SlideShare a Scribd company logo
A brief introduction to   Six Sigma - Data driven problem solving

June 27, 2012




Villads Haar Jakobsen, WTG Platform and Diagnostics




                                        1
Agenda



  • What is it actually

  • Purpose of Six Sigma

  • Example of a Six Sigma project organisation

  • What about all these belts?

  • Different Six Sigma methodologies

  • DMAIC

  • DFSS

  • Where could Six Sigma be used at DONG Renewables



                                        2
What is it actually?

                                           Six Sigma: A Metric
                                               The "six Sigma level" of a process can be used to express its capability
                                                 How well it performs with respect to customer requirements
                                                 Defects per million opprotunities




                                                                                       Six Sigma: A method
                                                                                           A well defined process and tool kit used for:
                                                                                             Product & service improvements
                                                                                             Design of products and services




        Six Sigma: Symbol, value, benchmark or goal
           Greek letter which defines standard deviation in statistics
           Standard deviation is a measure for spread/variation
             The average distance of the data points to the mean




                                                               3
Purpose of Six Sigma



    To reduce variation
    To work with continous improvements
    To reduce cost of poor quality / reliability
    To make decisions based on statistical basis


                                                    Release the full potential




                                        4
Example of a Six Sigma project organization




                              5
What about all these belts??
 White belt
   For managers
   What can we gain from six sigma projcts
   How do we understand the language of six sigma


 Yellow belt
   Know the stages in DMAIC
   Understanding of the 7 basic tools and some statistics


 Green belt
   Know about the basic statistical tools used in six sigma
   Six Sigma project manager


 Black belt
   High level of statistical tools used in six sigma
   Six sigma project manager
   Coaching green belts


 Master black belt
   Extensive experience from many six sigma projects




                                                               6
Different Six Sigma methodologies



                                 Six
                                Sigma

                 DMAIC                      DFSS
              Improvement methodology       Design Methodology




                                                DMADV




                                        7
DMAIC   Process flow and objectives




                         8
DMAIC                                     Examples of tools




  •   Problem statement       • Pareto                • FMEA                 • FMEA                  • Audit plan
  •   Process map             • Histogram             • Histograms,          • Pugh matrix           • SPC
  •   Fishone diagram         • Fishbone                Boxplots, Multi-     • Project               • Hypothesis testing
  •   Voice of the customer   • SIPOC                   vari charts, main      inplementation
  •   CTQ's                   • C&E matrix              effects plots,       • DoE
                              • Data collection         interaction plots,   • EVolutionary
                                plan                    etc.                   Operations (EVOP)
                              • Data quality (MSA)    • Hypothesis tests     • Process mapping
                              • Capability analysis          • T-tests       • Capability analysis
                              • Process stability            • ANOVA
                                (control charts)      • Regression
                                                        analysis




                                                               9
What is the project                                                   D
                                                                  Define



      Problem statement
      Define the goals
      What is the cost of the problem
      Who are the stakeholders
      What are the customer requirements
         Ojective statement




                                  How does the existing process work?

                                   How is the process and the flow


   Project charter




                                     10
Baseline and capability                       M
                                            Measure




   What is the current performance?

      Understand the process behaviour
      Which factors influence the output
      Verify the mesurement system
      Data collection
      Look for patterns
      Calculate the capability




                                       11
Potential root causes                                     A
                                                        Analyze




    What are the key root causes?

     Identify sources of variation
     Determine the critical process
      parameters
     Develop and confirm theories
      using data



             y = f (x1, x2, x3 . . . xn)
                       Critical Xs

        What possible solutions have been identified?

         Models with the highest explanatory power


                                       12
Developement of solutions and implement                                         I
                                                                            Improve


  Evaluate the solutions and optimize choosen solution

   Which solution provides the best output for the customer
      Cost benefit analysis
   Perhaps it is necessary with additional experiments         Quality/reliability
   Optmize chosen solution

                          Implement the solution
                                                         Time                         Cost
                           Risk management
                               Contingency plan
                           Make a pilot project and compare with initial data
                           Implement the solution
                           Calculate performance (capability)




                                      13
Sustainable solution                                       C
                                                         Control



  How do we keep the benefits of the new solution

   Optimize and refine solution
   Make a follow up plan for the implemented solution
   Monitor and control




                                     14
DFSS




                                              Concept and           Detailed product
  Charter                CTQ's                                                               Prototype
                                              design selection      design
  • Kano/survey          •   Brainstorming    • FMEA                • Robust design          •   Reliability test
  • Project management   •   FMEA screening   • Reliability                • Taguchi         •   FMEA
    tools                •   DoE                     • Robustness   • DoE                    •   Simulation
  • Project selection    •   QFD              • Risk analysis       • Specification design   •   SPC
  • Benchmarking         •   Benchmarking     • Pareto analysis     • Work design            •   Control plan
  • SIPOC                •   Process map      • Gap analysis        • Machine design
  • C&E matrix                                                      • Engineering design
  • CTQ matrix




                                                      15
Examples of areas where Six Sigma could be used at DONG



 Developement/implementation
                                            Within warranty phase          Out of warranty phase
           phase
• Choose optimum supplier /            • Concerns about the life of   • Optimize settings
  component                              components
                                                                      • RCA
• Evaluation of reliability            • Compare performance
                                                                      • Improve processes
• Quality inspection



                              • When working with experiments (DoE)

• RCA                         • Life tests – Robustness and Reliability

• Improve processes
• RCA

• Improve processes
                                                       16
Questions?




             17
Back up – Bathtub curve and the Weibull distribution


 λ
  Infant mortality             Normal life phase         Wear out phase
  Decreasing failure rate      Constant failure rate     Increasing failure rate



     β>5




              β<1                             β=1                   β>3




                                                                        Time
               Up to "5" yrs                           ≈ 20 ?



                                                 18
RCA – Analysis of life data




                              19

More Related Content

PDF
Flevy.com - Six Sigma Basics
PDF
Integrated methodology for testing and quality management.
PDF
Six Sigma
PPT
Problem Solving Methodology 2011 - 2014
PPTX
BA conf presentation 2010
PPT
01 introduction
PPTX
Lean vs-six-sigma
PDF
Simulation in logistics - SIMANDO
Flevy.com - Six Sigma Basics
Integrated methodology for testing and quality management.
Six Sigma
Problem Solving Methodology 2011 - 2014
BA conf presentation 2010
01 introduction
Lean vs-six-sigma
Simulation in logistics - SIMANDO

What's hot (20)

PDF
BABOK 2 Tasks & Techniques
PDF
Lean Six Sigma Green Belt Services Certification Brochure
PDF
Factory performance optimization
PDF
Logistic systems simulation (a presentation by SIMANDO)
PDF
TLMC Corporate Presentation-v 1.2
PPT
Yellow belt training program
PPT
D07 Project Charter
PDF
Guide to New Product Development (NPD)
PPT
01 intro
PDF
Lean Six Sigma Project Management Oct11th2010[ (1)
PDF
Babok2 Big Picture
PDF
PRODUCT DEVELOPMENT METHODOLOGY
PDF
it4116_04_scampi
PDF
NG BB 13 Voice of Customer
PDF
8D Training Material From VDiversify.com | 8D Training Material PDF Free Down...
PDF
293312230-Lean-Six-Sigma-Students.pdf
PPTX
Basis of Estimate for IT Services
DOCX
Ashwin Kumar_MSE 697 Final
PPT
Digital Six Sigma vs Directed Innovation
PPTX
Learning Investments: Making Workforce Future Ready
BABOK 2 Tasks & Techniques
Lean Six Sigma Green Belt Services Certification Brochure
Factory performance optimization
Logistic systems simulation (a presentation by SIMANDO)
TLMC Corporate Presentation-v 1.2
Yellow belt training program
D07 Project Charter
Guide to New Product Development (NPD)
01 intro
Lean Six Sigma Project Management Oct11th2010[ (1)
Babok2 Big Picture
PRODUCT DEVELOPMENT METHODOLOGY
it4116_04_scampi
NG BB 13 Voice of Customer
8D Training Material From VDiversify.com | 8D Training Material PDF Free Down...
293312230-Lean-Six-Sigma-Students.pdf
Basis of Estimate for IT Services
Ashwin Kumar_MSE 697 Final
Digital Six Sigma vs Directed Innovation
Learning Investments: Making Workforce Future Ready
Ad

Viewers also liked (11)

DOCX
Dimensión curricular 2016 i.e. san pablo
PDF
Personal injury lawyer in florida
DOCX
Act 8. tallerpractico10. edgar alberto patiño
PDF
Coresta16: Exhaled aerosol properties in a room following use of ecigs
PDF
Join Us. US Mission March 2016
PPT
MCO 2011 - Slide 29 - R.A. Stahel - Mesothelioma
PDF
오픈 데이터, 스마트 시티 그리고 인공지능
PDF
5 Minutes to a Better Resume
PPTX
Teen thriller films research
Dimensión curricular 2016 i.e. san pablo
Personal injury lawyer in florida
Act 8. tallerpractico10. edgar alberto patiño
Coresta16: Exhaled aerosol properties in a room following use of ecigs
Join Us. US Mission March 2016
MCO 2011 - Slide 29 - R.A. Stahel - Mesothelioma
오픈 데이터, 스마트 시티 그리고 인공지능
5 Minutes to a Better Resume
Teen thriller films research
Ad

Similar to A brief introduction to Six Sigma (20)

PPTX
Zero defect
PDF
Six Sigma Session For Production And Project Team By Lt Col Vikram Bakshi
PDF
A study on six sigma techniques and its application in reduction of seat reje...
PDF
6 Sigma
 
PPT
Profolio
PDF
Six sigma as foundation to cmmi
PPT
Six sigma-basics
PPTX
Lean six-sigma-project-charter-template
PDF
Aqm Programme Six Sigma
PPT
Six sigma overview
PDF
Six Sigma Presentation
PPTX
Final Six Sigma
PDF
Mpc mtcp six sigma [compatibility mode]
PDF
Integrating Six Sigma and Lean Manufacturing the Challenges & Benefits
PDF
Integrating Six Sigma and Lean Manufacturing the Challenges & Benefits
PPT
6sigmaweb
PPTX
Six sigma
PDF
Lean Six Sigma Yellow Belt Certification Brochure
PPT
Lean Six Sigma
PPT
Six Sigma
Zero defect
Six Sigma Session For Production And Project Team By Lt Col Vikram Bakshi
A study on six sigma techniques and its application in reduction of seat reje...
6 Sigma
 
Profolio
Six sigma as foundation to cmmi
Six sigma-basics
Lean six-sigma-project-charter-template
Aqm Programme Six Sigma
Six sigma overview
Six Sigma Presentation
Final Six Sigma
Mpc mtcp six sigma [compatibility mode]
Integrating Six Sigma and Lean Manufacturing the Challenges & Benefits
Integrating Six Sigma and Lean Manufacturing the Challenges & Benefits
6sigmaweb
Six sigma
Lean Six Sigma Yellow Belt Certification Brochure
Lean Six Sigma
Six Sigma

A brief introduction to Six Sigma

  • 1. A brief introduction to Six Sigma - Data driven problem solving June 27, 2012 Villads Haar Jakobsen, WTG Platform and Diagnostics 1
  • 2. Agenda • What is it actually • Purpose of Six Sigma • Example of a Six Sigma project organisation • What about all these belts? • Different Six Sigma methodologies • DMAIC • DFSS • Where could Six Sigma be used at DONG Renewables 2
  • 3. What is it actually? Six Sigma: A Metric  The "six Sigma level" of a process can be used to express its capability  How well it performs with respect to customer requirements  Defects per million opprotunities Six Sigma: A method  A well defined process and tool kit used for:  Product & service improvements  Design of products and services Six Sigma: Symbol, value, benchmark or goal  Greek letter which defines standard deviation in statistics  Standard deviation is a measure for spread/variation  The average distance of the data points to the mean 3
  • 4. Purpose of Six Sigma  To reduce variation  To work with continous improvements  To reduce cost of poor quality / reliability  To make decisions based on statistical basis Release the full potential 4
  • 5. Example of a Six Sigma project organization 5
  • 6. What about all these belts??  White belt  For managers  What can we gain from six sigma projcts  How do we understand the language of six sigma  Yellow belt  Know the stages in DMAIC  Understanding of the 7 basic tools and some statistics  Green belt  Know about the basic statistical tools used in six sigma  Six Sigma project manager  Black belt  High level of statistical tools used in six sigma  Six sigma project manager  Coaching green belts  Master black belt  Extensive experience from many six sigma projects 6
  • 7. Different Six Sigma methodologies Six Sigma DMAIC DFSS Improvement methodology Design Methodology DMADV 7
  • 8. DMAIC Process flow and objectives 8
  • 9. DMAIC Examples of tools • Problem statement • Pareto • FMEA • FMEA • Audit plan • Process map • Histogram • Histograms, • Pugh matrix • SPC • Fishone diagram • Fishbone Boxplots, Multi- • Project • Hypothesis testing • Voice of the customer • SIPOC vari charts, main inplementation • CTQ's • C&E matrix effects plots, • DoE • Data collection interaction plots, • EVolutionary plan etc. Operations (EVOP) • Data quality (MSA) • Hypothesis tests • Process mapping • Capability analysis • T-tests • Capability analysis • Process stability • ANOVA (control charts) • Regression analysis 9
  • 10. What is the project D Define  Problem statement  Define the goals  What is the cost of the problem  Who are the stakeholders  What are the customer requirements  Ojective statement How does the existing process work?  How is the process and the flow Project charter 10
  • 11. Baseline and capability M Measure What is the current performance?  Understand the process behaviour  Which factors influence the output  Verify the mesurement system  Data collection  Look for patterns  Calculate the capability 11
  • 12. Potential root causes A Analyze What are the key root causes?  Identify sources of variation  Determine the critical process parameters  Develop and confirm theories using data y = f (x1, x2, x3 . . . xn) Critical Xs What possible solutions have been identified?  Models with the highest explanatory power 12
  • 13. Developement of solutions and implement I Improve Evaluate the solutions and optimize choosen solution  Which solution provides the best output for the customer  Cost benefit analysis  Perhaps it is necessary with additional experiments Quality/reliability  Optmize chosen solution Implement the solution Time Cost  Risk management  Contingency plan  Make a pilot project and compare with initial data  Implement the solution  Calculate performance (capability) 13
  • 14. Sustainable solution C Control How do we keep the benefits of the new solution  Optimize and refine solution  Make a follow up plan for the implemented solution  Monitor and control 14
  • 15. DFSS Concept and Detailed product Charter CTQ's Prototype design selection design • Kano/survey • Brainstorming • FMEA • Robust design • Reliability test • Project management • FMEA screening • Reliability • Taguchi • FMEA tools • DoE • Robustness • DoE • Simulation • Project selection • QFD • Risk analysis • Specification design • SPC • Benchmarking • Benchmarking • Pareto analysis • Work design • Control plan • SIPOC • Process map • Gap analysis • Machine design • C&E matrix • Engineering design • CTQ matrix 15
  • 16. Examples of areas where Six Sigma could be used at DONG Developement/implementation Within warranty phase Out of warranty phase phase • Choose optimum supplier / • Concerns about the life of • Optimize settings component components • RCA • Evaluation of reliability • Compare performance • Improve processes • Quality inspection • When working with experiments (DoE) • RCA • Life tests – Robustness and Reliability • Improve processes • RCA • Improve processes 16
  • 18. Back up – Bathtub curve and the Weibull distribution λ Infant mortality Normal life phase Wear out phase Decreasing failure rate Constant failure rate Increasing failure rate β>5 β<1 β=1 β>3 Time Up to "5" yrs ≈ 20 ? 18
  • 19. RCA – Analysis of life data 19

Editor's Notes

  • #2: ViljaDONG siden april 2012 En af de gamle ;-)Jeg vil fortælle lidt om Six Sigma Jeg er det man kalder blackbelt – Men bare rolig, jeg er ikke farlig ;-)
  • #3: Agendaen ser således ud Den dækker over områder, der tilsammen, skal give et godt indblik i Six Sigma Jeg vil komme ind på …………Hvis der er spørgsmål undervejs, så spørg løs
  • #4: Der ernokligeså mange meningeromhvad 6S er, som der er folk, der eruddanet I 6S.Jegharsamletnogleafdem man høreroftest:Symbol, værdieller et benchmarkDeter et græskbogtsav, der bruges I statistikken for std. afv (somer den gennemsnitlige distance for alledatapunktertil mean)For andreerdet et mål der arbejdeshenimod6S er +/- 3 stdafv. – detsvarertil at allepunkter ligger indenfor 99,7% aftoleranceneller 3,4 defekterprmiomulighederFor migselvog mange andreerdet en metodeog en værktøjskasseOgdetergodt med flereforskelligeværktøjer I kassen,Som en bekendtplejer at sige, hvis man kun har en hammer I værktøjskassen, såtror man at alt er et søm
  • #5: Formålet er at reducere variationen vha datadrevet problemløsningDet ligger langt fra SPT, HUA og beslutninger truffet på baggrund af middelværdier alene (eksempel med lastbilkørsel)At arbejde med kontinuerlige forbedringerAt reducere omkostningerne for dårlig kvalitet og performance ved at forbedre processer og/eller pålidelighedenDet der er forskelligt ved 6S ift. andre metoder, er den tyngde statistikken harMarty Feldman ville have været en god udøver af 6SHan ville kunne fokusere klart på arbejdet med de bløde værktøjer med det ene øje, samtidigt med at han fokuserede på de hårde statistiske værktøjer med det andet
  • #7: For at beskrive videns- eller erfaringsniveauet indenfor 6S bruger man de bæltfarver, man også bruger indenfor kampsport, det er ret smart, så ved folk intuitivt hvad det er for et niveau den man taler med har.Yellow - Kender DMAIC modellen, de 7 grundlæggende værktøjer og har lidt kendskab til statistik (normalfordelingen og dens parametre)Green - Har styr på DMAIC modellen og de statistiske værktøjer der ofte bruges. Kan bruges som 6S projektlederBlack - Stort kendskab til statistiske værktøjer og erfaring med de bløde værktøjer og design af eksperimenter. Kan coache green belts.Master black - Gammel gubbe der har prøvet det heleWhite - Typisk ledere og andre der fx sidder i styregruppen, eller folk der har behov for at vide hvad de 6S folk er for nogle og hvad de kan bruges til.
  • #8: Der er to forskellige veje i 6S – en til ændring af noget eksisterende og en til design af noget nyt.Processer, produkter eller services.Til forbedring bruges DMAICTil design af noget nyt bruges DFSSModellen man bruger er DMADVNogen af værktøjerne er de samme, men ikke alle.Jeg vil gå mere i dybden med DMAIC, som er den mest kendte og brugte metode, til sidst fortæller jeg lidt om DFSS.
  • #9: Her er en illustration af procesflowet i DMAIC processen.Der er 3 projektfaser (MAI) og to &quot;aktiviteter&quot; (DC).D Definere problem og formålM Hvad er det der skal forbedres og kan vi måle detA Analyser processen og definer de faktorer der har indflydelseI Identificere og implementere forbedringerC Sikre at forbedringer er varige
  • #10: Jeg har en slide der viser nogle af de værktøjer der ofte bruges under de forskellige faser – det er selvfølgeligt ikke noget der ligger helt fast.I kender sikkert allerede mange af dem.Der er generelt to spor, et dataspor, hvor man bruger statistikken, og et processpor, hvor man bruger de blødere værktøjer (C&amp;E matrice, FMEA, fishbone, etc.)EVOP Optimeringsteknik, der bygger på DoE. Man eksperimenterer og finder den ved der fører til størst forbedring – iterativ proces.Jeg vil ikke gennemgå alle værktøjerne, hvis man har lyst til at se nærmere på hvad det er for værktøjer kan man hente præsentationen fra Pondus.
  • #11: I Define skal man:beskrive hvad problemet er og hvad det er man gerne vil forbedre eller ændrePrøve at kvantificere det i kroner og ørerIdentificere interessanterIdentificere hvad det er brugerne/kunderne gerne vil haveMan skal have et overblik over hvordan processen/produktet/servicen ser ud på nuværende tidspunktDe ting beskrives i et project charter, som så skal forelægges styregruppen – så får man forhåbentlig et GO til at gå videre
  • #12: I M-fasen skal man:Have en grundig forståelse for hvordan processen fungererHvad er det for faktorer og output der er i spilIdentificer kritiske processerVerificere målesystemet (vigtigt, dur ikke med beskidte eller upålidelige data)Lave en dataopsamlingsplanog opsamle dataBeregne den nuværende kapabilitetPræsenteres og få en godkendelse af styregruppen til at gå videre
  • #13: I A-fasen analyserer man data og undersøger deres sammenhæng med processen.Man identificerer de faktorer der giver anledning til variationen i processenKvantificerer hovedeffekter og interaktionerUdvikler modeller og be- eller afkræfter teorierIdentificerer mulige løsningerFår en godkendelse af styregruppen til at gå videre
  • #14: I I-fasen evaluerer man løsningerne.Man optimerer måske (DoE + EVOP (optimere udbytte/processbetingelser))Man inddrager evt. brugere/kunderLaver en costbenefitOg vælger den endelige løsningMan implementere løsningen, det kan være man:laver en laver en beredskabsplan, hvis noget går galtMåske laver man et pilotprojektFuld implementeringBeregner kapabilitetenDer laves en kontrol plan til brug i C-delen.Får en godkendelse af styregruppen til at gå videre
  • #15: I C-delen følger man løbende op på om forbedringerne er varige eller om der skal gøres yderligere.Efter nogen tid beregner man igen kapabilitetenDer sker måske et par småjusteringerDer laves en plan til løbende opfølgning for at sikre de varige forbedringer
  • #17: Udviklings-/implementeringsfasenSammenligning af leverandører, komponenter eller systemer (hypotesetest – er der en væsentlig forskel)QA (SPC, kapabilitetsanalyse, MSA, etc.)GarantiperiodenVed bekymring af restlevetiden for en komponent kan man lave tests (DoE+ALT)Sammenligne performance (variansanalyse)Perioden efter garantiudløbOptimere yield (EVOP)RCA (DMAIC)Forbedre processer, kontroller, beredskab, etc.
  • #18: Det var det jeg havde med, jeg håber det har givet et lille indblik i 6S-verdenen.Er der nogen der har nogle spørgsmål?
  • #19: 6S og pålideligheds-engineering passer rigtigt godt sammen, begge områderlænerisg meget op af mange af de samme statistiske værktøjer, og arbejder med datadrevne beslutninger.Så jeg har lige taget en slide med, der viser badekarskurven og nogle beta-værdier fra Weibull fordelingen.Som man kan så er der en rigtig stor forklaringsgrad mellem betaværdierne og faserne i badekarskurvenFasen for tidlige svigtβ &gt; 5 tyder på design fejl (høj β-værdi meget tidligere end designlevetiden = designfejl eller forkert drift)β &lt; 1: Der bliver bedre styr på produktionen og småfejl bliver rettet til Normal levetidsfasenβ = 1, det svarer til exponentialfordelingen. Det er normalt den der opgives af leverandører – man medregner aldrig den første fase, hvilket man bør gøre, medmindre man har været igennem et forløb robusthedstests og produktionstest.Udslidningsfasenβ &gt; 3, her har komponenten nået enden af sin levetid.
  • #20: 3 fejltyper på den samme komponent, hvoraf de to måske er afhængige af hinandenFjerner man den første fjerner man måske den anden, men det kræver lidt mere analyse at vurdere det