Multi-echelon multi-company inventory planning with limited information exchange

Supply chain planning concepts from multi-echelon inventory theory are generally based on some form of centralised planning of supply chains. Those multi-echelon models that do consider decentralised planning, assume complete information and/or a specific single objective function. This paper investigates how multi-echelon inventory theory can accommodate a setting with decentralised decision makers (a supplier and a number of retail groups) without complete information. We present a coordination procedure that does not require the retail groups to exchange demand information, but does allow using opportunities for demand pooling between them. We illustrate our ideas by way of a quantitative analysis of a two-echelon divergent supply chain, with both cooperative and non cooperative retail groups. We conclude that coordination across a supply chain with decentralised control and limited centralised information is feasible by using available algorithms with satisfactory service level and cost performance.     

An Example of Disbenefits of Centralized Stocking

This paper deals with a single-period, single-product inventory model. It is a multilocation problem with an opportunity for centralization. In the centralized system, the location's demands are satisfied from one central warehouse whose orders are processed on a first come, first served basis. The inventory size will be constrained to meet a specific maximum probability of being out of stock at each location. Under these assumptions, Stulman argues that a centralized system requires a lower total starting inventory than the equivalent decentralized one. We shall present a counter example showing that, under conditions where an 86% probability of stockouts is the maximum acceptable, a higher total starting inventory may be required by centralizing two locations' stock-holdings.     

Optimal Solutions in the Multi-location Inventory System with Transshipments

We consider a single-period multi-location inventory system where inventory choices at each location are centrally coordinated. Transshipments are allowed as recourse actions in order to reduce the cost of shortage or surplus inventory after demands are realized. This problem has not been solved to optimality before for more than two locations with general cost parameters. In this paper we present a simple and intuitive model that enables us to characterize optimal inventory and transshipment policies for three and four locations as well. The insight gained from these analytical results leads us to examine the optimality conditions of a greedy transshipment policy. We show that this policy will be optimal for two and three locations. For the n location model we characterize the necessary and sufficient conditions on the cost structure for which the greedy transshipment policy will be optimal.      

Optimal inventory management for a retail chain with diverse store demands

Item demands at individual retail stores in a chain often differ significantly, due to local economic conditions, cultural and demographic differences and variations in store format. Accounting for these variations appropriately in inventory management can significantly improve retailers’ profits. For example, it is shown that having greater differences across the mean store demands leads to a higher expected profit, for a given inventory and total mean demand. If more than one inventory shipment per season is possible, the analysis becomes dynamic by including updated demand forecasts for each store and re-optimizing store inventory policies in midseason. In this paper, we formulate a dynamic stochastic optimization model that determines the total order size and the optimal inventory allocation across nonidentical stores in each period. A generalized Bayesian inference model is used for demands that are partially correlated across the stores and time periods. We also derive a normal approximation for the excess inventory from the previous period, which allows the dynamic programming formulation to be easily solved. We analyze the tradeoffs between obtaining information and profitability, e.g., stocking more stores in period 1 provides more demand information for period 2, but does not necessarily lead to higher total profit. Numerical analyses compare the expected profits of alternative supply chain strategies, as well as the sensitivity to different distributions of demand across the stores. This leads to novel strategic insights that arise from adopting inventory policies that vary by store type.     

Effects of Centralization on Expected Costs in a Multi-location Newsboy Problem

This is a single-period, single-product inventory model with several individual sources of demand. It is a multi-location problem with an opportunity for centralization. The holding and penalty cost functions at each location are assumed to be identical. Two types of inventory system are considered in this paper: the decentralized system and the centralized system. The decentralized system is a system in which a separate inventory is kept to satisfy the demand at each source of demand. The centralized system is a system in which all demands are satisfied from one central warehouse. This paper demonstrates that, for any probability distribution of a location's demands, the following properties are always true: given that the holding and penalty cost functions are identical at all locations, (1) if the holding and penalty cost functions are concave functions, then the expected holding and penalty costs in a decentralized system exceed those in a centralized system, except that (2) if the holding and penalty cost functions are linear functions, and for any i≠j, P_ij, the coefficient of correlation between the ith location's demand and the jth location's demand is equal to 1, then the expected holding and penalty costs in a decentralized system are equal to those in a centralized system.     

Uniform distribution of inventory positions in two-echelon periodic review systems with batch-ordering policies and interdependent demands

We consider a two-level inventory system in which there are one supplier and multiple retailers. The retailers face stochastic, interdependent customer demands. Each location employs a periodic-review (R,nQ), or lot-size reorder point, inventory policy. We show that each location's inventory positions are stationary and the stationary distribution is uniform and independent of any other's.     

A Bayesian Approach to Estimating Decision Parameters in a Replacement Inventory System

This paper deals with the estimation of decision parameters in a special class of inventory systems in which the rate of stock depletion depends on the internal operations of the parent organization instead of market demands. Stocking decisions for this type of inventory systems are unique in that demands for the stocked items are internal demands, sporadic in some instances, infrequent in others, and almost never the direct results of conscious economic decisions. To cope with the peculiar situation, Bayesian point and interval estimation procedures are used. The estimated probability measures provide a basis on which stocking decisions pertinent to this type of inventory systems can be made or evaluated. Specific statistical postulates are made, and a numerical example is given for illustration. The possibility of applying the estimated results to one of the existing inventory models is briefly discussed at the end.     

An approximation algorithm for a facility location problem with stochastic demands and inventories

We propose a 2-approximation algorithm for a facility location problem with stochastic demands. At open facilities, inventory is kept such that arriving requests find a zero inventory with (at most) some pre-specified probability. Costs incurred are expected transportation costs, facility operating costs and inventory costs.     

Capacitated location model with online demand pooling in a multi-channel supply chain

This paper presents a location model that assigns online demands to the capacitated regional warehouses currently serving in-store demands in a multi-channel supply chain. The model explicitly considers the trade-off between the risk pooling effect and the transportation cost in a two-echelon inventory/logistics system. Keeping the delivery network of the in-store demands unchanged, the model aims to minimize the transportation cost, inventory cost, and fixed handling cost in the system when assigning the online demands. We formulate the assignment problem as a non-linear integer programming model. Lagrangian relaxation based procedures are proposed to solve the model, both the general case and an important special case. Numerical experiments show the efficiency of our algorithms. Furthermore, we find that because of the pooling effect the variance of in-store demands currently served by a warehouse is an important parameter of the warehouse when it is considered as a candidate for supplying online demands. Highly uncertain in-store demands, as well as low transportation cost per unit, can make a warehouse appealing. We illustrate with numerical examples the trade-off between the pooling effect and the transportation cost in the assignment problem. We also evaluate the cost savings between the policy derived from the model, which integrates the transportation cost with the pooling effect, and the commonly used policy, which is based only on the transportation cost. Results show that the derived policy can reduce 1.5–7.5% cost in average and in many instances the percentage of cost savings is more than 10%.     

A real-time decision rule for an inventory system with committed service time and emergency orders

In this paper, we study the inventory system of an online retailer with compound Poisson demand. The retailer normally replenishes its inventory according to a continuous review (nQ, R) policy with a constant lead time. Usually demands that cannot be satisfied immediately are backordered. We also assume that the customers will accept a reasonable waiting time after they have placed their orders because of the purchasing convenience of the online system. This means that a sufficiently short waiting time incurs no shortage costs. We call this allowed waiting time “committed service time”. After this committed service time, if the retailer is still in shortage, the customer demand must either be satisfied with an emergency supply that takes no time (which is financially equivalent to a lost sale) or continue to be backordered with a time-dependent backorder cost. The committed service time gives an online retailer a buffer period to handle excess demands. Based on real-time information concerning the outstanding orders of an online retailer and the waiting times of its customers, we provide a decision rule for emergency orders that minimizes the expected costs under the assumption that no further emergency orders will occur. This decision rule is then used repeatedly as a heuristic. Numerical examples are presented to illustrate the model, together with a discussion of the conditions under which the real-time decision rule provides considerable cost savings compared to traditional systems.     

An inventory sharing and allocation method for a multi-location service parts logistics network with time-based service levels

We consider a model to allocate stock levels at warehouses in a service parts logistics network. The network is a two-echelon distribution system with one central warehouse with infinite capacity and a number of local warehouses, each facing Poisson demands from geographically dispersed customers. Each local warehouse uses a potentially different base stock policy. The warehouses are collectively required to satisfy time-based service targets: Certain percentages of overall demand need to be satisfied from facilities within specified time windows. These service levels not only depend on the distance between customers and the warehouses, but also depend on the part availabilities at the warehouses. Moreover, the warehouses share their inventory as a way to increase achieved service levels, i.e., when a local warehouse is out of stock, demand is satisfied with an emergency shipment from another close-by warehouse. Observing that the problem of finding minimum-cost stock levels is an integer non-linear program, we develop an implicit enumeration-based method which adapts an existing inventory sharing model from the literature, prioritizes the warehouses for emergency shipments, and makes use of a lower bound. The results show that the proposed inventory sharing strategy results in considerable cost reduction when compared to the no-sharing case and the method is quite efficient for the considered test problems.     

Single-period Two-product Inventory Model with Reorder and Substitution

In market, excess demands for many products can be met by reorder even during one period, and retailers usually adopt substitution strategy for more benefit. Under the retailer's substitution strategy and permission of reorder, we develop the profits maximization model for the two-substitutable-product inventory problem with stochastic demands and proportional costs and revenues. We show that the objective function is concave and submodular, and therefore the optimal policy exists. We present the optimal conditions for order quantity and provide some properties of the optimal order quantities. Comparing our model with Netessine and Rudi's, we prove that reorder and adoption of the substitution strategy can raise the general profits and adjust down the general stock level.     

Priority centralization of the one period multi-location inventory system

In this paper, we will examine a multi-center one-period inventory system. The usual penalty cost for being out of stock will be replaced by an assurance of service constraint at each location. That is, we will constrain our inventory size to meet a specific maximum probability of being out of stock at each location. The centralized system we shall propose will define a priority rule which will cause us to satisfy the entire demand of high priority locations before we begin satisfying the demands of lower priority locations. This will allow us to find a minimum initial inventory level for the centralized system that will meet all of the assurance of service constraints. We will look at the special case where the variance of the total demand of several locations is non-decreasing in locations included in the total. In this case, we will show the computations required for finding the optimal centralized priority system are minimal. Finally, we will show that such a system is superior to a decentralized system.     

Analysis of s,S inventory systems with general lead time and demand distribution and adjustable reorder size

Stochastic Inventory systems of (s, S) type with general lead time distribution are studied when the time intervals between successive demands are independently and identically distributed. The demands are assumed to occur for one unit at a time and the quantity reordered is subject to review at the epoch of replenishment so as to level up the inventory to S. An explicit characterization of the inventory level is provided. The model is flexible enough to allow complete backlogging and or deal with shortages. A general method of dealing with cost over an arbitrary time interval is indicated. Special cases are discussed when either the lead time or the interval between successive demands is exponentially distributed.     

The determination of optimal partial fill policy for an inventory system with lumpy demand items

This paper presents a mathematical model developed for the synthesis of optimal replenishment policies for items that experience lumpy demands. In order to avoid disrupting the inventory system, a cutoff point of w units is introduced such that the system would only satisfy routinely customer orders with transaction sizes less than or equal to w units. For customer orders with transaction sizes larger than w units, the system would only supply the cutoff amount (w units). The excess units would be refused. The control discipline is the (s, S) inventory policy with continuous review, and the nature of the customer orders is approximated by a discrete stuttering Poisson distribution. The optimal values of the control parameters, w, s and S, are determined. The theoretical results obtained are illustrated with a numerical example.     

(s, S) Inventory systems with lost sales and Markov renewal demands

In the study of stochastic inventory systems it is generally assumed that the demand epochs are renewable in nature. The present paper deals with a single-item (s, S) inventory model with a finite number of different types of demands, in which the demand epochs form a Markov renewal process. The lead times are exponentially distributed and the demands that occur during stockout periods are not backordered. For this model the transient and limiting inventory level distributions are computed. Also the theory of point processes is used to obtain the mean reorder and shortage rates and their limiting values. For the special case of renewal demands, the problem of minimizing the long-run expected cost rate is analysed.     

Benefits of Centralized Stocking for the Multi-Centre Newsboy Problem with First Come,First Served Allocation

In this paper, a multi-centre, one-period inventory problem is examined. The usual penalty cost of being out of stock is replaced by an assurance of service constraint at each location. That is, the inventory size will be constrained to meet a specific maximum probability of being out of stock at each location. The objective will be to show how to find the necessary starting inventory of a centralized system where orders are processed on a first come, first served basis. This can then be used to find the savings associated with the centralized inventory over a non-centralized one.     

Capacitated replenishment and disposal planning for multiple products with resalable returns

We consider a replenishment and disposal planning problem (RDPP) that arises in settings where customer returns are in as-good-as-new condition. These returns can be placed into inventory to satisfy future demand or can be disposed of, in case they lead to excess inventory. Our focus is on a multi-product setting with dynamic demands and returns over a finite planning horizon with explicit replenishment and disposal capacities. The problem is to determine the timing of replenishment and disposal setups, along with the associated quantities for the products, so as to minimize the total costs of replenishment, disposal, and inventory holding throughout the planning horizon. We examine two variants of the RDPP of interest both of which are specifically motivated by a spare part kitting application. In one variant, the replenishment capacity is shared among multiple products while the disposal capacity is product specific. In the other variant, both the replenishment and disposal capacities are shared among the products. We propose a Lagrangian Relaxation approach that relies on the relaxation of the capacity constraints and develop a smoothing heuristic that uses the solution of the Lagrangian problem to obtain near-optimal solutions. Our computational results demonstrate that the proposed approach is very effective in obtaining high-quality solutions with a reasonable computational effort.     

An Inventory System with Postponed Demands

Abstract

In this article we consider a continuous review perishable inventory system in which the demands arrive according to a Markovian arrival process (MAP). The items in the inventory have shelf life times that are assumed to follow an exponential distribution. The inventory is replenished according to an (s, S) policy and the replenishing times are assumed to follow a phase type distribution. The demands that occur during stock out periods either enter a pool which has capacity N (<∞) or leave the system. Any demand that arrives when the pool is full and the inventory level is zero, is also assumed to be lost. The demands in the pool are selected one by one, if the replenished stock is above s, with interval time between any two successive selections is distributed as exponential with parameter depending on the number of customers in the pool. The joint probability distribution of the number of customers in the pool and the inventory level is obtained in the steady state case. The measures of system performance in the steady state are derived and the total expected cost rate is also calculated. The results are illustrated numerically.     

电子商务环境下考虑退货的选址库存问题模型与算法

选址库存问题(location inventory problem, LIP)是物流系统集成的经典问题之一,也是企业需要面对的管理决策难题。本文考虑在电子商务环境下无质量缺陷的退货商品可简单再包装后重新进入销售市场这一现实情况,对设施选址和库存控制进行集成优化,构建随机需求下有退货的LIP模型。针对此问题求解的复杂性,设计了改进的自适应混合差分进化算法对模型进行整体求解。最后,通过多组算例验证了模型和算法的实用性和优越性,可为设施选址、库存控制和商品配送回收决策提供重要参考依据。