Jin-Hyeon Sung

Mr. Jin-Hyeon Sung

orcid

Research Focus & Metrics

Academic Career Length

2Years (2025-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
2 April 2025 in Mathematics
Verified
Journal Article
Open access
PDF file
JSJin-Hyeon Sung
Jin-Hyeon Sung
SJSeong-Hyeon Ju
Seong-Hyeon Ju
SCSeung-Wan Cho
Seung-Wan Cho
HJHak-Jong Joo
Hak-Jong Joo
KSKyung-Min Seo
Kyung-Min Seo
BKBonggu Kang
Bonggu Kang
The optimization of overhead hoist transport (OHT) systems in semiconductor manufacturing plays a crucial role in improving production efficiency. In this study, the development of a discrete event simulation model to analyze the physical and control characteristics of an OHT system is presented, focusing on building a modular simulation framework for evaluating operational strategies by applying various optimization techniques. Additionally, a step-by-step analysis is introduced to optimize OHT operation using the developed model. The simulation model is broadly divided into three parts according to their purposes. The physical system encompasses the physical entities such as the equipment and vehicles. The experimental frame comprises a generator, which triggers experiments, and a result analyzer. Finally, the system controller is structured hierarchically and consists of an upper layer, known as the manufacturing control system, and subordinate layers. The subordinate layers are modularly divided according to their roles and encompass a main controller responsible for OHT control and a scheduling agent manager for dispatching and routing based on SEMI commands. The proposed simulation model adopts a structure based on the discrete event systems specification (DEVS). Since the hierarchical system controller may face challenges such as computational overhead and adaptability issues in real-world implementation, the modular design based on DEVS is utilized to maintain independence between layers while ensuring a flexible system configuration. Through an exploratory analysis using the simulation model, we adopt a step-by-step approach to optimize the OHT operation. The optimal operation is achieved by identifying the optimal number of OHT units and pieces of equipment per manufacturing zone. The results of the exploratory analysis for the three scenarios validate the effectiveness of the proposed framework. Increasing the number of OHT units beyond 17 resulted in only a 0.08% reduction in lead time, confirming that 17 units is the optimal number. Additionally, by adjusting the amount of equipment based on their utilization rates, we found that reducing the amount of equipment from 12 to five in process E-1 and from seven to three in the OUT process did not degrade performance. The proposed simulation framework was thus validated as being effective in evaluating OHT operational efficiency and useful for analyzing key performance indicators such as OHT utilization rates. The proposed model and analysis method effectively model and optimize OHT systems in semiconductor manufacturing, contributing to improved production efficiency and reduced operational costs. Furthermore, this work can bridge the gap between theoretical modeling and practical complexities in semiconductor logistics.

1 July 2025 in Systems
Verified
Journal Article
Open access
PDF file
JWJi-Min Woo
Ji-Min Woo
SJSeong-Hyeon Ju
Seong-Hyeon Ju
JSJin-Hyeon Sung
Jin-Hyeon Sung
KSKyung-Min Seo
Kyung-Min Seo
In recent manufacturing environments, the use of digitally retrofitted equipment has grown substantially, yet this trend also amplifies the challenge of ensuring stable operation through effective anomaly detection. Retrofitted systems suffer from two critical obstacles: a severe scarcity of labeled data and substantial variability in operational patterns across machines and products. To overcome these issues, this study introduces a novel anomaly detection framework that integrates Model-Agnostic Meta-Learning (MAML) with a Long Short-Term Memory Autoencoder (LSTM-Autoencoder) under a multi-machine-based task formulation. By constructing meta-tasks from time-series datasets collected on multiple five-axis computer numerical control (CNC) machines, our method enables rapid adaptation to unseen machines and production scenarios with only a few training examples. The experimental results demonstrate that, even under data-scarce conditions, the proposed model achieves an accuracy of 98.02% and an F1-score of 94.74%, representing improvements of 4.2 percentage points in accuracy and 16.9 percentage points in F1-score over conventional transfer learning approaches. Furthermore, in cross-validation on entirely new machine data, our framework outperforms existing models by 18.1% in accuracy, evidencing superior generalization capability. These findings suggest that the proposed multi-machine-based Model-Agnostic Meta-Learning Long Short-Term Memory Autoencoder (MAML LSTM-Autoencoder) can significantly enhance operational efficiency and reduce maintenance costs in retrofitted manufacturing equipment, thereby improving overall productivity and paving the way for real-time industrial deployment.

Verified
Journal Article
Open access
PDF file
HPHee-Mun Park
Hee-Mun Park
JSJin-Hyeon Sung
Jin-Hyeon Sung
HPHong-Sun Park
Hong-Sun Park
YLYeong-Hyun Lim
Yeong-Hyun Lim
JSJoono Sur
Joono Sur
KSKyung-Min Seo
Kyung-Min Seo
Swarm Unmanned Surface Vehicles (SUSV) must sustain real-time coordination as fleet scale and formation change. We propose hierarchical and modular architecture that decouples mission-planning algorithms from interface evolution, composed of an Interface Adapter System (IAS), a Mission Execution System (MES), and an Interoperation Integration Verification System (IIVS). The IAS standardizes and integrates data from diverse sensors and subsystems through modular adapters, facilitating flexible subsystem integration. The MES employs a discrete-event system specification (DEVS)-based modeling approach, providing independent and efficient mission execution capability without necessitating interface modifications. The IIVS utilizes LabVIEW-based analytical methods to verify and validate subsystem interoperability continuously, enabling rapid and reliable scaling. The architecture was implemented in an operational SUSV program and evaluated through simulation experiments and sea trials. During scale-up from 10 to 20 USVs, mission-cycle deadlines (≤250 ms) were met in 99% of cases, while integration lead time for scale-up decreased by about 80%. Message-level tests confirmed robust interoperation under increased load, and algorithm-level tests showed stable plan re-computation under dynamic tasking. These results indicate improved interoperability, scalability, and reliability, offering a practical blueprint for mission planning in maritime swarms.