Models forrobusttacticalplanninginmulti-stageproductionsystemswith uncertain demands

We considertheproblemofdesigningrobusttacticalproductionplans,inamulti-stageproduction system, whentheperiodicdemandsofthefinishedproductsareuncertain.First,wediscusstheconcept of robustnessintacticalproductionplanningandhowweintendtoapproachit.Wethenpresentand discuss threemodelstogeneraterobusttacticalplanswhenthefinished-productdemandsarestochastic with knowndistributions.Inparticular,wediscussplansproduced,respectively,byatwo-stagestochastic planning model,byarobuststochasticoptimizationplanningmodel,andbyanequivalentdeterministic planning modelwhichintegratesthevariabilityofthefinished-productdemands.Thethirdmodeluses finished-product averagedemandsasminimalrequirementstosatisfy,andseekstooffsettheeffectof demand variabilitythroughtheuseofplannedcapacitycushionlevelsateachstageoftheproduction system. Anexperimentalstudyiscarriedouttocomparetheperformancesoftheplansproducedbythe three modelstodeterminehoweachoneachievesrobustness.Themainresultisthattheproposedrobust deterministic modelproducesplansthatachievebettertrade-offsbetweenminimumaveragecostand minimum costvariability.Moreover,therequiredcomputationaltimeandspacearebyfarlessimportant in theproposedrobustdeterministicmodelcomparedtothetwoothers.