Automating the Part Identification Method of Automotive Assembly Lines Through RFID Technology

Sammanfattning: Barcode scanning has been used for many decades in the assembly process to identify individual parts. Besides the fact that scanning is a non-value-adding operation, it also is prone to error. It cannot beensured with certainty that the scanned part will be the part installed. Furthermore, if any part is interchanged, it leaves the manufacturer with the challenge of detecting this and correcting the data. The genealogy data is important as it enables precisely tracing which parts are built into which vehicle. Strong confidence in the integrity of the genealogy data allows a manufacturer to minimize the scope assembly line uptime. When scan errors occur, the factory execution system could stop the production line to fix the issue and ensure high quality. Therefore, this thesis proposes an alternative and innovative approach to the part identification and verification process in an assembly line. The approach is to replace the traditional barcode with a passive ultra-high frequency RFID label. It automates the identification process when a part is installed in the vehicle, which makes manual scanning redundant. The suggested approach also proposes a final traceability scan. Hereby the completely assembled vehicle and its components with the RFID tags are read again to verify the same parts are still installed. The result would be enhanced genealogy data of each vehicle. This thesis aims to determine the technical feasibility of both processes and investigate the economic feasibility. The conducted empirical research of this thesis is based on a literature review about RFID technology and its applications. To prove the technical feasibility, a series of experiments were carried out for the in-station part identification and the final traceability verification. With a determined number of test parts, a total of 498 experiments were conducted in a real production environment. Moreover, the proposed dual-antenna approach and software logic enables accurate part identification. Lastly, for the assessment of the economic feasibility, a comprehensive data model was developed to assess the production impact of scanning.  Literature and a theoretical investigation show that most of the already consumed scan results can be related to human errors. The experiments for the automated in-station identification reveal; that it is possible to accurately identify the installed part under at least one setup with the suggested dual-antenna approach. However, every single part needs its setup adjusted to the environment in which it is assembled. There is not one out-of-the-box solution that suits every individual application. The finding from the final traceability scan experiment is that all tested parts are identified by the determined setup. It becomes apparent that reading the individual parts even after a car is completed is possible, despite the interference of the metal chassis and radio frequency waves. The conclusion from the economic feasibility is that although the RFID tags are more expensive than barcode labels, the implementation could still offer significant financial benefits to a manufacturer. To summarize the topic, the proposed method based on RFID technology is an innovative approach that is technically feasible and offers a variety of benefits.