Mohsen Elhafsi

Production and Inventory Control of a Single Product Assemble-to-Order System with Multiple Customer Classes

We consider the optimal production and inventory control of an assemble-to-order system with m components, one end-product, and n customer classes. A control policy specifies when to produce each component and, whenever an order is placed, whether or not to satisfy it from on-hand inventory. We formulate the problem as a Markov decision process and characterize the structure of an optimal policy. We show that a base-stock production policy is optimal, but the base-stock level for each component is dynamic and depends on the inventory level of all other components (more specifically, it is nondecreasing). We show that the optimal inventory allocation for each component is a rationing policy with different rationing levels for different demand classes. The rationing levels for each component are dynamic and also nondecreasing in the inventory level of all other components. We compare the performance of the optimal policy to heuristic policies, including the commonly used base-stock policy with fixed base-stock levels, and find them to perform surprisingly well, especially for systems with lost sales.

Optimal Control of a Production-Inventory System with Both Backorders and Lost Sales

We consider the optimal control of a production inventory-system with a single product and two customer classes where items are produced one unit at a time. Upon arrival, customer orders can be fulfilled from existing inventory, if there is any, backordered, or rejected. The two classes are differentiated by their backorder and lost sales costs. At each decision epoch, we must determine whether or not to produce an item and if so, whether to use this item to increase inventory or to reduce backlog. At each decision epoch, we must also determine whether or not to satisfy demand from a particular class (should one arise), backorder it, or reject it. In doing so, we must balance inventory holding costs against the costs of backordering and lost sales. We formulate the problem as a Markov decision process and use it to characterize the structure of the optimal policy. We show that the optimal policy can be described by three state-dependent thresholds: a production base-stock level and two order-admission levels, one for each class. The production base-stock level determines when production takes place and how to allocate items that are produced. This base-stock level also determines when orders from the class with the lower shortage costs (class 2) are backordered and not fulfilled from inventory. The order-admission levels determine when orders should be rejected. We show that the threshold levels are monotonic (either non-increasing or non-decreasing) in the backorder level of class 2. We also characterize analytically the sensitivity of these thresholds to the various cost parameters. Using numerical results, we compare the performance of the optimal policy against several heuristics and show that those that do not allow for the possibility of both backordering and rejecting orders can perform poorly.

Optimal Integrated Production and Inventory Control of an Assemble-to-Order System with Multiple Non-Unitary Demand Classes

We study a pure assemble-to-order system subject to multiple demand classes where customer orders arrive according to a compound Poisson process. The finished product is assembled from m different components that are produced on m distinct production facilities in a make-to-stock fashion. We show that the optimal production policy of each component is a state-dependent base-stock policy and the optimal inventory allocation policy is a multi-level state-dependent rationing policy. Using numerical experimentation, we first study the system behavior as a function of order size variability and order size. We show that the optimal average cost rate is more sensitive to order size variability than to order size. We also compare the optimal policy to the first-come first-serve policy and show that there is great benefit to inventory rationing. We also propose two simple heuristics and show that these can effectively mimic the optimal policy which is generally much more difficult to determine and, especially, to implement.

Demand Allocation in Multiple-Product, Multiple-Facility, Make-to-Stock Systems

We consider the problem of allocating demand arising from multiple products to multiple production facilities with finite capacity and load-dependent lead times. Production facilities can choose to manufacture items either to stock or to order. Products vary in their demand rates, holding and backordering costs, and service-level requirements. We develop models and solution procedures to determine the optimal allocation of demand to facilities and the optimal inventory level for products at each facility. We consider two types of demand allocation, one in which we allow the demand for a product to be split among multiple facilities and the other in which demand from each product must be entirely satisfied by a single facility. We also consider two forms of inventory warehousing, one in which inventory locations are factory based and one in which they are centralized. For each case, we offer a solution procedure to obtain optimal demand allocations and optimal inventory base-stock levels. For systems with multiple customer classes, we also determine optimal inventory rationing levels for each class for each product. We use the models to characterize analytically several properties of the optimal solution. In particular, we highlight eight principles that relate the effects of cost, congestion, inventory pooling, multiple sourcing, customer segmentation, inventory rationing, and process and demand variability.

An Operational Decision Model for Lead-Time and Price Quotation in Congested Manufacturing Systems

This paper considers the problem of determining the lead-time and price to be quoted to a single order in a make-to-order manufacturing setting. The manufacturing system consists of several processing centers all subject to random failures and repairs. Because of a time window constraint on the delivery of the order, the latter has to be split among several processing centers to meet the constraint imposed on its delivery date. The assignment of lots to the processing centers is based on minimizing the operating cost associated with the entire order (i.e., its price). Two major cases are studied: the case of rushed order and the case of regular order. Each case has two options: Partial deliveries allowed and partial deliveries not allowed. Exact and heuristic algorithms are developed for each situation. Numerical results are used to draw conclusions and test the performance of the heuristics.

The use of flowlines to simplify routing complexity in two-stage flowshops

Flexible manufacturing systems are often designed as flowshops supported by automated material handling devices that facilitate routing among any two processors of adjacent stages. This routing structure is complex, and results in excessive capital investment and costs of management. In this paper we propose a decomposition of two-stage flowshops into smaller independent flowlines that allow for unidirectional routing only. We solve optimally the problem of minimizing makespan on two parallel flowlines, by means of a Dynamic Programming algorithm (DP). Based on DP we develop lower bounds on the throughput performance of environments that consist of more than two flowlines. We present several heuristic algorithms and report their optimality gaps. Using these algorithms, we show that the decomposition of two stage flowshops with complicated routing into flowline-like designs with unidirectional routing is associated with minor losses in throughput performance, and hence significant savings in material handling costs.

Optimal Lead-Time Planning in Serial Production Systems with Earliness and Tardiness Costs

We consider a production system consisting of N processing stages. The actual leadtimes at the stages are stochastic. The objective is to determine the planned leadtimes at each stage so as to minimize the expected total inventory costs, tardiness penalties, and a backlog penalty for not meeting demand due date at the last stage. Recursive relationships are used for automatic generation and efficient computation of the objective function. The efficiency of the proposed algorithms allows us to obtain new insights regarding operating policies, leadtime delivery reliability, and production line design. The problem is formulated as a convex nonlinear programming problem. The latter is then solved using classical convex optimization algorithms. For the special case of exponentially distributed leadtimes, the objective function is derived in closed form.

Optimal Control of an Assembly System with Multiple Stages and Multiple Demand Classes

We consider an assembly system with multiple stages, multiple items, and multiple customer classes. The system consists of m production facilities, each producing a different item. Items are produced in variable batch sizes, one batch at a time, with exponentially distributed batch production times. Demand from each class takes place continuously over time according to a compound Poisson process. At each decision epoch, we must determine whether or not to produce an item and, should demand from a particular class arise, whether or not to satisfy it from existing inventory, if any is available. We formulate the problem as a Markov decision process and use it to characterize the structure of the optimal policy. In contrast to systems with exogenous and deterministic production lead times, we show that the optimal production policy for each item is a state-dependent base-stock policy with the base-stock level nonincreasing in the inventory level of items that are downstream and nondecreasing in the inventory level of all other items. For inventory allocation, we show that the optimal policy is a multilevel state-dependent rationing policy with the rationing level for each demand class nonincreasing in the inventory level of all nonend items. We also show how the optimal control problem can be reformulated in terms of echelon inventory and how the essential features of the optimal policy can be reinterpreted in terms of echelon inventory. Subject classifications: production and inventory control; systems with multiple echelons; inventory rationing; Markov decision processes; make-to-stock queues.

Negotiating price/delivery date in a stochastic manufacturing environment

We study a make-to-order manufacturing system consisting of several processing centers that are subject to failures and repairs. Our objective is to build a model that can be used as a tool for negotiating the delivery date and the price of a certain upcoming order. The model takes into account the congestion level of the shop floor at the time the order is placed. Based on the workload of the processing centers, the model splits the order into lots and assigns them to the processing centers so as to determine the order completion time associated with the minimum operating cost. The efficiency of the solution method for the model allows real-time decision-making while negotiating the price and delivery date of the order to be placed. Since the decisions are made based on a snapshot of the congestion level at the shop floor, using this model will reduce the conflict between the marketing and the production activities in manufacturing organizations.

Optimal production control of a dynamic two-product manufacturing system with setup costs and setup times

This paper deals with the optimal control of a one-machine two-product manufacturing system with setup changes, operating in a continuous time dynamic environment. The system is deterministic. When production is switched from one product to the other, a known constant setup time and a setup cost are incurred. Each product has specified constant processing time and constant demand rate, as well as an infinite supply of raw material. The problem is formulated as a feedback control problem. The objective is to minimize the total backlog, inventory and setup costs incurred over a finite horizon. The optimal solution provides the optimal production rate and setup switching epochs as a function of the state of the system (backlog and inventory levels). For the steady state, the optimal cyclic schedule is determined. To solve the transient case, the systems state space is partitioned into mutually exclusive regions such that with each region, the optimal control policy is determined analytically.