6. Conclusionandfutureresearch
This paperproposesahybridinventorypolicywithsplitdelivery
underregularandsurgedemand.Splitdeliveryconsidersthescenario
wherethereisasinglesupplierwhoreplenishesallordersbutthe
suppliercouldsplitthedeliveryoftheordersusingmultipleship-
ments.Thispaperidentifies aninventorymanagementpolicythat
minimizes costswhilemaintainingahighservicelevel.Thelevel
crossingtheoryisusedtoformulatethisproblemandcalculatethe
equilibriumequationsforthestationaryinventorylevelprobability
densityfunction(PDF)underagiveninventorypolicyfeaturedbythe
orderingpointsandorderquantitiesofbothordertypes.ATabu
search-basedheuristicsisdevelopedinordertoidentifysolutionsfor
this problem.NumericalexperimentsdemonstratethattheTabu
search finds highqualitysolutionsinashortertimethantheexplicit
enumeration.Numericalresultsgeneratedfromdifferentinstances
alsoshowthatproposedhybridinventorypolicywithsplitdelivery
couldoutperformhybridinventorypolicieswithstandarddeliveryby
saving3.5–42.1%oftotalcost.
The paperonlyconsidersasingle-itemsingle-locationinventory
system.Onefutureresearchdirectioncouldbetoextendthemodelto
a multiple-locationcase.Thepresentedmodelassumesthatitemsare
notperishable,thoughmanymedicineshave finite shelflifeinprac-
tice.Thishybridinventorymodelcouldbeextendedtoperishable
productswitha fixedorvariableshelflife.Theoptimizationapproach
reliesonTabusearchbasedheuristics.Therefore,anotherpotential
extensionisdevelopingexactsolutionalgorithmswhich find optimal
inventorypolicies.Insummary,thispapermakesacontributiontothe
inventorymanagementliteratureregardingtheresponsetotwotypes
of demandpatterns.Inpractice,themodelandsolutionapproach
presentedinthispapercanhelphealthcareprovidersorhumanitarian
logistics providersinmanagingtheir inventories.Theproposedhybrid