美國的一個倉庫管理系統(tǒng)

Internet-basedDataEnvelopmentAnalysisforWarehousingOutlineTheproblemThecurrentsolutionTheneedAnewsolutionHowitworksInternetdeploymentandresultstodateFuturedirectionsPerformanceAssessmentHowwellareyouperforming?Doyouhaveopportunitiestoimprove?WarehouseOperationsReceivingFunction(inbound)UnloadInspectPutAwayStorageFunctionShippingFunction(orderfulfillment)LoadPackOrderPickStorageFunctionStorageFunctionSingleFactorProductivityMetricsProductivity=outputinputTraditionalPerformanceMetricsFillrateInventoryturnsLines/hourOrders/hour$/line$/orderWorkokwhenRequirementsarenotchangingTechnologyisnotchangingCompetitionisnotchangingIt’sveryhardtointerpretasinglefactorproductivitymetricwhentheenvironmentissubjecttorapidchangeinproducts,customerrequirements,technology,orcompetition.Butinadynamicworld…Can’tcompareovertimeCan’tcompareacrosslocationsCan’tcomparetoothercompaniesAtleastnotwithoutalotofadditionalexplanatorydataandinformation!BenchmarkingRelativeperformancelevel{Best(effective)practices}INPUTSOUTPUTSSystem-orientedPerformanceMeasureTheNeedResourcesServicesActivitiesTotalFactorProductivity?Can’tsolvethepricingproblemINPUTSOUTPUTSONEPERFORMANCEINDEXDataEnvelopmentAnalysisResourcesServicesActivitiesSystem-basedassessmentmethodResources:capital,labor,overheadActivities:inbound,orderfulfillmentServices:lines/qtyshipped,fillrate,etcComparetootherwarehousesAllotherwarehousesAllotherwarehousesinyourindustryAllotherwarehousesinyourcompanyYourwarehouseinthepastProductionFunctionTheoryforoneinput,oneoutputResource/InputProduction/OutputSystemEfficiencyConceptResource/InputProduction/OutputOBASystemefficiencyofwarehouseBistheratioOAOBDEAModel:Charnes,Cooper,andRhodesConstantReturnstoScaleDataEnvelopmentAnalysisAllowsustoconsidermultiple““inputs”Allowsustoconsidermultiple““outputs””Determinesthereferencepointontheproductionfunctionbyconstructingahypothetical““bestpractices””warehouseusingrealwarehousedataBestpossible*notaverage**fromdataDEAPerformanceScoreContributiontoProfitInput/OutputSpecification(theFrazelle/Hackmanmodel)EfficiencyWarehouseLinesShippedStorageFunctionAccumulationTotalStaffingEquipment““Replacement”CostWarehouseareaHtmldocumentsSolverDatabaseAtyoursiteGTServerWeb-basedToolOvertheinternetOver150qualifiedusersResultstoDateExperienceExistingdatabaseMorethan150warehousesNotsegmentedbyindustry(yet)No““descriptive””datatouseforsegmentingCansegmentbasedoninputsandoutputsOutputSegmentationbrokencase:49fullcase:32pallet:13mix:65total:159Input-oriented,all159togetherBrokenCase,InputEfficiencyComparedtoAll(49/159)BrokenCase,InputEfficiencyComparedWithin(49/49)PickRateforBrokenCasePickingAve=17SD=27Fullcase,InputEfficiencyComparedtoAll(32/159)Fullcase,InputEfficiencyComparedWithin(32/32)PickRateforCasePickingAve=14

SD=27.7Pallet,InputEfficiencyComparedtoall(13/159)Pallet,InputEfficiencyComparedwithin(13/13)PickRateforPalletPickingLines/Laborhour(pallet)02468100.010.020.030.040.050.0100.0150.0200.0Morelines/laborhourFrequencyAve=25

SD=27.7Mixed,InputEfficiencyComparedtoall(65/159)Mixed,InputEfficiencyComparedWithin(65/65)PickRateforMixedPickingLines/LaborHour(mix)01020304050600.0020.0040.0060.0080.00100.00lines/laborhourFrequencyAve=10.6SD=23AggregatePickRateforAll159Wheredowegofromhere?ManyOpportunitiestoImprovetheBenchmarkingToolEnhancethebasicinput/outputmodelEnhancetheabilitytobenchmarkfortechnology,practice,&requirementsSomeSuggestedMetricsInputsSpaceCapitalLaborInventory#ofskusturnsOutputsInboundreceiptmixreceiptvariabilityreturnstimetoavailabilityFulfillmentpickvolumepickvariabilitypickaccuracyfillratebut…Sorta……“Marker”AnalysisPerformance“Marker”AttributeDEAPerformanceScorePerformance“Marker”PracticeDEAPerformanceScoreResultsBiggerisnotalwaysbetter,atleastwithregardtoequipmentandlabor.Thereis,however,someevidencethatmorewarehousespaceleadstobettersystemefficiency.Laborhourswasnotfoundtobeasignificantfactor,byitself,inpredictingsystemefficiency.However,theinteractionoflaborwithinvestmentwasfoundtobesignificantinthesensethatlaborhoursmitigatestheeffectofinvestment(inotherwords,thoughhighinvestmentwarehousestendedtobelessefficientthanlowinvestmentwarehouses,thedifferencesbecomeslessprominentthehigherthelaborhours).MoreResultsTheinteractionofinvestmentandareawasfoundtobesignificant.Thismeansthathighinvestmentwarehousesareevenlessefficientiftheyarealsolarge.Nomatterhowwesegmentthedata,averylargeproportionofwarehousesareoperatingatorbelow50%systemefficiency.Whilethismayreflectseasonalfluctuationsincustomerorders,itstillrepresentsaverysignificantopportunityforimprovement.MoreResultsTheopportunityforimprovementseemslargestforthesegmentofwarehousesdoingpredominantlyfullcasepicking.Inthatsegment,asmallerproportionofthewarehousesare"efficient'thaninanyothersegment,andalargerproportionareoperatingbelow50%efficiency.OtherDataRequirementsType(wholesale,retail,manufacturing)Industry(pharma,auto,electronic