| Professor Marc E. Posner
Industrial and Systems Engineering faculty, The Ohio State University
|
|
【主题】Modern Factory Layout Design
【时间】2005-6-28上午9:30
【地点】经管学院伟伦楼中402
【语言】英文
【内容摘要】
The Functional, Cellular and Flow line layouts are traditional facility layouts that are discussed in the literature and implemented in industry. This research is motivated by the inability to design an acceptable layout for a semiconductor fabrication facility. Due to the variety of products produced, none of the three traditional layouts provides a design that fits the material flow network.
We develop a detailed formulation that models the trade-offs between the problem costs. These costs include WIP, material handling, machine setup, and processing. Then, we construct the facility layout as a network of layout "modules". Each module has the material flow pattern of one of the traditional layouts.
One portion of this problem is to partition machines into parallel work centers, and simultaneously assign job types to work centers. Each job type has a distinct Poisson arrival rate and a distinct WIP weight. Each work center is a queuing system and each machine has an exponential service time distribution. The goal is to minimize the total WIP cost. The problem is polynomial if there are identical machines, only two work centers, and job types can be divided between work centers. More general problems are NP-hard. Analysis demonstrates that a simple heuristic finds an optimal solution for a large class of problem instances. The heuristic is extended to the multiple work center case where the relative error is bounded above by 1+e. When machines are not identical, heuristics are constructed and analyzed for the random available and fastest available machine assignment disciplines.
【主讲人简介】
Marc E. Posner is a Professor of Operations Research in the Industrial, Welding and Systems Engineering Department at The Ohio State University. He received a B.A. in Mathematics from Brandeis University, an M.S. and Ph.D. in Operations Research from the University of Pennsylvania. He has published in most of the major operations research journals on a variety of topics ranging from the construction of statistical decision rules to decomposing nonlinear programming problems. His research is primarily in the field of deterministic optimization with an emphasis on integer programming. He is interested both in heuristic and exact methods. An area of focus is on scheduling and production problems. Currently, he is the Department Editor in Scheduling and Logistics for IIE Transactions and is an Associate Editor for Naval Research Logistics. |
