Overview
Masters '04
Masters '02
Masters '01
Masters '00
Masters '99
Masters '98
Masters '97
Masters '96
Masters '95
Masters '94
Masters '93
Masters '92
Masters '91
Masters '90
Masters '89
Doctoral '90-'01
alumni

Masters Abstracts (2003)

KAWLI, VIRAJ
Ph.D Candidate, Univ of Missouri, Rolla, MO

Email: vkawli@yahoo.co.in

Scheduling for High Mix Low Volume Electronics Manufacturing Environments.

A typical surface mount Printed Circuit Board (PCB) assembly line consists of various operations that include solder paste deposition, component placement, and reflow soldering. Previous studies have shown that, typically, the component placement machine governs the throughput of the entire PCB assembly line. Hence, the component placement operation is considered as the bottleneck operation in a PCB assembly line. Enhancing the throughput of a PCB assembly line requires streamlining of the production process and efficient line balancing of component placement machines. These areas are strongly interrelated and cannot be addressed in isolation. Production planning and scheduling are crucial for enhancing the throughput of the component placement process in PCB manufacturing. They can be effective only when preceded by throughput enhancement techniques such as grouping of products and the appropriate arrangement and assignment of machine resources. Each of these throughput enhancement techniques cannot be applied individually as they are interrelated. Hence, the problem pertaining to the throughput enhancement of a component placement machine can be divided into three sub-problems - grouping of products, feeder template generation, and nozzle assignment. This research focused on improving the throughput of the component placement machine, and consequently the throughput of the entire PCB assembly line. Hence, the objective of this research was to develop a production planning and scheduling system to enhance the throughput of a PCB assembly line, through the use of throughput enhancement techniques such as Group Technology (GT) and feeder template generation. The process of enhancing throughput of a placement machine in a PCB assembly line was divided into sub-problems - grouping of products, selection of the best feeder arrangement, and sequencing of individual placements. Algorithms were developed for the grouping products, feeder template generation, and production scheduling. Based on these algorithms, software systems were developed in VC++ and C++. The design and implementation of these software systems were carefully analyzed, and their effects on the production process were reviewed. The system of grouping similar products together to reduce setup time and cycle time enhanced throughput by 15 percent, which is a significant achievement in a high volume PCB assembly environment. Feeder template generation was used to enhance the throughput and utilization of component placement machines. It was observed that the priority feeder template generation method helps to increase the throughput by 15 to 200 percent depending on the complexity of the product, type of product, and number of products in a group. The scheduling system that was developed for the component placement machine takes into account a composite objective function. This system was able to provide a schedule that was closer to the 'optimal' solution. It is an improvement on the current scheduling system that uses Moore's algorithm.

 

 
sitemap | contact us