供应链库存管理和共享信息的价值外文翻译Word下载.docx

供应链库存管理和共享信息的价值外文翻译Word下载.docx

《供应链库存管理和共享信息的价值外文翻译Word下载.docx》由会员分享，可在线阅读，更多相关《供应链库存管理和共享信息的价值外文翻译Word下载.docx（8页珍藏版）》请在冰豆网上搜索。

SupplyChainInventoryManagementandtheValueofSharedInformation

MaterialSource:

http:

//mansci.journal.informs.org/cgi/content/abstract/46/8/1032

Author:

Ge´

rardP.Cachon•MarshallFisher

Intraditionalsupplychaininventorymanagement,ordersaretheonlyinformationfirmsexchange,butinformationtechnologynowallowsfirmstosharedemandandinventorydatequicklyandinexpensively.Westudythevalueofsharingthesedatainamodelwithonesupplier,Nidenticalretailers,andstationarystochasticconsumerdemand.Thereareinventoryholdingcostsandback-orderpenaltycosts.Wecompareatraditionalinformationpolicythatdoesnotusesharedinformationwithafullinformationpolicythatdoesexploitsharedinformation.Inanumericalstudywefindthatsupplychaincostsare2.2%loweronaveragewiththefullinformationpolicythanwiththetraditionalinformationpolicy,andthemaximumdifferenceis12.1%.Wealsodevelopasimulation-basedlowerboundoverallfeasiblepolicies.Thecostdifferencebetweenthetraditionalinformationpolicyandthelowerboundisanupperboundonthevalueofinformationsharing:

Inthesamestudy,thatdifferenceis3.4%onaverage,andnomorethan13.8%.Wecontrastthevalueofinformationsharingwithtwootherbenefitsofinformationtechnology,fasterandcheaperorderprocessing,whichleadtoshorterleadtimesandsmallerbatchsizes,respectively.Inoursample,cuttingleadtimesnearlyinhalfreducescostsby21%onaverage,andcuttingbatchesinhalfreducescostsby22%onaverage.Forthesettingswestudy,weconcludethatimplementinginformationtechnologytoaccelerateandsmooththephysicalflowofgoodsthroughasupplychainissignificantlymorevaluablethanusinginformationtechnologytoexpandtheflowofinformation.

（SupplyChain;

Multi-EchelonInventoryManagement;

PeriodicReviewPolicies;

ElectronicDataInterchange）

1.Introduction

Informationtechnologyhashadasubstantialimpactonsupplychains.Scannerscollectsalesdataatthepoint-of-sale,andelectronicdatainterchange（EDI）allowsthesedatatobesharedimmediatelywithallstagesofthesupplychain.Theapplicationofthesetechnologies,especiallyinthegroceryindustry,hassubstantiallyloweredthetimeandcosttoprocessanorder,leadingtoimpressiveimprovementsinsupplychainperformance（seeCachonandFisher1997,ClarkandHammond1997,KurtSalmonAssociates1993）.

Asaresultofthesesuccessstories,thereisnowageneralbeliefwithinindustrythatcapturingandsharingreal-timedemandinformationisthekeytoimprovedsupplychainperformance.Thepurposeofthisresearchistotestthisbeliefbyrigorouslymeasuringthevalueofinformationsharingandcomparingthisvaluetotwoothersourcesofsupplychainimprovement:

reducingleadtimesandincreasingdeliveryfrequencybyreducingshipmentbatchsizes.Notethatthesameinformationtechnologythatfacilitatesinformationsharingalsocontributestothereductionofleadtimesandshipmentfrequencybyreducingthetimeandcosttoprocessorders.Thus,thequestionaddressedinhereisnotwhetherinformationtechnologyimprovessupplychainperformance,buthow.Specifically,doestheprimarygaincomefromsharinginformationorfromallowingproductstoflowmorequicklyandevenlyinthesupplychain?

WeaddressthisquestionwithinthecontextofasupplychainwithonesupplierandNidenticalretailersthatfacestationarystochasticconsumerdemandwithaknowndistribution.Therearefixedtransportationtimesbetweenlocations,andshipmentquantitiesequalamultipleofabasebatchquantity.Thereareholdingcostsatalllevelsandback-orderpenaltycostsatthelowestlevel.Thismodelprovidesareasonablerepresentationofsupplychainssellinganestablishedproductunderconstantpricingconditions.

Weconsidertwolevelsofinformationsharing.Withtraditionalinformationsharingthesupplieronlyobservestheretailers’orders.Withfullinformationsharingthesupplierhasimmediateaccesstotheretailers’inventorydata.Wedevelopaninventorypolicyforeachinformationsharinglevel.Reorderpointpoliciesareusedwithtraditionalinformationsharing.Theretailersalsousereorderpointpolicieswithfullinformation,butthesupplierdoesnot.Instead,thesupplierusesitsadditionalinformationtobetterallocateinventoryamongtheretailersandtoimproveitsorderdecisions（i.e.,tobettertimeitsownreplenishments）.

Thedifferencebetweensupplychaincostsundertraditionalandfullinformationisonemeasureofthevalueofsharedinformation.However,theremayexistevenbetterpoliciesforeitherinformationlevel,thatis,optimalpoliciesareunknownforeachlevel.Toaccountforthispossiblebias,wedevelopasimulation-basedlowerboundoverallfeasiblepolicies,nomatterwhatthelevelofinformationsharingis.Thecostdifferencebetweentraditionalinformationandthelowerboundisthemaximumvalueofsharedinformation.

Inanumericalstudywithawiderangeofparametervalueswefindthatinformationsharingreducessupplychaincostsby2.2%onaverage,andthegapbetweentraditionalinformationpolicycostandthelowerboundis3.4%onaverage.Cuttingleadtimebynearlyhalf（fromfivetothreeperiods）reducescostsby21%onaverage,andcuttingbatchsizeinhalfreducessupplychaincostsby22%.Werecognizethatthiscomparisonismeaningfulonlyifthoseleadtimeandbatchsizereductionscanbereasonablyexpectedfromtheimplementationofinformationtechnology.Infact,wedidobservecomparablereductionsatCampbellSoupCompanywhenitimplementedinformationtechnologytoimproveitssupplychain.1Therehasalsobeenotherdocumentationontheimpactofinformationtechnologyinthegroceryindustry:

Barilla,theworld’slargestpastaproducer,reduceditsleadtimefromoveroneweektotwodays（HarvardBusinessSchoolcase9-694-046）;

andH.E.B.,alargegrocerychainbasedinTexas,eliminated6to10daysfromitsleadtime（HarvardBusinessSchoolcase9-195-125）.Weconcludethatwhileinformationsharingdoesreducecosts,simplyflowinggoodsthroughthesupplychainmorequicklyandmoreevenlyproducesanorderofmagnitudegreaterimprovement.

2.LiteratureReview

Thefollowingpapersshowhowsharingdemandandinventorydatacanimprovethesupplier’sorderquantitydecisionsinmodelswithknownandstationaryretailerdemand:

Bourlandetal.（1996）,Chen（1998）,Gavirnenietal.（1999）,andAvivandFedergruen（1998）.Leeetal.（2000）usesharedinformationtoimprovethesupplier’sorderquantitydecisionsinaserialsystemwithaknownautoregressivedemandprocess.Liljenberg（1996）studieshowtousesharedinformationtoimprovethesupplier’sallocationofinventoryamongtheretailers.Inourmodelsharedinformationisexploitedforbothuses:

bettersupplierreplenishmentsandbetterallocationstotheretailers.

Wefocusonsharingdemandandinventorydata,butthereareotherdatathatcanbesharedinasupplychain.Gavirnenietal.（1999）measurethebenefitofsharingtheparametersoftheretailer’sorderingpolicywiththesupplier.Aviv（1998）exploresthebenefitsofsharingforecastsforfuturedemand.

Inourmodel,asintheotherstudiesmentioned,itisassumedthatinformationisalwayssharedtruthfully.CachonandLariviere（1997）studyforecastsharingwhentheforecastproviderhasanincentivetoprovideanoverlyoptimisticforecastofdemand.

BothLeeetal.（2000）andGavirnenietal.（1999）assumethereexistsaperfectlyreliableexogenoussourceofinventory;

informationsharinghasnoimpactontheretailerbecauseitsordersarealwaysreceivedinfullafterafixednumberofperiods.Intheotherpapers,asinourmodel,thesupplieristheonlysourceofinventory.Therefore,informationsharingmayimpacttheretailersbychangingthesupplier’sorderquantitiesorallocations.

Gavirnenietal.（1999）andAvivandFedergruen（1998）allowforlimitedsuppliercapacity,whereascapacityisunrestrictedinourmodelandintheotherpapers.

Thereportedbenefitsofinformationsharingvaryconsiderably.Liljenberg（1996）findsthatbetterallocationlowerssupplychaincostsby0%to3.9%.Chen（1998）findsthatsupplychaincostsareloweredupto9%,andonaverageby1.8%.AvivandFedergruen（1998）reportbenefitsof0%–5%.Incontrast,Leeetal.（2000）findthatinformationsharingloweredsupplychaincostsbyabout23%intheirscenariowiththehighestdemandnonstationarity.However,Graves（1999）studiesasimilarmodel,withtheexceptionthatthereisnooutsideinventorysource,andconcludesthatinformationsharingprovidesnobenefittothesupplychain.Gavirnenietal.（1999）reportthatsharingtheretailer’sdemanddatareducedthesupplier’scostby1%–35%.Theimpactonthesupplychain’scostwouldbelowerbecauseinformationsharingintheirmodelhasnoimpactontheretailer’scosts.

Thereisotherresearchrelatedtoourwork.Leeetal.（1997）findthatsharinginformationreducesthesupplier’sdemandvariance,whichshouldbenefitthesupplychain,buttheydonotquantitativelymeasurethisbenefit.Therearemanystudiesthatinvestigateasupplychainmodelwithonesupplier,Nretailers,stochasticconsumerdemand,andbatchordering.Someofthemassumetraditionalinformation（e.g.,Axester1993,Cachon1995,ChenandSamroengraja1996,LeeandMoinzadeh1986,SvoronosandZipkin1988）,whileothersassumefullinformation（e.g.,ChenandZheng1997,Graves1996,McGavinetal.1993）.Becauseofdifferentassumptionsandtestproblems,itisnotpossibletomeaningfullycomparesupplychaincostsacrossthosetwosetsofstudies.Severalres