dc.description.abstract | Distribution centers have been playing an important role in delivering goods from manufacturers to the end users. The development of computer, information, and communication technologies has made everything go faster than it used to. For a distribution center, this means more frequent and faster delivery of goods to its customers. For years, industrial engineers have been looking for ways to optimize the order-picking operations in a distribution center. It has been shown by researchers that the cost of order picking comprises about 50% or more of the total operation cost of a distribution center (Hwang et al., 1988). And, it has been shown that a distribution center can boost its competitive advantages if its order-picking cost can be reduced (Ballou, 1992, Morabito et al., 2000).
The purpose of this paper is to look for way to make the order-picking operations more efficient and effective. Problems investigated in this paper include order-batching problem, route-planning problem, and the order-splitting problem. For each problem, several solutions will be proposed. For order-batching problem, four seed-order selection algorithms (i.e., RD, LNVL, LNI, MRV) and four accompanying-order selection algorithms (i.e., MAD, MCVL, MCI, SNAA) will be proposed and investigated. They will make up 16 different combinations of seed-order selection algorithm and accompanying-order selection algorithm. As for the route-planning problem, four methods (i.e., NCGI, LG, NCRI, MTLI) will be investigated. The first method is obtained from Ho and Lo (2001), while the second one is called the largest-gap algorithm that was originally proposed by (Hall, 1993). And, the last two methods are proposed by us. It is a two-stage approach that finds an initial route first and then uses a heuristic optimization technique to improve the result. Finally, for the order-splitting problem, three algorithms (i.e., RD, MDPD, MIV) will be proposed and investigated. An order will be divided into two orders if its required space is greater than the remaining space in the picking cart.
To understand the performance of the above algorithms and methods, simulation experiments (using Arena) will be conducted. The simulation results not only can tell us which algorithm can perform better than others but also can help us understand their relationships. For example, one may see how one seed-order-selection algorithm is affected by another accompanying-order selection algorithm. It is our hope that the knowledge learned from this research can help us find the best solution for each of the above problems so that more effective and efficient order-picking operations can be obtained. | en_US |