Project 6 Sigma Pre-DMAIC Process DMAIC Process Define Measure Analyze Improve Control Cp, Cpk, Pp, Ppk Sigma DMAIC
PROJECT
6 Sigma? What is Sigma? ( ) Sigma (=) 6 Sigma /service, Sigma - Greek letter - data, 6, 6 Sigma - Data - Data - Processsigma process,, service, 2 Sigma*: /Service69146% 308,538 () 4 Sigma: /Service99379% 6,210 () 6 Sigma: /Service9999966% 34 () * Short term
6 Sigma ( ) 6 Sigma methodology performance excellence process, culture Performance excellence - Transformational Change Across-the-board: process, culture, customer - Operational Change Business processes: tool, methodology Critical Customer Requirement : CCR Defect: Service unacceptable to customer Product or Service Output
6 6, QC 7,,,
Sigma 6 Sigma output supplier, input, process defect 6 Sigma Market Suppliers Inputs Business Process Process Output Gap Critical Customer Requirement Defects Defect (root cause) defect Process output variationdefect
Pre-DMAIC Process Issue, gap process - or gap? - or? - or,? - or process? 6 Sigma Issue - - - -, - goal - communication - Process - process
6 Sigma process Key Performance Measure process Business Strategy Key Performance Measure * Chosen Market Leadership Processes Business Process Process Output Process Sigma Critical Customer Requirement
6 Sigma 6 Sigma gap Analyze Plan Focus Build Activity Output gap Gap cost Improve readiness, sigma multiply Analyze Control Measure gap Define - (needs), - - - - (ROI,, -, ) - - Quick wins - Six SIGMA Breakaway -,, - - ( ) - - - - - - - - - - - - - - Six sigma - - - Build Focus Analyze Plan
Black Belt, Master Black Belt Master Black Belts : Expert Consultant Black Belts : Team Leader and Coach - Black Belt - / tool - - Business management team consultant member - Sigma - lead - Change agent Senior Leadership Team : Business Unit Leadership Team : Improvement Project Team : Set Direction Prioritize & Deploy Teams Deliver Improvement Results -, Sigma - - review - communication - - - manage - champion - needscommunicate - Team Supporter Employees : Provide Support - impact - Process (moral support)
DMAIC
6 Sigma Process 6 Sigma, data, data,, pilot process 5 6 Sigma 5 Define Opportunities Measure Performance Analyze Opportunities Improve (Develop Solutions) Control (Implement) 11 Validate Business Opportunity 21 Determine What to Measure 31 Process Stratification and Analysis 41 Generate Improvement Ideas 51 Develop Pilot Plan 12 Document and Analyze Processes 13 Define Customer Requirements 14 Build Effective Teams 22 Manage Measurement 23 Understand Variation 24 Determine Sigma Performance 25 Excel Team Performance 32 Determine Root Causes 33 Validate Root Causes 34 Manage Creativity 42 Evaluate and Select Solutions 43 Present Recommendations 44 Implementing Change 52 Plan and Implement Solution 53 Process Integration 54 Closure and Recognition Tool & Technique - Affinity Diagram - Brainstorming - Customer Research - Deployment Mapping - Gantt Chart - Kano Analysis - Team Charter - Qualitative Process Analysis - Top Down Mapping - VOC Translation - Quality Function Analysis - Control Charts - Data Charts - Gauge R&R - Histogram - Indicators - Run Charts - Sigma Calculation - Stratification - Variation - Cause & Effect Analysis - Correlation - Design of Experiment - Fault Tree - Pareto Charts - Process Simulation - Quantitative Process Analysis - Regression - Stratification - Structure Tree - Value Analysis - Challenge Assumption - Error Mode Effect Analysis - CDAM* - Process Control System - Cost Benefit Analysis - Replication - Criteria Selection Matrix - Standardization - Force Field Analysis - Work Planning - How-By Pursuit - Lateral Thinking - Mind Mapping - Random Word - Storyboard - Solution Mapping * Combine, delete, add, modify
Define Define process - - Business process - Critical Customer Requirements - project Business Documentation & Process Analysis Define Customer Requirement - - - - -, team charter Mapping - - (top-down) VOC translate - Critical Customer - - Requirements review Quick-win - - Process map - VOC, CCR, CTQ - Team role profile - Team charter : Top down chart Flow - Barrier : Functional deployment chart - Value analysis - Quick win list
11 Business (operation) (value) impact customer impactbusiness impact Business Process : Theme : : - / / (Referrals) - / /
project team charter Charter / Team Charter Team Charter ( ) 1 Business Case ( ) Business Case - 5000, - boundary, Project 50%,, 2 (Business ) 9 SMART, -? -? Specific : 3 ( )? Project -? Process Measurable : - - : :? 20% - : Attainable - 50% Process 4 Project (Boundary) :? - - process - -Business 50%? - 6 - R&D business -? - Engineering Relevant : Biz -?? - Project Time bound 5 Project : - /? Team 1234? - Champion: - - Key job communication plan - - Master BB: - - BB : - - 6 Team - : - - : -? - - : - - -? - : - - Communication - : -
12 Documentation & Process Analysis process - (top-down chart) (functional deployment process) Process MappingValue Analysis Process Input, Output, Process activity value Activity Value Analysis ( ) Value to Customer Value to Operation Non Value Process Mapping Supplier Input: a1 A B C a b c Process a2 a3 Output: or Service Customer Activity 1 Activity 2 Activity 3 Activity 4 Activity 5 Process - Top down charting - Functional deployment process - - - activity - 6 Sigma CTQ Non valueactivity, Quick Win
S-form Pick List Work order / Top Down Mapping ( ) carrier / Release, /QC, / Least-cost Routing 3, A/R Functional Deployment Process Map ( ) Routing A4 A3 A1 A5 A11 A2 /QC A6 A7 A8 A12 A10 A13 A14 A15 A16 A17 A18 D3 A9 D1 D2 A19 A20 A21 A22 A/R
13 Define Customer Requirement needs CCR, CBR Impact output indicator CTQ CTP Process Output Indicator segmentation CTQ * CCR *** Key Issue Voice of Customer Voice of Biz Key Issue CBR CCR output - point, bias -,,, issue - Business process CTP ** CCR -, output, :, point - issue : : - : : Business impact * Critical To Quality ** *** Critical TO Process Critical Customer s Requirement
CCR Key Issue Voice of Customer - 2kgf/ 2 leak - Dead cell - / - - lot range - - 20 - - 300 / - - lot 70% - Voice of Business Key Issue CBR - Leak - Seam, Ball, Gasket - 2kgf/ 2 leak - - - Ni-Strip 2kg - Ni-Strip - - - Slittingwave - Loading - Slitting - slitter
14 Project, relationship,, process Team Mission Project 1 : Champion:? 2 :,? Black Belt:,, 3 Relationship:, Full Time Part Time,? 4 Process?,? Champion: Black Belt: Full Time:
Measure process CCRCTQ input, proces output indicator variation sigma Understand Variation Sigma Performance - CCR - - - - - -, - - - - - - - 6 - operational definition baseline - Input, process, output - Data - data graph - sigma indicator operational definition :Run chart - Special cause - Indicator :Histogram :Control chart - Special cause data graph
Measure 21 CCR output indicator ( =measurement) input, process indicator Process Measurement Indicator Process measurement - - Supplier Input Process Output CCR - - - Input Indicator Process Indicator Output Indicator - - - - - - CCR - 3 Process Indicator Input Indicator - A B C - Output Indicator a b c Process output indicator key indicator
22 Manage Measurement key indicator operational definition data data Data Process -? -? -? -? - -? -? -? -? Indicator operational definition Data Data display Data : ProcessCycle Time,,,, Data Data Data criteria Operational Sample Data Measurement source Definition Data -?? Fax? / ( ) -? Cycle time 289,,? or Fax -? Fax Center -???? -?, / ( )
23 histogram, run chart, control chart data special causecommon cause - -,, - Sigma : Process performance : Customer requirements - Sigma special cause, common cause output Histogram - ( ) - Data Run Chart - data Control Chart - Run chart 3 Process - Data process variation - / / special cause -Special cause data
24 Sigma sigma data spec data ( data, data), (, ) Sigma Process (CCR) output Sigma - CCR output - output sigma - CCR * - Sigma : data or data - Output : Operational definition : - (Short term) : 30-50 : performanceshift - (Long term) : 100-200 : performanceshift Datalong-term, short-term : / special cause long-term Sigma Value - data *, data (P> 005) - CTQ Cp, Cpk, Z Sigma : Short Term Zst = 3Cp : Short Term Zst = 3Cpk : Long Term Zlt = 3Pp : Long Term Zlt = 3Ppk - 1 3 CTQ :CTQSigma normalize eg CTQ 1 = 95%, CTQ 2 = 96%, CTQ 3 = 97% 3 CTQ = 095096097 = 096 = 96% 175 Sigma Short Term 175 + 15 = 325 Sigma RTY: Rolled Throughput Yield = 1 (95%) 2 (93%) 3 (90%) = 79% * Data - - n = P = = ( s ) P = = ( )
CTQ sigma CTQ normalized sigma Paste resin for Glove CTS : 1) Pinhole 2) CTQ (NFNo) New Spec (n=30) 60 15 Z=145 40 50m Cpk : 096 08% 08013 Cp : 112 3,300( ) 3,800( ) 3,550220 220 Cp : 115 Z=345 Cp,Cpk,Z Sigma / Sigma Z Z=145 92647% 0963 =288 99801% 1123 =336 99961% 345 99972% Z=345 99972% 5 Y norm = 92647% 99801% 99961% 99972% 99972% = 98426% Sigma lt=2151** ** DataLong-Term
sigma /, servicesigma, Service sigma /OBU Sigma Value Core Process Manufacturing Transactional Sigma A B 80% 20% 30 40 100% 32
Define Opportunity Define CCR, CTQ, Team Charter CTQ, Data ( ) Data : variation Data :, variation Chart(Run, Histogram) N Sigma Y Common Cause, Special Cause Sigma Sigma : Cp, Cpk : Pp, Ppk Special Cause Special Cause Data (Quick Win point) Measure team charter Analyze Root Cause
Analyze D M A I C Measure special cause data data, Ishikawa diagram, Process - - Measure - - - /, - Improve -, -, -, - : Output:, - - Ishikawa Diagram, - Correlation, Regression Tree Diagram ANOVA, Sample t - - DOE
D M A I C 31 Process Definition - - (,, ) - - - :,,,, Tool Pareto - : 80% 20% - : 2~3 focus - Define process - - Data Criteria +Measurable + (how, where, (Poor Biz (What is how much, Result) Wrong?) when) Vendor 70 down -,? - Gap ( ) - -,, - 2 + Focus on Customer Gap ( &Should be) Down zero Matrix = System down of 70 hrs in the last 2 month have resulted in 50 customer complaints and vendors to not receive payment
D M A I C 32 (Root Cause) Output (symptom) / tool Ishikawa diagram tree diagram / - -?- - - identify - - / - - Environment - Position - type - Material - Differences in products - Symptom 2 Ishikawa Diagram?? - 1 / - 1 check because 2 / 3 Tool Tree Diagram - :Man,Method, Machine, Material, Environment Why? Why? Why? :Brainstorming Why? Because Because Because - Diagram 1 2 3~5 mixing wax mainta- nance
D M A I C 33 problem tool process Tool - - Process (Correlation) Linear Regression (Regression) Multiple Linear Regression - - - Y X variables - r - r 2 - r 2 : -1<r<1 : 0<r 2 <1, % : 0<r 2 <1, % r 0 : Y ~% x : Y ~% x - P<005 r - P<005 r 2 - P<005 r 2 - VIF*>10x - : data - Minitab Practice Stat>Basic Statistics> Correlation variables OK - Minitab Practice Stat> Regression> Regression Response Y variable Predictor X Variable OK, OK - Minitab Practice Stat> Regression> Regression Response Y variables Predictor X Variables OK, OK
D M A I C ANOVA (Analysis of Variance)Y x, x Y Main Effect Plot Interaction Plot ANOVA* - Y - Y - Main Effect Plot, Interaction Plot, Cube plot x - Minitab Practice - Minitab Practice Stat>ANOVA> Main effect plot> Responses Y Factors x > OK : - Minitab Practice Stat>ANOVA> Interaction plot> Responses Y Factors x > OK Stat>DOE> Create Factorial Design> Design OK - Stat>DOE> Factorial Plots> Main Effect set up OK : Interaction set up : OK X1 X2 X3 Cube set up OK OK 1 2 X1 X2 3 X3 1 X1, X2 2 X1, X3 3 X2, X3 DOE ( ) One-way ANOVA - data - Minitab Practice Stat>ANOVA>One-way> Response Column OK - P< 005 P> 005 General Linear Model - Y x - Minitab Practice Stat>ANOVA> General Linear Model> responses Y Model x Covariates> Covariates x >OK>OK - P< 005 * Analysis of Variance
D M A I C Sample T (? ) 1-sample T - Data : : 1250hr? - Minitab Practice Stat> Basic Statistics> 1-Sample t Variables Column, Test Mean Alternative not equal Confidence 95 or 99 > OK (41) C1 T-Test of the Mean Test of mu = 4100 vs mu not = 4100 Variable N Mean StDev SE Mean T P C1 10 4440 1389 439 077 046 P005 (=not equal) = 2-sample T - Data :? : A B? - Minitab Practice Stat> Basic Statistics> 2-Sample t Samples in different Column> ( : Sample ) First column Second column Alternative not equal Confidence 95 or 99 > OK Two sample T for vs N Mean StDev SE Mean C1 10 444 139 44 C2 10 442 150 48 95% CI for mu C1 - mu C2:(-135, 139) T-Test mu C1 = mu C2 (vs not =):T=003 P = 098 DF = 17 P>005 : 80 60 40 20 0 Boxplots of C1 and C2 (means are indicated by solid circles) C1 C2 Box 50%Data Median
D M A I C Analyze Process Flow - focusing - yx - 24 Sigma Calculation 31 32 33 Improve Y - y data -, - Sub ( ) sigma - - x 5M 1E, 5 (x ) - y, x ( )
Improve / idea,, stake holder commitment Idea Present Recommendation Implementing Change - - - - - - (Change Idea stakeholder - () management methods) - - -,, - - () feasibility test,, - Output: - Idea - - Stake holder -
41 Idea brainstorming,, idea, Idea Matrix Tool Description Traditional Idea Generation Brainstorming - focus -, - idea - idea -, Non-Traditional Idea Generation - (assumption) - (Challenge Assumptions) - Mapping - idea - - Mind Map - mental picture Six Thinking Hats - - Idea - issue brainstorm - - - 6 - - Non-linear thinking -,
42, 1filtering sigma,,, Process 1 2 3 4 n 1 2 3 4 n Pre Filter - Show Stoppers - Organizational Fit 3 4 5 6 - Show Stopper : -? - / -? -? 250 200 100% 80% : -? 150 60% :? : scope -? 100 40% 50 20% - Organizational Fit -? 0 #1 #2 #3 #4 0% :? - :?? :,? : Process 1 2 3 4? 40% 25% 20% 10% 1 80% - ( :2) (1) (3) (1) 2 20% 80% Go to next page 3 4 30% 100% 1: 0408 = 32% 2: 02502+0208 = 21% 3: 02503+011 = 175% 1, 2 53%
From previous page risk control, Risk / control
, stake holder / Step 1 Step 2 Step 3 Step 4 Step 5 Step 6 Step 7 Presentation - Team charter condition ( ) - Data - -, 6 - - (2 ) Gantt Chart Presentation Tools, milestone Action Plan?,,?? Storyboard chartsample Define text Customer Team Charter As-Is Process Requirements Measure flow chart Problem Statement Data Collection Plan Analyze Improve Ideas Solution Implement Plan Data Collected Verified Root Causes 1 2 3 Should-Be Process
44 Implementing Change 12 Drivers -, -, -,, - - - - data - - -
Control pilot, Pilot - -Pilot - pilot - - - - - - -, / -Pilot - process system Output: Pilot -Pilot - -Pilot - Pilot - Risk assessment - - Project lessons - - - FMEA learned - EMEA - - Process control system -
51 Pilot Pilot Pilot Pilot Pilot Activity - - Scope - Impact - Pilot criteria data, customer, vendors/suppliers - /staff, technology / - Contingency System feedback - Training - - Data - - - Communication - / - - - - test - Improve - Problem Statement Analyze - Opportunity Goal Define - - - - Pilot gap - pilot Section 30 Analyze Section 40 Develop Solutions - Pilot gap - gap Pilot - Test site - pilot
52 Pilot FMEA, EMEA tool system,, Plan Element Next Page Risk Assessment,, -, Risk Assessment EMEA,, FMEA :,, (,, ) (,, ) / control -,, data -, /,, -, -,,, -, -,?, - /, - (10) solution benefits in 40 -,,, -, milestone, * : * : : : :
EMEA* * EMEA: Error Mode and Effect Analysis FMEA: Failure Mode and Effect Analysis # Error Mode Effect Analysis (EMEA) : (Degree of Effort) - (Error) (Cause) (Effect) (Process Step) outbound - - -, -, 421 - - 4 3 9 EMEA pallet - - pallet pallet - -, 422 -, 1 1 6 pallet - FMEA* Example: Appeal Process 0+2=2 2+2=4 [ ] (Occurrence Prevention [Countermeasur]) - - - pallet Process -, Action Clerk Sales Rep Checks Incomplete Delay 5 8 1 40 Rushed upon Receipt Appeal Application Appeal Committee Incorrect Missing Members Bad Decision Bad Decision Delay 8 8 5 Sales Rep Untrained Entry error, Research error Business Travel Illness Vacation 8 5 10 Sign off Territory Manager Review by Territory Manager Schedule MF 1 9 1 RPN* 64 360 50 Create a list of official checks for Territory Manager JRK 2/16/99 RPN
Process Control System Process Description: Purpose -, Process Flow chart Position -, Step or Time - - - -, - Process Description Process Customer CCR Output Indicator(s) Process flow chart Checking Misc Information process Indicator Output Indicator DATE, Revision Description Input/ process/ output Indicator Charts and Graph Type Process Control System Control Limit Specs/ Target What to Check Checking When to Check Data : File Name: Process Customer: CCR: Output Indicator(s): Checking Item Frequency Responsibility Who Checks Contingency Plan Action req d for exception Misc Information Include - Abbreviations - Procedures - Remarks, Etc Implementation Timeline: Tasks #4 Process Touch Points: TMSD, Metro/District Managers and Local Field Personnel, Third party Provider and Procurement REV # DATE Revision Description By Appr :Process control system :,, / : : CCR : (deployment flowchart) :, (,,, ),, : Process control system control system,,,,, : :CCR, :Process control system,
53 Process Pilot test (replication) Pilot (Standardization),, 3 cycle time 3 cycle time $5,000,000 $1,000,000 job $1,000,000 air filters
54 Project Lessons Learned Team Evaluation Recognition Compile Categorize Communicate - feedback / document - -,,, teaming skill,, /, -, / - - - -, 6 - - task - -,, - - - /
1 Cp, Cpk, Pp,Ppk Process Capability Ratios: Cp The greater the design margin, the lower the Total Defects Per Unit(TDU) Design margin is measured by the Process Capability Index(Cp) Cp = (Maximum Allowable Range of Characteristic) (Normal Variation of Process -- Short Term) Note: It is often suggested that the short-term standard deviation should be corrected for certain theoretical bias(which may be present) prior to calculation of Cp, Cpk, Pp, or Ppk Such an adjustment is made by dividing the short-term standard deviation by C4, a constant used in conjunction with control charts The theoretical merits of such a corrective device can be mathematically argued; however, the practical implications of such argument is highly unwarranted when given a sufficient sample -3sst m 0 Process Width +3sst Cp = USL-LSL 3s st Design Width Z ST = ST 3 Cp Cp LSL T USL Note: Pp is based on the same equation as Cp with one exception: namely, Pp employs the long-term standard deviation(whereas Cp employthe shortterm standard deviation)
Cpk Process Capability Ratios: Cpk Cpk = Cp (1 - k) Note: Ppk is based on the same equation as Cpk with one exception: namely, Pp employs the long-term standard deviation(whereas Cp employs the short-term standard deviation) Where k is the proportion of the tolerance zone consumed by the static mean shift Example: Cp = 2, k = 25 k = T - m (USL-LSL)/2 Cpk = 2( 1-25 ) = 15 m 0 m 1 0 ppm LSL LSL 6s 6s st st T T 45s 45s st st USL USL 34 ppm
Pp Process Capability Ratios: Pp Note: Pp is based on the same equation as Cp with one exception: namely, Pp employs the long-term standard deviation(whereas Cp employs the short-term standard deviation) Pp = USL-LSL 3s lt Z LT = LT 3 Pp Pp Pp = (Maximum Allowable Range of Characteristic) (Normal Variation of Process -- Long Term) Short-term distribution displays only pure error, Ie, white noise only Mean is artifically centered on the target value(by virtue of the equation) Long-term distribution displays white and black noise In this instance, black noise are the nonrandom variations in the process mean which tends to expand the standard deviation In the instance of Pp, the mean is artifically centered on the target value(by virtue of the equation) -3slt m 0 Process Width Design Width +3slt LSL T USL
Ppk Process Capability Ratios: Ppk Ppk = Pp (1 - k) Note: Ppk is based on the same equation as Cpk with one exception: namely, Pp employs the long-term standard deviation(whereas Cp employs the short-term standard deviation) Where k is the proportion of the tolerance zone consumed by the static mean shift Note: The k value is computed the same way as in the Cpk metric Again, the only difference is that Cpk utilizes the short-term standard deviation whereas Ppk employs the long-term standard deviation k = T - m (USL-LSL)/2 Long Long Term Term Short Short Term Term m 0 m 1 Long Long Term Term with with Static Static Offset Offset LSL LSL T T USL USL
2 Sigma Zlt 1 Pp Z Example : LDPELB7500 MT Long Term Data LSL T=195 USL LSL 17 Z T=195 USL 22 Pp = USL - LSL 6* slt Zlt = 3*Pp = 222 = 074 2 Ppk Z 232% 072% LSL Z T=195 USL Ppk = Min ( = 066 USL - X bar 3 * slt, X bar-lsl 3 * slt ) 17 X bar =192 slt = 011263 22 17 X bar =192 1 Z Zlt = 3*Ppk = 198 22 P(X<LSL) + P(X>USL) = 00304 00304 304% Z 0 1875 Zlt = 1875 Z
3:DMAIC Package Define Measure Analyze Improve Control CCRCTQ, CCR Baseline Sigma 1Pilot 1 Y=f(X)=X1+X2+X3--- Pilot Activity BAP(APF Output 1Process 1 Idea Pilot B) Input,,Output ( ), 2 Activity 1 2 Quin-win (OD) Idea EMEA,FMEA Team Charter 2 Quin-win 2Process 2 3,Fault Tree 3 Top Down Charting, 3 PilotSolution Sigma, Non- value Added Special Cause 4 Quick-win Quick-win Time Line 3 4 Sigma, 3 (VOC) CCR Sigma, CCR 1 1Input,Process,Output (Z,DPMO,RTY-Nomal) 4 1Data 1 1Process Control System Kano 2Process Map 5 indicator 2Quick-winProcess 2 Story Board Process Map 2Pilot CTQ 3Quick-win 2Operational Definition Map 3, 3 Key 4 4CCR,CTQ 3Quick-win 3 4 4 deliverables (VOC) 4Data 4 5Story 4 Board 5Pilot CCR Format 6Change Map 6 Kano 5Process 7 4 6 1 CTQ 1 / 1 Vital Few 1 DOE 1 / 2 () () DOE, Check 2 Data 2Quick-win, Point CTQ 3 () 20 Data 3() R 2 3Special Cause2 Breakthrough? CTQ 4 4 2CTQ 4 Scope () DOE 3 Red-X 5 5 3 () (), 4 Follow up & Auditing Daily:BB/ ( Issue ), Weekly: (,,, ) () X-R, Monthly: (,Transition, ) 4
Define Measure Analyze Improve Control CCRCTQ, CCR Baseline Sigma 1Pilot Y=f(X)=X1+X2+X3--- Pilot Activity BAP(APF 1 Idea Pilot B),, 2 Activity 1, Idea EMEA,FMEA Team Charter Quin-win 2Process Anderson s Regression 2 Special 3 DOE Top Down Charting, cause PilotSolution Quick-win Vital-Few Sigma, Non- value Added Sigma 4 Quick-win Time Line 3 3 (VOC), CCR 1 1Input,Process,Output 1Process Control System Kano 2Process T-test Map indicator Story Board T-test CTQ -Square 2Pilot ANOVA 3Quick-win 2Operational Definition -Square ANOVA, 3 Key 4 4CCR,CTQ 3Quick-win DOE DOE 4 deliverables (VOC) 4Data 4 Logistic Regression 5Pilot Logistic Regression CCR Format Regression Co-Relation Regression Co-Relation 6 Kano 5Process 7 4 6 Check Point 1 CTQ 1 / 1 Vital Few 1 DOE 1 / 2 () () DOE, 2 Data 2Quick-win, CTQ 3 () 20 Data 3() R 2 3Special Cause2 Breakthrough? CTQ 4 4 2CTQ 4 Scope () DOE 3 Red-X 5 5 3 () (), 4 () X-R 4