Author:  Wan, Chunglai Johnny 
Title:  Postponement strategy in supply chain management 
Year:  2004 
Subject:  Hong Kong Polytechnic University  Dissertations Business logistics Industrial procurement  Management Materials management Physical distribution of goods 
Department:  Dept. of Logistics 
Pages:  x, 211 leaves : ill. ; 30 cm 
Language:  English 
InnoPac Record:  http://library.polyu.edu.hk/record=b1781116 
URI:  http://theses.lib.polyu.edu.hk/handle/200/4726 
Abstract:  Postponement is a supply chain strategy that calls for the design of a supply chain to achieve both low cost and fast response by combining some common processes and delaying other product differentiation processes such as packaging and labelling. The point that separates the differentiation processes from the common processes is known as the point of product differentiation. Recent research studies present four common postponement strategies, namely pull, logistics, form and price postponements. They aim at balancing the costs and benefits of mass production and customization. In this thesis, four models are presented to evaluate the impacts of pull and form postponement strategies under various supply chain structures. First, we developed an EOQbased model to examine the impact of pull postponement. We formulated a total average cost function for ordering and keeping n endproducts in a supply chain, in which their demands are known and deterministic. The function consists of fixed ordering cost, variable cost, inventory holding cost and backorder cost. By optimization technique, we showed that postponed customization of endproducts will result in a lower total average cost and a lower EOQ. In the second model, we developed an EPQbased model to examine the impact of postponement. We formulated a total average cost function for producing and keeping n endproducts in a supply chain, in which their demands are known and deterministic. The function consists of fixed and variable production costs, and inventory holding cost. By optimization technique, we showed that postponed customization of endproducts results in a lower total average cost in certain circumstances. The third one is a stochastic model of a single endproduct supply chain that consists of a supplier, a manufacturer and a number of customers. The customer demand per period is discrete and uniformly distributed, and the order leadtime is constant. Two distinct models were developed to represent the inventory system of the manufacturer. The first one is a lot sizereorder point (r, q) inventory system and the second one is a postponement system. We aimed at finding an optimal inventory policy that can minimize a total average cost function per period that comprises fixed ordering cost, variable cost, inventory holding cost and backorder cost. We employed a Markov chain analysis to determine the exact average inventory level and the exact average accumulated backorder per period at the steady state so that the total average cost can be evaluated analytically. Also, we designed an algorithm for searching for a near optimal total average cost per period. The optimal total average cost was compared with the total average cost of a postponement system by both optimization and simulation techniques. Our results showed that the postponement system is more cost effective when the leadtime is zero, while the (r, q) inventory system is more effective when the leadtime is greater than zero. In the last model, we aimed at conducting a simulation experiment of a twoendproduct supply chain, for which customer demands are discrete and independent. Customer demands follow a uniform, Poisson or normal distribution. Two simulation models, namely one is a postponement system while the other is a nonpostponement system, were designed for comparing their performance and total cost after t periods. In the postponement system, we assumed the manufacturer orders and stores workinprocess inventory (WIP) of the two endproducts based on an (r, q) policy. When the customer demands are known, WIP is transformed into required endproducts in a short processing time. In the nonpostponement system, the manufacturer orders and stores the endproducts based on two independent (r, q) policies. Given a set of (r, q) and a demand distribution, the postponement system outperforms the nonpostponement system in terms of average order frequency, average onhand inventory, average backorder and average fillrate. Thus, this system provides some cost benefits when the net postponement cost is low. Besides mathematical models, a study of a Hong Kong based manufacturer, who has adopted postponement, was presented to support our theoretical results. 
Files  Size  Format 

b17811168.pdf  5.212Mb 


As a bona fide Library user, I declare that:  


By downloading any item(s) listed above, you acknowledge that you have read and understood the copyright undertaking as stated above, and agree to be bound by all of its terms. 