Collaborative agent based mass customization manufacture information system research

• Main Authors: Lu, Tung-Hong, 盧東宏
• Other Authors: Amy J. C. Trappey
• Format: Others
• Language: zh-TW
• Published: 2009
• Online Access: http://ndltd.ncl.edu.tw/handle/86845309083885770880

博士 === 國立清華大學 === 工業工程與工程管理學系 === 97 === The production model is heading from mass production toward mass customization in these days. Due to the shorter product lifecycle, the demand of product customization and variety is increasing, and which leads the production model to a new and chanllengin direction. This research dedicate on development a mass customization information platform to support comsumers on customization process and suppliers on manufacturing activities as well. This platform provides functions to coordinate different behaviors including comsumer requirment collection, purchase order tranformation, purchase order management and suppliers manufacturing execution by employing the multi-agents based system which is called "Agent based Mass Customization Information Service Platform (AMCISP)". The content of this research can be divided into three parts. To be more specific, the first part of this reseach is focused on modeling of the mass production and custmization process and the associated analysis done by using Petri-Net. The process model came from the process analysis named as “As-Is process model”. Then, the mass customization service process is proposed based on the results of As-Is process analysis. The process model is so-called “To-Be process model”. The second part of this research is focused on the information requirment analysis which is mainly extracted from the To-Be process model. The following part of this reseach describes the implementation details of the AMCISP such as the intelligent agent system architecture, communication models, and collabortation behaviors. In addition, the RFID technology is involved for the improvement of data acquisition effectiveness in the distributed operation. A new application model of RFID systems in manufacturing environment is presented in this stage. Finally, a customization bicycle scenario is used to demostarte the customization requirements collection process, automatic parts selection and the execution detail of the AMCISP. The contributions of this research are developing the new methodology for constructing an information platform to support the mass customization manufacturing, and a new distributed mechanism of monitoring the manufacturing process among outsourcing partners in real time using agent technology.