Research Initiation Award: Analysis and Control of Assembly Systems

点击翻译按钮获取中文摘要

9308290 Duenyas In the past decade, there has been a growing research interest in queueing network models of manufacturing systems. This interest is due to the ability of queueing network models to capture and evaluate the effects of variability and congestion. The recent interest in pull manufacturing systems such as kanban has increased the use of queueing network models. However, an important feature of many manufacturing systems that has not yet been satisfactorily incorporated in queueing network models is assembly operations. Assembly operations are operations that require as input the output of several different production lines. The aim of this project is to develop analytical models for the analysis and control of assembly systems. Bounds and approximations will be developed for such performance measures as the throughput, variance of the number produced in a given period, mean and variance of cycle time and mean and variance of queue lengths. Our preliminary results indicate that approximating assembly systems by equivalent tandem systems provides a promising appraoch. The analytical models will be used in algorithms for the optimal design of assembly systems. We will answer design questions such as how work should be allocated between different lines ad work centers, how to set production quotas, and Work In Process Inventory levels in these systems and how variability affects the optimal design and performance of assembly systems. The researchers will work with industry to test and implement our results. The result of the research will aid manufacturers signficantly in designing more efficient assembly systems.