Inventory Management In Closed-loop Supply Chains With Dependent Return And Demand

The most important feature of closed-loop supply chain is that there exist product returns in it.Product returns have become more and more common in many industries.Whether it be the independent remanufacturers actively collect products sold by original equipment manufacturers and used them to remanufacture,or comsumers return products they don’t want any longer,or product reuse in the operations of firms,nonnegligible problems in the management exist.From a practical perspective,the increasing number of returned products add huge complexity and bring tough challenge to the inventory management in closed-loop supply chains.It makes how to transit from comventional inventory management mode to one with product returns an important issue.From an academic perspective,although many papers have studied inventory models with product returns,most of them assume the returns are independent of demand and derive some results according to that assumption.However in reality,the returns are dependent on demand and the number of returns wouldn’t exceed that of sold products in previous time.Aiming at the inadequacy of existing literature,this dissertation assumes the returns are dependent on demand and studies some problems in inventory management with product returns.This dissertation contains three contents.In the first part when there exists remanufacturing competition,we study an original equipment manufacturer how to make use of strategic inventory to deter an independent remanufacturer.In the second part when there exists commercial returns,we study how many checkers should be allocated to minimize the expected total costs in which the holding cost of returned products is included.In the third part when there exists reuse,we study the optimal ordering policy.Although the emphases of the three parts are respectively different,they are under the same framework that returns are dependent on demand.In the first part,because in many industries independent remanufacturers abound and their remanufactured products impose an external threat on the newproducts of original equipment manufacturers.To cope with the threat,conventional wisdom suggests original equipment manufacturers involve in the remanufacturing process.However,in reality there are still many original equipment manufacturers who are unwilling to adopt the remanufacturing strategy worrying their own remanufactured productswould cause an internal competition and cannibalize with their new products.In this paper,we propose a new strategy for original equipment manufacturers to deal with the dilemma.Specifically,we use a two-period model to show by building up strategic inventory in the first period,original equipment manufacturers can successfully deter the entrance of independent remanufacturers.The strategic inventory strategy works because it can be used as a commitment device to convey a credible information that there will be enough new products in the second period and entering the market will be rendered unprofitable for independent remanufacturers.In the second part,we assume the sojourn time of the products in the market and check time of the returned products both are exponentially distributed,the checkers will automatically adjust the checking speed according to the number of return products in the system.We establish a costs control model making use of finite source queuing theory and discuss how many checkers should be allocated to minimize the expected total costs when product returns are dependent on demands and the returned products must be checked before reselling.Through the analyses of numerical examples,we find more checkers are needed when it is more likely that the returned products could be resold after detection.This is because if there are more returned products could be resold after detection,it would be better to let them pass the detection more quickly.That could decrease not only the returnable product inventory cost but also the production cost.The benefit can offset the increased costs of detection and serviceable product inventory due to the increased number of checkers.On the other hand,more checkers are needed when the fewer sojourn time of the products in the market.This is because if products spend less time in the market,there would be more returns.Let the returns pass the dectection more quickly could decrease the cost of production.The benefit alone can offset the increased costs of detection,returnable and serviceable product inventories due to the increased number of checkers.In the third part,we assume there is only single product in the market,the returned products in every period are collected form products sold in previous periods.By establishing a periodic review inventory control system with product return,we study how to find the optimal ordering policy to minimize the expected total costs in every period.Our results demonstrate there exists an optimal ordering up to level which is related to the two system state variables in a single period.If inventory level in the beginning of the period is less than the optimal ordering up to level,the best strategy is to order and maintain the inventory to that level.Otherwise do not order.If the planning horizon increases to two periods,there exists an optimal ordering up to level which is related to the two system state variables in each period.However,the two optimal ordering up to levels do not have a unified form.This is because product return ruins the monotonicity existed in inventory replenishment periods in conventional inventory control models.Furthermore,the dependency between demand and returns brings a huge difficulty to the analysis of the N-period problem.Accordingly,in this part we further assume the stochastic demand in every period is large enough such that the returns are not enough to satisfy the demand and additional new products have to be ordered.Under this assumption,we propose an ordering policy which have a unified form in N periods to aid managers.