(Special) Smart Factory Applications for The 4th Industrial Revolution
Chair: Prof. Namhun Kim (Ulsan National Institute of Science and Technology)
An Analysis of Quality Monitoring and Control System using Real-Time, Integrated Cost Effectiveness and Support Vector Machine Mr. YeongGwang OH, Kasin RANSIKARBUM, Moise BUSOGI, Daeil KWON and Namhun KIM (Ulsan National Institute of Science and Technology, Ubonratchatani University)
The quality monitoring and control (QMC) has been an essential process in the manufacturing industries to ensure product reliability. With the advancements in big-data analytics, machine-learning based QMC has become more and more popular in various manufacturing industries, such as automotive and electronic companies. At the same time, the cost effectiveness (CE) of the QMC is perceived as a main decision criterion that explicitly accounts for inspection efforts and has a direct relationship with the QMC capability. In this paper, the integrated support vector machine (SVM)-based automated QMC system with the adjusted CE, CEadj, model is proposed. Unlike existing models, the proposed model explicitly incorporates inspection-related expenses (i.e., warranty cost, rework cost, inspection cost) and error types (i.e., type-I and -II errors) in the CEadj framework to guide the SVM algorithm. The proposed automated QMC system is verified and validated using a door-trim manufacturing process of an automotive industry. Next, a designed experiment is performed to assess the sensitivity analysis of the proposed framework. The proposed model is found to be effective and could be exploited as an alternative or complementary tool for the traditional quality inspection system.
Adaptive SVM-based Real-time Quality Assessment for Primer-Sealer Dispensing Process of Sunroof Assembly Line Mr. Moise BUSOGI, YeongGwang OH, Kasin RANSIKABUM, Daeil KWON and Namhun KIM (Ulsan National Institute of Science and Technology, Ubonratchatani University)
Quality assessment in many production processes typically relies on manual inspections due to a lack of reference data and an effective method to classify defects in a systematic way. Recently, the real-time, automated approach for product quality assessment has been regarded an important aspect for smart manufacturing applications, such as in the automotive industry. In this research, we develop and implement the self-evolving quality assessment system based on the adaptive support vector machine (ASVM) model in the real production system of the primer-sealer dispending process of sunroof assembly line. An adaptive process is a feedback control that ensures the effectiveness of the SVM algorithm over time and enables the real-time improvement of SVM-based quality assessment. Next, an industrial case study of a primer-sealer dispensing process in a sunroof assembly line of an automobile is illustrated to verify and validate the applicability and effectiveness of the proposed ASVM-based quality assessment system. Defective patterns are analysed using an infrared thermal image of primer-sealer dispensing in a manufacturing process, which contains multi-modal data of dimensional information and temperature deviation from the dispending patterns in our study.
A Digital Signal-based Prognostic Approach to Factory Automation Mr. Jinwoo LEE and Daeil KWON (UNIST)
A data network from a control system is usually consist of complex wiring systems to communicate with manufacturing lines. The electrical connections by the wiring systems can be deteriorated by wire faults such as chafing in the wiring system. During the life cycle of the wiring system, field stress conditions such as mechanical stress condition often cause and intensify the wire faults . The progress of wire fault can cause other wire failures such as cutoff or arcing. Furthermore, failures in the connected wires may eventually cause malfunctions in facilities of manufacturing lines. In order to prevent serious failures, health of the wiring system can be monitored in real time to repair or replace the damaged wires before the time to failure. However, the conventional approaches to wire health monitoring often require additional connections to the wiring system due to additional monitoring devices. Thus, the operation of the facilities may be interfered when the monitoring devices monitor wires. As a result, the conventional approaches, which require additional devices, have difficulty detecting the extent of wire damages in real time, and may end up neglecting the progress of chafing.
In this study, a method for wire health monitoring is developed to prevent wire failures by monitoring wire damages in real time. Digital signal is affected adversely by impedance discontinuity on the transmission line. By monitoring the integrity of digital signal continuously, time to failure can be predicted in real time depending on the extent of wire damage. In addition, an accelerated wire abrasion test was designed to damage wires gradually. During the abrasion test, the integrity of the transmitted digital signal was continuously deteriorated. The monitored signal makes it possible to extrapolate the degradation pattern of the signal parameters depending on the extent of wire damage. The monitored signal makes it possible to extrapolate the degradation pattern of the signal parameters depending on the extent of wire damage. Thus, the results in this study validate that the proposed method is capable of monitoring the wire damage to diagnose a wiring system with real-time information.
Lifetime Prediction of Optocouplers in Digital Input and Output Modules based on Bayesian Tracking Approaches Ms. Insun SHIN and Daeil KWON (Ulsan National Institute of Science and Technology)
In recent years, reliability of DIO modules has been drawing much attention from manufacturing companies under the growing complexity of automation systems for smart factory establishment. In manufacturing systems, DIO modules have been widely used to pass sensor measurements and configuration input signals for controlling actuators. Because sensor measurement and control signals pass through DIO modules, the faults of DIO modules would cause malfunctions or failures of the smart manufacturing systems and eventually lead to unexpected downtime in the manufacturing process. For predictive maintenance of DIO modules, this paper proposes a method of predicting the remaining useful life of a critical component in DIO modules based on the Bayesian tracking approaches. Optocouplers are one of the critical components in DIO modules that uses a short optical transmission path including light sources and photo-sensors to transfer an electrical signal. The performance of optocouplers may be degraded overtime with damages in a light source or a photo-sensor and eventually cause the faults of control systems. Extended Kalman Filter and Particle Filter are used in nonlinear degradation modeling to predict the lifetime of optocouplers, evaluating those filters by accuracy-based prognostic metrics and showing the effectiveness of Bayesian tracking approaches for lifetime prediction of optocouplers.
A Study on a Reliability of a Mixed-model Assembly Line based on Complexity Mr. Moise BUSOGI, YeongGwang OH, Wooyeol LEE and Namhun KIM (Ulsan National Institute of Science and Technology)
In a manufacturing sector, most of manufacturers have concentrated their capability to develop a flexible manufacturing system which produces various products in small volume with limited resources. Especially in auto industries, maintaining higher diversity in models and options of their products is one of priorities in production to guarantee the market competitiveness. However, diversification of products causes a dramatic increase of manufacturing complexity and task difficulty in mixed-model assembly system when choosing proper part as variety-driven manufacturing environments. Nevertheless, quantitative indexes which estimate the manufacturing complexity are relatively insufficient. This research, thus, proposes a reliability-based complexity metric for effective assessment of mixed model assembly lines. We present the model and analysis of process complexity based on information entropy to estimate the manufacturing complexity of mixed-model production system in manufacturing industry. The proposed model is illustrated with a small scale assembly line with operators’ manual tasks to verify and validate its applicability and usefulness.