Six Sigma 12 Steps (六西格玛的12个步骤)

MethodologyOverviewPhaseII:ProcessMeasurementPhaseIII:ProcessAnalysisPhaseIV:ProcessImprovementPhaseV:ProcessControlSixSigmaPhasesPhaseI:ProcessDefinitionMAICDSUPPLIERProcessStep1Processstep2CUSTOMERProcessstep3RECORDDEFECTSRECORDDEFECTSRECORDDEFECTSRECORDDEFECTSRECORDDEFECTSMEASURECONTROLIMPROVEANALYZECONTINUOUSIMPROVEMENTMODELPROCEDURETIMEGOALTRENDCHARTACTIONPLANACTIONNAMEDATEVitalFewIdentificationContinuousImprovementThefollowingvisualillustrateshowthesystemcanworkDEFINESixSigmaBreakthroughStepsDefineMeasureAnalyzeImproveControlStep1-SelectOutputCharacteristic-IdentifyProcessInput/OutputVariablesStep2-DefinePerformanceStandardsStep3-ValidateMeasurementSystemStep4-EstablishProcessCapabilityStep5-DefinePerformanceObjectivesStep6-IdentifyVariationSourcesStep7-ScreenPotentialCausesStep8-DiscoverVariableRelationshipsStep9-EstablishOperatingTolerancesStep10-ValidateMeasurementSystemStep11-DetermineProcessCapabilityStep12-ImplementProcessControlsSixSigmaMethodologyStep1-DefinePhaseYouhavetohaveareasontodoaprojectandthatreasonshouldbesomethingthatisbotheringacustomer.Ifyoucan’tidentifywhatisimportanttoworkon,don’tworkonanything.Ifyoucan’tputanameonityoucan’tdoit.Ifyoucan’tdefinetheunitsofmeasure-becareful!Youwillenduphavingto:Definetheunitsofmeasureandcreateameasuringsystemfortheissueatthebeginningofyourproject.Step2-MeasurePhaseIfyoudon’tknowwhatmakessomethinggoodorbad,howcanyoumeasuretheproblemintermsofdefects(DPMO)?Ifthedefinitionofwhatmakesadefect(thespecification)isnotinthesameunitsofmeasureastheselectedYfromstep1,youhavethewrongspec.ThedefinitionneedstobecrystalclearandbriefSelecttheOutputCharacteristic(the“Y”)Whatistheproblem?Whatisadefect?DefinethePerformanceStandardofthe“Y”Whatisthespecificationorlevelofperformancerequiredbythecustomer?SixSigmaMethodologyStep3-MeasurePhaseIfyoucan’tmeasuresomething,howdoyouknowwhereyouare,whereyouhavebeen,orwhereyouaregoing?Ifyourmeasurementsystemisincapable,STOPandFIXITbeforeproceeding.Note:MeasurementSystemAnalysis(MSA)mustbedoneonanyandalldefectsyouwishtocount(watchoutifyouhaveavisualinspectionof6characteristics)Step4-MeasurePhaseIknowwhatIwanttofix(step1),whatdefinesitasgoodorbad(step2),andIcanaccuratelymeasureit(step3)so:NowIcansayhowmyprocessisperformingintermsoflongandshorttermZandDPMOwhichismybaselinedefectrate.Iknowcanimprovefromhere.ValidatetheMeasurementSystemCanweaccuratelyandconsistentlymeasuretheY?Canyoutrustyourdata?EstablishProcessCapabilityWhatisourdefectratetoday?SixSigmaMethodologyStep5-MeasurePhaseWheredoIwanttobeattheendofmyproject?Thereshouldbeastretchgoalwithlogicbehinditsselection.Theobjectiveshouldmakeitcleartoallatwhatpointtheprojectwillbeconsideredcompleteandsuccessful.Step6-AnalyzePhaseDefinePerformanceObjectivesGiventhecurrentdefectrate,howmuchshouldweexpecttoimprove?IdentifyVariationSources(“Xs”)Whataresomepotentialcausesofthedefect?Listallpotentialinputs(KPIV,X)thatcouldaffectmyoutput(KPOV,Y).Thisisfillingthetopofthefunnel.Sometoolsincludefishbone,C&Ematrix,FailureModesandEffectsAnalysis(FMEA),detailedprocessmaps.Thinkoutsidethebox-ifyoufixitthesamewayyoudidlastyear,youwillhavethesameproblemyouhavenow.SixSigmaMethodologyStep7-AnalyzePhaseStep8-ImprovePhaseScreenPotentialCausesCanIfigureoutwhichofthepotentialcausesareactuallyaffectingmydefectrateusingstatisticaltests?ScreeningisdoneusingGraphicalToolsandHypothesisTeststoidentifyandprovewhicharethevitalX’s.ThisisthemiddleofthefunnelforprojectswithvariablerelationshipsbetweenX’sForsomesimplerprojectswheretherearenovariablerelationshipsbetweenmyX’s(orIhaveonly1X)thisisalsostep8–thebottomofthefunnel,thefinalvitalX’sDiscoverVariableRelationshipsNowthatIknowwhichonesactuallycausethedefect,howshouldtheybesetsotheyproducetheleastamountofdefects?EvaluateifanyofmyvitalX’sareaffectedbyotherVitalX’s.ThisisprimarilydonethroughtheuseofDesignofExperiment(DOE)formoderatetocomplexprojects.DOEisanexperimentalprocessthatdeterminestheimpactofmultipleinputsonaselectedoutput.Thisleadstothebottomofthefunnel,IknowwhichX’saffectmyYandhowtheyaffecteachother.Transactional-useOFATapproach.SixSigmaMethodologyStep9-ImprovePhaseStep10-ImprovePhaseIknowwhichX’sareimportant.WhatrangeofsettingscanIusetomakegoodproduct?InthecaseofavariableX(e.g.PSIonanairfeed),Ihavetoprovideasettingtolerance.EstablishOperatingTolerancesCanIestablishatoleranceformyinputsthatwillstillyieldalowdefectratetoleavemyselfsomeflexibilityandfacilitatecontrol?ValidateMeasurementSystemNowthatIhaveimproved,canIstillmeasurethedefectorthecausesgoingforward?InthecaseofavariableX(e.g.PSIonanairfeed),Ineedtovalidatethatitcanbemeasured(avitalXMSA).Also,ImighthaveimprovedmyYsomuchthatIcannolonger“read”myprocess,andmayhavetoimprovemygagesystemtotrulymeasurewheremyimprovementshavebroughtme.SixSigmaMethodologyStep11-ImprovePhaseStep12-ControlPhaseDetermineProcessCapabilityWhatisthedefectratenowthatIhaveimplementedtheimprovements?ThismeasuresthecapabilityofcontrollingmyXswhereIhavesetthem.Thisisalsothetimewhenwedetermineformalresultsbycom