《SixSigma12Steps六西格玛的12个步骤》由会员分享，可在线阅读，更多相关《SixSigma12Steps六西格玛的12个步骤（24页珍藏版）》请在金锄头文库上搜索。
1、MethodologyOverview1Phase II: Process MeasurementPhase III: Process AnalysisPhase IV: Process ImprovementPhase V: Process ControlSix Sigma PhasesPhase I: Process DefinitionMAICD2SUPPLIERProcess Step 1Processstep 2CUSTOMERProcessstep 3RECORDDEFECTSRECORD DEFECTSRECORD DEFECTSRECORDDEFECTSRECORDDEFECT
2、SMEASURECONTROLIMPROVEANALYZECONTINUOUS IMPROVEMENT MODELPROCEDURETIMEGOALTREND CHARTACTION PLANACTIONNAMEDATEVital Few IdentificationContinuous ImprovementThe following visual illustrates how the system can workDEFINE3Six Sigma Breakthrough StepsDefineMeasureAnalyzeImproveControlStep 1 - Select Out
3、put Characteristic - Identify Process Input/Output VariablesStep 2 - Define Performance StandardsStep 3 - Validate Measurement SystemStep 4 - Establish Process CapabilityStep 5 - Define Performance ObjectivesStep 6 - Identify Variation SourcesStep 7 - Screen Potential CausesStep 8 - Discover Variabl
4、e RelationshipsStep 9 - Establish Operating TolerancesStep 10 - Validate Measurement SystemStep 11 - Determine Process CapabilityStep 12 - Implement Process Controls4Six Sigma MethodologyStep 1- Define PhaseYou have to have a reason to do a project and that reason should be something that is botheri
5、ng a customer.If you cant identify what is important to work on, dont work on anything.If you cant put a name on it you cant do it.If you cant define the units of measure - be careful! You will end up having to:Define the units of measure and create a measuring system for the issue at the beginning
6、of your project.Step 2 - Measure PhaseIf you dont know what makes something good or bad, how can you measure the problem in terms of defects (DPMO)?If the definition of what makes a defect (the specification) is not in the same units of measure as the selected Y from step 1, you have the wrong spec.
7、The definition needs to be crystal clear and briefSelect the Output Characteristic (the “Y”)What is the problem? What is a defect?Define the Performance Standard of the “Y”What is the specification or level of performancerequired by the customer?5Six Sigma MethodologyStep 3 - Measure PhaseIf you can
8、t measure something, how do you know where you are, where you have been, or where you are going?If your measurement system is incapable, STOP and FIX IT before proceeding. Note: Measurement System Analysis (MSA) must be done on any and all defects you wish to count (watch out if you have a visual in
9、spection of 6 characteristics)Step 4 - Measure PhaseI know what I want to fix (step1), what defines it as good or bad (step 2), and I can accurately measure it (step 3) so:Now I can say how my process is performing in terms of long and short term Z and DPMO which is my baseline defect rate. I know c
10、an improve from here.Validate the Measurement SystemCan we accurately and consistently measure the Y? Can you trust your data?Establish Process CapabilityWhat is our defect rate today?6Six Sigma MethodologyStep 5 - Measure PhaseWhere do I want to be at the end of my project? There should be a stretc
11、h goal with logic behind its selection.The objective should make it clear to all at what point the project will be considered complete and successful.Step 6 - Analyze PhaseDefine Performance ObjectivesGiven the current defect rate, howmuch should we expect to improve?Identify Variation Sources ( “Xs
12、”)What are some potential causes of the defect?List all potential inputs (KPIV, X) that could affect my output (KPOV,Y).This is filling the top of the funnel. Some tools include fishbone, C&E matrix, Failure Modes and Effects Analysis (FMEA), detailed process maps.Think outside the box - if you fix
13、it the same way you did last year, you will have the same problem you have now.7Six Sigma MethodologyStep 7- Analyze PhaseStep 8 - Improve PhaseScreen Potential CausesCan I figure out which of the potential causes areactually affecting my defect rate using statistical tests? Screening is done using
14、Graphical Tools and Hypothesis Tests to identify and prove which are the vital Xs.This is the middle of the funnel for projects with variable relationships between Xs For some simpler projects where there are no variable relationships between my Xs (or I have only 1 X) this is also step 8 the bottom
15、 of the funnel, the final vital Xs Discover Variable RelationshipsNow that I know which ones actually cause the defect, how should they be set so they produce the least amount of defects?Evaluate if any of my vital Xs are affected by other Vital Xs. This is primarily done through the use of Design o
16、f Experiment (DOE) for moderate to complex projects. DOE is an experimental process that determines the impact of multiple inputs on a selected output. This leads to the bottom of the funnel, I know which Xs affect my Y and how they affect each other.Transactional - use OFAT approach.8Six Sigma Meth
17、odologyStep 9 - Improve PhaseStep 10 - Improve PhaseI know which Xs are important. What range of settings can I use to make good product?In the case of a variable X (e.g. PSI on an air feed), I have to provide a setting tolerance.Establish Operating TolerancesCan I establish a tolerance for my input
18、s that will still yield a low defect rate to leave myself some flexibility and facilitate control?Validate Measurement SystemNow that I have improved, can I still measure the defect or the causes going forward?In the case of a variable X (e.g. PSI on an air feed), I need to validate that it can be m
19、easured (a vital X MSA).Also, I might have improved my Y so much that I can no longer “read” my process, and may have to improve my gage system to truly measure where my improvements have brought me.9Six Sigma MethodologyStep 11 - Improve PhaseStep 12 - Control PhaseDetermine Process CapabilityWhat
20、is the defect rate now that Ihave implemented the improvements?This measures the capability of controlling my Xs where I have set them.This is also the time when we determine formal results by comparing a new capability analysis with the baseline capability analysis (step 4) and our goals (step 5).I
21、mplement Process Controls How do we make sure it stays fixed?The Xs you have determined as vital, their settings, and other actions you have taken to make the improvement must be:Nailed downSet in concreteFully implemented (NOT just agreed to)Put into a rigorous audit scheduleBEFORE you can say a pr
22、oject is closed!10Advanced Tools YIELD Accelerated ResultsFive Phases and Key ToolsControl PlanSPCMistake ProofingAutomated Control$MeasureAnalyzeImproveControlMetricsGage StudiesCapabilityProcess MapC & E MatrixFMEAMulti-VariHypothesis TestingANOVAVariation Tree DOE:Full Factorials2K FactorialsFractional FactorialsDefineOutput CharacteristicsOutput Variables11
