifia!le ?e-g-3 reduce machine set up time !" ONS3 increase throughput !" MES3 reduce floor space !" BES@) mphasis should !e on identif"ing and eliminating *aste3 and then standardi6ing at the improved level) Aenchmar+ *hen possi!le
Step N ; Conduct Training

) Select hands=on training that is compati!le *ith the project ) Do training on the first morning ) Provided !" an internal or external expert
Step P ; Do itX
) Perform in B to N da"s
xample D ; (niversal .oint QC4

Categor" Changeover Tool Change
Target savings M-R hours ?ONS@ DM-Q hours ?ONS@ Actual savings M-N hours ?PRS@ DB-Q hours ?OES@ Target savings LJRD per da" Actual savings LJPQ per da" Total savings LQOR LQJE
LJOJ3EMD per "ear
xample J ; Tu!e Mill SM D

) Changeover savings per "ear2 LNDE3EEE ) 5educed changeover time from M hours3 ME minutes to J hours3 DD minutes ) mplo"ees happier ; !onuses !ased on changeovers ) Management happier ; more changeovers3 more production3 more cash in the door
7o* to Start and Sustain "our #ean .ourne"
) A journe" of one thousand miles3 starts *ith a single step
Q 'a"s to 9et Started

D-Aaseline Assessment or 9ap Anal"sis J-$alue Stream Map B-Training in #ean M-Aasic Auilding Aloc+s N-<ai6en Alit6 P-Pilot Projects O-Change Management Q-4
D- Aaseline Assessment
Aaseline Assessment or 9ap Anal"sis performed !" experienced #ean experts ) (se
; ; ; ;
Intervie*s 4!servations Process mapping Anal"sis of relia!le data
) Create a 09ap Anal"sis1 *ith focus on eliminating the ight 'astes ) 9enerate an Action Plan for implementing #ean improvements
J- $alue Stream Mapping

$alue Stream Mapping ) Assem!le the cross functional team ) 7ave a $alue Stream Manager ) Determine a Product Famil" ) Create the Current State Map ) Create the Future Stats Map ) Develop the Plan to get there3 tie=in *ith !usiness o!jectives ) 5evie* the Plan3 sta" on course ) Gour Future State then !ecomes "our Current State ) xpand to Multiple $alue Streams

B- Training in #ean
Training in #ean
) 0Massive1 training in #ean ) %eed to !uild a critical mass of trained emplo"ees ) Perform the training just !efore implementation ) #ean Champions should have advanced s+ills in #ean
M- Aasic Auilding Aloc+s

) Start *ith one of the Aasic Auilding Aloc+s of #ean
Continuous Improvement & <ai6en Alit6 TPM Po+a="o+e P4(S #a"out .IT Cellular & Flo* Self Inspection Pull S"stem & <an!an Autonomation $ S M
Aatch Si6e 5eduction Standard 'or+
Change Management
Aasic Auilding Aloc+s

Aasic Auilding Aloc+s ) Start *ith the implementation of the Aasic Auilding Aloc+s ) Auild up la"er !" la"er until TPM3 Cellular Manufacturing and PullC<an!an are esta!lished ) Then continuousl" improve using <ai6ens3 suggestion s"stems and periodic $alue Stream Maps
xample ; Implementing NS
) Plan
; Identif" NS Champions & Teams ; Decide ho* to roll= out to entire organi6ation ; 5esources re,uired NS Implementation Map Q R O Q Q R R DE DB Q Q R R DE N D M J DD D D D M B N P DJ M M
) Train
; Train just !efore <ai6en ; Train=the=trainer
) Do ) Improve & 5epeat
N- <ai6en Alit6
Select a <ai6en Alit6 project ) Perform in B=N da"s ) Focus on speed and elimination of *aste ) Can perform on 0lo* hanging fruit1 ) 9enerates ,uic+ victories and improvements ) Do not continue in an Ad=hoc approach3 use "our $alue Stream Map
P- Pilot Projects
Pilot Projects ) Implement #ean Pilot Projects *here !ottlenec+s have !een identified ) (se cross=functional teams ) PDCA methodolog" is !est ) Can use !enchmar+s and !est practices for goal setting ) Communicate results ) Migrate lessons learned to other areas
O- Change Management
Change Management
) Aegin *ith cultural Change Management !efore rolling out #ean ) Address the human side of #ean in "our three to five "ear Master Plan ) (se internalCexternal change agents ) Communicate the need for change2 ultimatel"3 #ean has to !ecome integrated into dail" *or+ life ) 4pen up channels for sharing ideas
Q- 4
4verall ,uipment ffectiveness ?4 @ anal"sis can identif" *here to start "our #ean journe" ) Pareto the time spent on2
; ; ; ; ; ; ; ;
Area+do*ns Setups Tool changes Idling time Slo*er speed Minor stoppages Producing defects3 re*or+ Start=up issues
) This exercise *ill self=identif" the 0!iggest !ang for the !uc+1 and *here to start
Plan
) Gou must have a plan ) (tili6e a Steering Committee3 Design Teams and #ean Champions ) Tie the #ean 4!jectives *ith the Ausiness 4!jectives ) Commit resources ?time3 people3 !udgets@
Stages of #ean Implementation
9enerall"3 organi6ations can use this model for the stages of implementing #ean
; sta!lish ta+t time and meet it
) Ta+t ) Flo*
; After meeting ta+t time3 then create Flo*
) Pull
; 'here "ou can/t Flo*3 Pull
7o* to Sustain #ean8

) #ean *ill not !e sustaina!le *ithout proper training in #ean and satisfied emplo"ees ) Internali6e into dail" *or+ ) (nderstand that it is a never=ending process or philosoph"2 no turning !ac+ ) Create disciplineCmotivationCincentives ) Standardi6e so as not to slip !ac+
7o* to Sustain #ean8

) Continued3 visi!le management commitment ) 4pen communication channels ) mphasi6e accounta!ilit"
) (se #ean performance metrics ) 5ole of #ean champions ) .o! rotation
7o* to Setup the #ean Team
) Steering Committee ) Design Teams ) Champions
#ean Team 5oles & 5esponsi!ilities

Design Team NS3 $isual3 P4(S Design Team #ean Steering Committee 7PT Design Team PullC<an! an Design Team
Cellular Design Team
Champions
Champions
Champions
Steering Committee
5oles & 5esponsi!ilities ) Set the #ean polic" ) Provide resources ; time3 people3 #ean Steering !udgets & remove !arriers Committee ) Develop and share the #ean $ision ) Develop the Communication Plan and then deplo" it ) Mem!ers usuall" ) 0'al+ the tal+1 ever"da" and full" from top support the #ean initiatives management3 !ut can include 0$alue ) Determine the Design Team ma+e= Adders1 up and mem!ers ) 5evie* the *or+ of the Design Teams
Design Teams
5oles & 5esponsi!ilities ) Deplo" the #ean polic" ) Determine the resources re,uired ; Time3 people3 !udgets3 etc ) Determine the !est *a" to implement the #ean Auilding Aloc+s in "our organi6ation ) 5eport to the Steering Committee on progress ) Support "our #ean Champions
Design Team
) Mem!ers usuall" from management or 0$alue Adders1 ) 9roup li+e Auilding Aloc+s together
Design Team Agenda

) Design Teams decide ho* ) Design Teams ma" change #ean *ill !e implemented over time at their facilit" ; At first the" oversee ) Ta+e into account2 the implementation plan ; 4rgani6ational Culture ; Then the" support the sustaining ; Change management efforts ; 5esources ; The" ma" dis!and3 ; Current s+ills & !e a!sor!ed into needed s+ills another Design ; Si6e and timing of Team or morph projects into a ne* design Team ; Metrics & 9oals
#ean Champions
5oles & 5esponsi!ilities ) Deplo" the #ean polic" via training3 implementation and <ai6en Alit6 ) Feed!ac+ information to the Design Team on progress ) Ae given the time to support the #ean efforts ) Ma+e presentations and communicate the results of the #ean projects
Champions
) Mem!ers usuall" from management or 0$alue Adders1 ) Motivated to learn3 lead and improve their organi6ation
Champion/s Agenda
) Ae read" to commit time to #ean projects ) Ae *illing to learn and continuall" improve ) Aecome a #ean content expert ) 7ave s+ills in training3 pu!lic spea+ing3 project management and !e a team
'IIFM
) 9ain ne* s+ills ) $alued !" the organi6ation ) xciting3 ne* assignments
) #earn other aspects of #ean
Multi=facilit" Deplo"ment xample

Small Facilit" #ean Steering Committe e Design TeamC Champio ns 4rgani6at ion #ean Steering Committe e Large Facilit" #ean Steering Committe e Desi Desi gn gn Tea Tea m m
Desi gn Tea m
Champions
Desi gn Tea m
Momentum
) 7ave to !uild a 0critical mass1 of emplo"ees trained in #ean and appl" principles ) Auild 0!u"=in1 and get people on!oard ) 0Aand*agon1 affect
4rgani6ational Alignment
Dealing *ith 4!jections to #ean

) Put "ourself in their shoes ) 7elp ans*er 0'IIFM1 ) Communicate3 communicate3 communicate and then communicate some moreX ) Create an 0 levator Speech1
9etting People on Aoard

se i d a r a P #ean
Status Quo #and
#ean nterprise vs#ean Manufacturing
) Ta+ing #ean !e"ond the shop floor
#ean nterprise
) Move from the shop floor to nterprise=*ide #ean implementation ) Man" of the Auilding Aloc+s are essential for efficient office functions ; NS3 $isual3 P4(S3 Standard 'or+3 #a"out3 Self Inspection3 Po+a= "o+e ) The goal is to reduce or eliminate the *astes to reduce lead times and to enhance responsiveness3 competitiveness and customer satisfaction
+hank youX
***-as,-org

ASQ Six Sigma

Uploaded by

Copyright:

Available Formats

ASQ Six Sigma

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

ASQ Six Sigma

Uploaded by

Copyright:

Available Formats

Workshop on Structured Problem Solving Using DMAIC Methodology

The volution of Qualit"

.uran/s Definition of Qualit"

IS4 Definition of Qualit"

Cros!"/s Definition of Qualit"

) Qualit" is conformance to re,uirements) 5e,uirements are ans*ers to +e" organi6ational ,uestions2

Qualit" volution2 Medieval 9uilds

Qualit" volution2 Product 4rientation

Qualit" volution2 Process 4rientation

uProcesses !ecame uShe*hart identified

Qualit" volution2 'artime

issueuThe militar" developed a sampling inspection s"stem and trained suppliers-

Qualit" volution2 Total Qualit" Movement

'- d*ards Deming

; Qualit" planning ; Qualit" control ; Qualit" improvement

Total Qualit" Management

Total ,ualit" management ?TQM@2

Customer relationships Continuous improvement Compan"=*ide participation

#evels of Customer Satisfaction

'- d*ards Deming on Qualit"

A Qualit" Approach Aenefits - - 2mployees

Qualit" Studies and Standards

xternal and Internal Customers

) Increased customer satisfaction

conomic gro*th and sta!ilit"

Increased emplo"ment opportunities

Introduction to Six Sigma

'hat is #ean Six Sigma8

'hat is Six Sigma8

#o*er Process Control Center

(pper Control 6 Sigma

Process Capabil ity

Measurem ent System

7arvesting the Fruit of Six Sigma

================================== M 'all = Must Improve Internall"

M 'all = Must Improve Internall"

+o( ,anging )ruit

B 'all = Demand improvement B 'all = Demand improvement

Logic and ntuition

Process Characterization and Optimization

The Focus of Six Sigma

- - - K% Independent Input=Process Cause Pro!lem Control

Six Sigma Defines Pro!lems Statisticall"

Should we focus on the process outputs (Y) or inputs ( )!

G = 4utputs uDependent u4utput u ffect uS"mptom uA!le to Monitor

KD - - - K % = Inputs uIndependent uInputs3 process uCause uPro!lem uControlla!le

The process is used to control the product-

RRS 9ood ?B-Q Sigma@

RR-RRRPPS 9ood ?P Sigma@

The Define Phase

Projects Must !e I Meaningful And Manageable!

Sorting Projects from Messes

Project Selection is Critical

#ean Six Sigma Project Criteria

'hat is Customer Satisfaction8

Customer %eeds Translated into 0Critical To1 ?CT@ Characteristics

Project Selection Criteria

Aligned *ith core o!jectives

Process Improvement Data Availa!ilit"

Project Selection Matrix Diagram

!ohit "adav ()+,-*(( .++/