• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.41 no.4
• /
• pp.59-69
• /
• 2018

This study deals with the case study on the pallet quantity determination problem for the flexible manufacturing system producing 32 different types of aircraft wing ribs which are major structures of an aircraft wings. A Korean company has constructed the WFMS (wing rib flexible manufacturing system) that is composed of several automated equipments such as the 5-axis machining centers, the RGV (rail guided vehicles)s, the AS/RS (automated storage and retrieval system), the loading/unloading stations, and so on. Pallets play a critical role in the WFMS to maintain high system utilization and continuous work flow between 5-axis machining machines and automated material handling devices. The discrete event simulation method is used to evaluate the performance of the WFMS under various pallet mix alternatives for wing rib manufacturing processes. Four performance measures including system utilization, throughput, lead-time and work in process inventory level are investigated to determine the best pallet mix alternative. The best pallet mix identified by the simulation study is adopted in setting up and operating a real Korean aircraft parts manufacturing shop. By comparing the real WFMS's performances with those of the simulation study, we discussed the cause of performance difference observed and the necessity of developing the CPS (cyber physical system).

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.41 no.3
• /
• pp.120-128
• /
• 2018

OLED Display fabrication system is one of the most complicated discrete processing systems in the world. As the glass size grows from $550{\times}650mm$ to $1,500{\times}1,850mm$ in recent years, the efficiency of Automated Material Handling System (AMHS) has become very important and OLED glass manufacturers are trying to improve the overall efficiency of AMHS. Aiming to meet the demand for high efficiency of transportation, various kind of approaches have been applied for improving dispatching rules and facility layout, while simultaneously considering the system parameters such as glass cassettes due date, waiting time, and stocker buffer status. However, these works did not suggest the operational policy and conditions of distribution systems, especially for handling unnecessary material flows such as detour. Based on this motivation, in this paper, we proposed an efficient algorithm for improving detour transportation in OLED FAB. Specifically, we considered an OLED FAB simplifying OLED production environment in a Korean company, where four stockers are constructed for the delivery of Lot in a bay and linked to processing equipments. We developed a simulation model using Automod and performed a numerical experiment using real operational data to test the performance of three operation policies under considerations. We showed that a competitive policy for assigning alternative stocker in case of detour was superior to the current dedicated policy using a specified stocker and other considered policies.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2014.10a
• /
• pp.331-337
• /
• 2014

Paradigm shift to cloud computing has increased the importance of security. Even though public cloud computing providers such as Amazon, already provides security related service like firewall and identity management services, it is not suitable to protect data in cloud environments. Because in public cloud computing environments do not allow to use client's own security solution nor equipments. In this environments, user are supposed to do something to enhance security by their hands, so the needs of visualized security management arises. To implement visualized security management, developing near realtime data handling & packet classification mechanisms are crucial. The key technical challenges in packet classification is how to classify packet in the manner of unsupervised way without human interactions. To achieve the goal, this paper presents automated packet classification mechanism based on naive-bayesian and packet Chunking techniques, which can identify signature and does machine learning by itself without human intervention.