Development of an RFID-based quality management system in distributed manufacturing network

• Main Author: Lv, Yaqiong
• Other Authors: Lee Ka Man, Carman
• Format: Theses and Dissertations
• Language: English
• Published: 2013
• Subjects: DRNTU::Engineering::Manufacturing::Quality control; DRNTU::Engineering::Industrial engineering::Quality engineering
• Online Access: https://hdl.handle.net/10356/53651
• Institution: Nanyang Technological University

Due to globalization and outsourcing, products are fabricated by different parties in the distributed manufacturing network and enterprises face the challenge of assuring the quality of semi-finished product in each manufacturing process in different geographical locations. A Radio Frequency Identification (RFID)-based quality management system in the distributed manufacturing network has been proposed in this research. The feature of the proposed system is to model and simulate large distributed manufacturing networks to streamline the mechanical and production engineering processes with RFID technology, which can keep track of process variants The proposed RFID-based quality management system consists of three sub-systems including Petri-net modeling system, fuzzy quality management system and RFID system. In Petri-net modeling system, the components of the manufacturing systems we consider exhibit concurrency or parallelism, which means that activities of one component may occur simultaneously with other components. To deal with concurrency and synchronization, the Hierarchical Colored Timed Petri Net formalism for modeling is described in this study. This method can help to model graphically and test the discrete events of concurrent operations. Fuzzy quality management system can help for knowledge representation, so as to provide knowledge-based decision assistance in distributed manufacturing environment. The skill and experience of quality from production engineers was acquired and stored in knowledge bases, to ensure the product with verifiable quality. Fuzzy clustering approach is adopted by in this research. In the first stage, the raw data during manufacturing process can be classified based on its own nature, and with the classification results, operation decision can be made in the second stage by running fuzzy inference engine without relying on highly skilled expert. As the information transmitting in the system is with vague information, fuzzy quality engine becomes a more useful tool for represent knowledge in the real world with an approximate method. In the thesis, how to define the fuzzy rule for the system and the associated membership functions’ development is presented. In the RFID system, a new RFID-based CPN modeling method where the colored tokens are evolved to color tagged tokens carrying the product information of real-time status. With this new real-time modeling method, the performance of manufacturing systems such as yield rate and throughput can be realized. RFID enable automatic data capture and monitor the quality of the production process in the distributed manufacturing environment. In order to illustrate how the proposed quality management system works, two case studies in a worldwide company with the alias name as GG Ltd , is presented in the last part of the thesis. This research is contribute to design and develop a RFID-based quality management system with the distinct feature of fuzzy logic and statistics analysis to achieve progressive knowledge acquisition and creation to the entity it is associated with, thereby assuring the quality of the products. In particular, this proposed system is characterized by the inclusion of RFID technology which has been recognized as one of proven technology that can streamline the logistics operations. This proposed RFID-based quality management system may have significant impact on the smart track and control the manufacturing process by virtue of its capabilities to progressively introduce machine intelligence to support relevant activities within the distributed manufacturing network.