Lean 6sigma and DMAIC


Published on

I've been asked to put together a basic (and therefore relatively quick) introduction to Lean Six Sigma & DMAIC. While it’s not yet finished, I thought I would put it out there for people to comment on. Since the presentation is supposed to be training material there’s more text on the slides than I would prefer, but there are a few exercises and games to get the trainees involved.
I've put the PowerPoint version on my blog:

Published in: Business, Technology
No Downloads
Total views
On SlideShare
From Embeds
Number of Embeds
Embeds 0
No embeds

No notes for slide

Lean 6sigma and DMAIC

  1. 1. 20%80% of process defectsarise from 20% of theprocess issuesCapacity = WORK + WASTELean Six Sigma
  2. 2. What is Lean Six Sigma…2 successful management strategies for improving business processes combinedLEANLEAN 6σ6σ
  3. 3. Plan, Do, Check, ActPlan = Define + Measure +AnalyseCheck + Act = Control Do = ImplementCan be usedalongside DMAICmethodsTraditional PDCAmethodLean six sigmamethod
  4. 4. Benefits of Lean Six SigmaMotorola has reported over US$17 billion in savings as of 2006. Another early adopter ofLean Six Sigma is General Electric (introduced by Jack Welch). By the late 1990s, abouttwo-thirds of the Fortune 500 organisations had begun Lean Six Sigma initiatives with theaim of reducing costs and improving quality.
  5. 5. Lean 6σ ProjectKaizen blitzDifferent approachesFix specific problem orworkflow issue withinthe organisationKaizen=ContinuousimprovementLarge scale projectwork to find routecauses & fix hiddenand obvious problems
  6. 6. ConsPros Happens in sporadic bursts Root causes not thoroughlyanalysed to optimise the results Fixes the obvious issues at theexpense of learning about thelatent opportunities Kaizen is seductive & efficient. Delivers large scale results quickly Utilises the entire team’s collectiveinsight & experience Inspires curiosity about waste,defects and constraints tothroughputKaizenLean Six Sigma, or DMAIC, fills in the gaps that Kaizen fails to address.It supplies the analytical discipline necessary to thoroughly understandthe nature of processes and problems.Short, focused bursts of continuous improvement
  7. 7. DMAICKaizenJust do itOngoing quick hitsLow risk &investmentDiscover and deliver newvalue3 – 6 monthsSolve problems 2 – 4 monthsGreater $$ and / or riskScaleApproachProject / Risk relationship~ 1 weekStreamlineRemove waste
  8. 8. Lean Control
  9. 9. What is Lean Control…Lean is a philosophy and set of management techniques focused on continuous“elimination of waste” and smoothing work flows so that every process, task or workaction is made “value adding” when viewed from customer perspectiveHow much product does thecustomer really need?Anything more is wastedinventoryExtrastoredproductNeededstoredproduct
  10. 10. Examples of waste (muda)123456OverproductionInventoryMotionWaitingTransportationOver processingDefects7
  11. 11. ExerciseBrainstorming &negativeBrainstormingExamples of mudaIn pairs, use brainstorming and negative brainstormingtechniques to find a solution for one example of one of thedifferent types of muda in your workplacesOne of the toolsused in the DMAICmethodInventoryMotionDefectsBrainstormingNegativeBrainstorming
  12. 12. Streamlining the processIn addition to looking at wastes, lean also looks at streamlining the process using push /pull processes and kanban systems to ensure the product is made just in time with aslittle work in progress (WIP) as possibleCustomer Service Production Quality Control LogisticsInformation flowMaterialflowMaterialflowMaterialflowInformation flow Information flowPush process
  13. 13. Chocolate experimentExperiment 1Each person will be givena role / workstation.As part of the work youmust pass the product onto the next person in thechain.After 2 minutes counthow many products arestill inside the productionprocess (WIP) and howmany have been given tothe customerSupplierProduction 1ShippingQuality controlProduction 2CustomerProduction 3Watch dog
  14. 14. Chocolate experiment 1 – part 1Production 3SupplierSupply a singlechocolate & orderto production 1every secondProduction 1Production 2Count eachchocolate thensend a batch of 4 toproduction 3Supply a batch of 3chocolates toproduction 2 every3 secondsWrap chocolate inpaper, then send abatch of 3 toQuality controlShippingQuality controlSupply 1 wrappedchocolate to thecustomerCheck eachchocolate & pass 1to shipping eachsecond
  15. 15. Chocolate experiment 1 – part 2Watch dogCustomerAccept a delivery ofchocolate fromshippingMonitor the timetaken, and themovement of thepeople during theexercise. Count theWIP and goodsreceived by thecustomerCustomerProduction 2Production 1Production 3ShippingQuality controlSupplierWatch dogRoom layout
  16. 16. Watchdog feedbackPush productionWhat did you learn?How much work was in process (WIP)?Was creating batches efficient or did it slow the process down?Was the work station layout effective?How could we improve things?Learning aboutflow, push and pullproduction in Lean
  17. 17. Takt timeTaktzeit=cycle timedaTTT TdTime demand (customerdemand), e.g. [unitsrequired / day]TaNet time available to worke.g. [minutes of work / day]
  18. 18. ExampleIf Customer demand = 400 units/dayTo make product in 1 shift,takt time must = 1 part / minuteShift time = 8hrs (480mins)Break time = 30mins (2 x 15mins)Lunch break = 30minsTeam brief = 10minsMaintenance checks = 30minsNet Available Time to Work480 - 30 - 30 - 10 - 10 = 400minsIf Customer demand = 600 units/dayTo make product in 1 shift,takt time must = 1 part / 40secs
  19. 19. Chocolate experiment 2Experiment 2Try the simulation withthe layout of the room asshown…Send the chocolatesthrough the systemwithout batching.Note there are new tasksfor some roles!What happens now?CustomerProduction 3SupplierProduction 1Production 2ShippingQuality ControlWatch dogRoom layout
  20. 20. Chocolate experiment 2 – new rolesWatch dogCustomerOnly accept adelivery if the firstone arrives in 10seconds. 10 mustarrive within 1minuteDon’t accept if theyare wrapped!Monitor the timetaken, and themovement of thepeople during theexercise. Count theWIP and goodsreceived by thecustomerQuality controlDue tomaintenance issuesyou can now onlycheck a chocolateevery 8 seconds
  21. 21. Watchdog / customer feedbackPush productionWhat did you learn?Was there a bottleneck?Was this a push or a pull process?How could we improve the production flow?Was the customer satisfied?What about the company profits and losses?How could we find out what the customer really wants?Which processes added value and which steps had overprocessing wastes?Learning aboutinventory control,WIP, kanbanVoice of the customerValue added processes
  22. 22. Chocolate experiment 3Experiment 3Use kanban stations tocontrol the flow of goodsthrough the processFind out what thecustomer really wants(VOC) and compare it tothe voice of the process(VOP)See how a pull systemworks!CustomerProduction 3SupplierProduction 1Production 2ShippingQuality ControlWatch dogRoom layoutK KKKK KKOnly put a chocolate onto the next kanbanstation if the kanban station is empty!!!
  23. 23. Watchdog feedbackPush productionWhat did you learn?How much WIP was there in the process now?Where was the process being pulled from?Was the customer satisfied?Does VOC = VOP now?Is it necessary to keep ANY over processing steps in theproduction cycle?Learning aboutinventory control,WIP, kanbanVOC / VOPBusiness necessary steps &value added steps
  24. 24. Origins - LeanTaiichi Ohno, Shigeo Shingo and Eiji Toyodadeveloped the TPS system between 1948 and1975Lean derives from Toyota Production Systemor Just In Time Production, Henry Ford andother predecessorsToyota, Aichi, JapanDetroit, USA
  25. 25. Lean timelineTAYLORStandardisedwork / Time studyWHITNEYInterchangeablepartsGILBRETHProcess charts/ Motion studyFORDAssembly / LowlinesDEMINGSPCTQMTOYODA/OHNOTPS Just in time, nostock productionLEANLean used inmanufacturing1850 1900 1950 2000
  26. 26. Six Sigma
  27. 27. ExamplePareto principleThe Pareto principle is named after an Italian economist Vilfredo Pareto, who observed in1906 that 80% of the land in Italy was owned by 20% of the populationMicrosoft noted that byfixing the top 20% of themost reported bugs, 80%of the errors and crasheswould be eliminatedUniversity of Florence,Italy
  28. 28. What is 6σ…Six Sigma seeks to identify and remove these 20% of causes of defects and errors inbusiness processes. It uses a set of management methods, including statistics and createsa dedicated infrastructure of people within the organisation who are experts in thesemethods.Production linesPercentageyieldLines where there is asignificant defectTargetBy focussing resources onthe areas where significantdefects occur, we canstabilise and improveperformance!
  29. 29. ExerciseNormal distributioncurvesNormal distribution curvesA normal distribution of data means that most of the examplesin a set of data are close to the "average," while relatively fewexamples tend to one extreme or the other.Each person should take one of the targets and some sillyputty. Affix the target to the wall, and throw ten pieces ofputty at the target and see where they landWe are then going to use this data to plot our first statisticalgraph!Mean and Sigma
  30. 30. TargetTarget scoringScore your results basedon the diagram shownThen count up thenumber of times you hiteach ring of the targetand plot the results onthe graph512430
  31. 31. Sample target resultsTarget ringNoofhits24681001 2 3 4 5Paul SimonJackSarahWhat factors (inputs, Xs) affectthe outcome (output, Y)?How can you calculate the mean? =2.85
  32. 32. HistogramsNoofpeopleHeight in metresHistogram showingno of people ateach heightHeightinmetresMean height = 1.8mMinimum = 1.6mMaximum = 2.0mImagine performing a survey of everyone’s heightin a factory. The results would be similar to thegraph below
  33. 33. Normal distribution curveIf you draw a line over thetop of the histogram, youget what is known as anormal distribution curveNot all data sets will have graphs that look this perfect. Some will have flatcurves, others will be steep. Sometimes the mean will lean a little bit to one sideor the other. But all normally distributed data will have a similar "bell curve"shape.12Normaldistribution curveMean
  34. 34. The standard deviation σ is a measure of howmuch variation there is in the dataStandard deviationσStatistics show that 68 percent of thedata in this group falls within 1σ of themean (either side)34When the data is tightly bunched togetherand the bell shaped curve is steep, thestandard deviation is small.When the data is spread apart and the bellcurve is relatively flat, the standard deviationis largeSmallstandarddeviationLargestandarddeviationσ
  35. 35. Sigma levelsWe can measure the sigma level of a process bygraphing the number of defects / dayThe data should show normal distributionA process with a 1σ score will have 69% defects and 31%yieldTo have a 2σ score, there would be 31% defects and69% yieldσLevel%Defects%Yield12345669 3131 696.7 93.30.6 99.40.02 99.980.0003 99.995
  36. 36. But what is 6σ?The term "six sigma process" comes from the notion that if you can measure sixstandard deviations between the process mean and the nearest specification limit,practically none of the products will fail to meet the specification6σσ 2 3 4 5 6 σ- σ-2-3-4-5-6 σ μUpperspecificationlimit (USL)Lowerspecificationlimit (LSL)68% of the data
  37. 37. Six sigma methodologiesWe use the six sigma methodologies, then, to find the causes of the defects in ourprocesses to allow us to get a higher sigma score…Some causes of defects are obvious, others hidden…Traditional Quality costsTangibleEasy to measureHidden Quality costsIntangibleDifficult to measureInspectionWarrantyScrapReworkRejectsMore setupsExpediting costsLost salesLate deliveriesLost customer loyaltyExcess inventoryLong cycle timesCostly engineering changes
  38. 38. INPUTSXsOUTPUTSYsPROCESS IProcess parametersMost of the time, we know that we have a problem, although we are not alwaysexactly sure where. Using six sigma we can Define the problem exactly, and thenMeasure the inputs (X) and outputs (Y) of the processOnce we can identify the anomalous data we can Improve and Control it. IF theanomalous data is affecting costs!We Analyse this data to determine if our performance is reliable and consistent, itfollows a pattern, or if we have an unreliable or out of control process with significantlevels of defects and / or special events
  39. 39. Origins – Six Sigma 6σ6σ started in Motorola by Bill Smith 1986Illinois, USAIn 1920 Walter Shewhart showed that 3σ fromthe mean is the point where a process requirescorrection
  40. 40. DefineAnalyse ImproveControlMeasure?The DMAIC method
  41. 41. What is the DMAIC method?ImproveA structuredapproach to problemsolvingTeam led project workA Japanese fad thatwon’t last long?
  42. 42. Exercise15 word flip chartOne of the toolsused in the DMAICmethodWhat is the DMAIC method?In teams, choose one of the following topics, then using postits and a flip chart, write 15 words to describe:• What you think the DMAIC method is• What are the origins of DMAIC• Why DMAIC is important to businessTextTextText
  43. 43. The “D” (Define) in the DMAIC process focuses on selecting high-impact projects and understanding which underlying metric(s) willreflect project successThe M-A-I-C steps in DMAICcomprise a number of toolsaimed at understanding andcontrolling root causes.Projects create additional work, andthe management team must beconvinced that the project is worththe additional work required todeliver breakthrough results.A good DMAIC project is focussed and team driven
  44. 44. Who is who in Lean Six Sigma…Champions / SponsorUsually high up, who sponsors a Lean Six Sigma project through the business. The first fewLean Six Sigma projects may encounter some technical, political, or institutional obstaclesin the business environment. Its critical that someone is there to help the project and theproject members overcome any obstacles and resistance.Black BeltsDoes Lean Six Sigma "full time." Their entire work effort is focused on finding defects, whereverthey might be, and eliminating them from the business processes. They move from department todepartment, heading Lean Six Sigma projects.Green BeltsThis is a title for someone who is involved with a Lean Six Sigma project "part-time." They have ajob with normal duties, added involvement in a Lean Six Sigma project.
  45. 45. Brief overviewDefineAnalyseImproveControlMeasure Measure the process to determine current performanceDefine the project goals and customer deliverablesImprove the process by eliminating defectsAnalyze and determine the root cause of the defectsControl future process performanceDMAIC
  46. 46. DMAIC in Everyday LifeD M A I CMy baseline weightis 192 pounds, andmy target weight is182 pounds. Themeasurement erroris acceptablePack healthy snacksso we don’t use thevending machine.Pre-plan 1 eveningsnack to develop agood routineWeigh-in 1/wk andtrack the results. Ifweight creepsabove 185 pounds,log food intake andmake adjustmentsUsing the DMAIC, how wouldyou tackle a probleminvolving trying to loseweight?I want to loseten pounds…We take in 24% ofour daily calorieswith eveningsnacks, & another19% by hitting thevending machine atworkI want to improve my health. Reaching ahealthy body weight will increase my energylevel and prevent a number of problemsdown the road. The research I have donesays that I need to lose ten pounds.”
  47. 47. PillarsThe original Toyota ProductionSystem could be described as ahouse with pillars.JIT (just in time)Smooth out production /resource utilisationJidokaQualityproduction,defect freeLean controlRespect for peopleHEIJUNKA(levelled production)KAIZENSTANDARDWORKFlow, pullprocesses, wastereduction
  48. 48. Paper tower exerciseA4 sheet of paperUsing a single sheet of A4 paper only & no othermaterials, construct the tallest free standing structure, in5 minutesCreative thinkingWorking underpressure, planning,team working
  49. 49. Selecting a project teamOnce you have identifiedthat you have a problem,you need a team!All teams are only as strong astheir weakest member!Should all of the project team be processexperts?More team members means more time!It’s difficult to engage operatorsMost project teams end up being the same usualsuspects doing all of the work
  50. 50. 5-wordsWhat makes a good team player?2 teams must battle it out to see who is themost creative, innovate and show betterteam skills!Choose 5 words:1. Animal2. Mineral3. Famous place4. Food5. Clothing / JewelleryGive your words to the opposing team andthen in 15 minutes write a script containingthe 5 words you have been given. When thetime is up, be ready to perform!During the game, thinkabout the way you andyour team membersinteract. What was goodand what is bad?!
  51. 51. StoryboardsGood storyboards are essential!Why use it?To track data, decisions, andactions and create a graphicalor pictorial record of yourDMAIC project.
  52. 52. StoryboardsWhat does it do?• Facilitates decision making• Helps maintain forwardmomentum• Helps prevent rework• Provides a quick, visualsummary of a team’s work• Can also be used aspresentation materials.• Other employees can haveaccess to the improvementteam’s work.