Hyo-Young Kim
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Hyo-Young Kim has a total of 37 co-authors. Here are the top co-authors by number of shared publications:
- Kihyun Kim
Department of Mechatronics Engineering, Tech University of Korea, Siheung, Korea
8 shared publications - Kihyun Kim
Department of Mechatronics Engineering, Korea Polytechnic University, Siheung-si 15073, Korea
4 shared publications Eunjong Choi
Department of Mechatronics Engineering, Tech University of Korea, Siheung, Korea
3 shared publicationsJinsoo Choi
Department of Mechanical Engineering, Korea Advanced Institute of Science and Technology, Daejeon, Korea
3 shared publications- Chang-Hoon Seo
Mechanical Engineering, Ajou University 2 , 206, Worldcup-ro Yeongtong-gu, Suwon, Gyeonggi-do 16499, South Korea
2 shared publications Hak-Jun Lee
2 shared publications
Latest Publications
- Open Access
- 24 June 2026
- Published by Elsevier in SSRN Electronic Journal
Mutual Information-Based Wavelength Selection for Micro-Leak Detection in N2 SemiconductorProcess Environments
In semiconductor manufacturing processes such as etching, maintaining chamber integrity is essential for process stability and yield. Micro-cracks or gaps in the process chamber, arising from long-term material degradation and mechanical wear of chamber components, allow external air to infiltrate, degrading plasma conditions and production yield. Among available sensors, Self-Plasma Optical Emission Spectroscopy (SPOES) enables real-time in-situ monitoring during active processes, but is susceptible to noise and spectral interference in N2 process environments. To address spectral interference in N2 environments, this study proposes a leak detection framework combining a Mutual Information (MI)-based correlation wavelength selection method with SPOES. The optimal wavelength (245.2 nm) selected through MI-based data-driven analysis was spectroscopically confirmed to correspond to the NO 𝛾-band emission. This indicates that the system automatically identified the emission wavelength of the leak-induced reaction product (NO) without prior chemical knowledge, demonstrating that leakspecific signals can be extracted even in environments dominated by N2 background emission. The proposed framework, combining wavelet denoising and logistic regression, achieved 98.42% accuracy and a 0.99% false alarm rate over a leak concentration range of 10–40 ppm. A singlesample inference time of 22.9 𝜇s confirmed feasibility for real-time processing within the 50 ms SPOES measurement cycle.
- 24 June 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessMutual Information-Based Wavelength Selection for Micro-Leak Detection in N2 SemiconductorProcess Environments
In semiconductor manufacturing processes such as etching, maintaining chamber integrity is essential for process stability and yield. Micro-cracks or gaps in the process chamber, arising from long-term material degradation and mechanical wear of chamber components, allow external air to infiltrate, degrading plasma conditions and production yield. Among available sensors, Self-Plasma Optical Emission Spectroscopy (SPOES) enables real-time in-situ monitoring during active processes, but is susceptible to noise and spectral interference in N2 process environments. To address spectral interference in N2 environments, this study proposes a leak detection framework combining a Mutual Information (MI)-based correlation wavelength selection method with SPOES. The optimal wavelength (245.2 nm) selected through MI-based data-driven analysis was spectroscopically confirmed to correspond to the NO 𝛾-band emission. This indicates that the system automatically identified the emission wavelength of the leak-induced reaction product (NO) without prior chemical knowledge, demonstrating that leakspecific signals can be extracted even in environments dominated by N2 background emission. The proposed framework, combining wavelet denoising and logistic regression, achieved 98.42% accuracy and a 0.99% false alarm rate over a leak concentration range of 10–40 ppm. A singlesample inference time of 22.9 𝜇s confirmed feasibility for real-time processing within the 50 ms SPOES measurement cycle.
- Open Access
- 7 April 2026
- Published by JMIR Publications Inc. in JMIR Preprints
Goal Distance and Reward Contingency Predict Within-Session Cessation: Evidence from Large-Scale App Logs (Preprint)
BACKGROUND The mobile health (mHealth) and fitness app market has grown substantially, with global downloads of health and fitness applications reaching 3.6 billion in 2024. Despite this growth, the motivational dynamics that shape within-session engagement in mission-based fitness apps remain poorly understood. Although goal-gradient theory predicts effort acceleration as users approach a reward-contingent endpoint, it is unclear whether nonlinear motivational patterns emerge across the sequential stages of a single structured exercise session. OBJECTIVE This study aimed to examine whether within-session engagement in a mission-based fitness app exhibits (a) nonlinear motivational dynamics across sequential sets and (b) goal-proximal acceleration as users approach a reward-contingent completion boundary. METHODS We analyzed large-scale behavioral log data from Daysquat, a mission-based squat fitness app in which users complete four sets of ten squats per day to earn redeemable points. Stopping behavior across the structured four-set daily mission was modeled using a discrete-time proportional hazard model with covariates including set position, previous-day activity, user tenure, and platform-level engagement. A total of 221,461 user-day observations with at least one completed set, drawn from 10,725 registered users over a 187-day observation period, were included. Motivational intensity was additionally indexed by inter-set completion intervals among sessions in which all four sets were completed (n = 132,688). RESULTS The results revealed a non-monotonic, reverse-J-shaped engagement pattern: stopping risk declined across the first three sets (HR = 0.58 at set 3 vs. set 2) but increased sharply at the reward-contingent completion boundary (HR = 49.74 at set 4). Timing analyses converged with these findings; mean inter-set intervals peaked at the midpoint transition (set 2→3: 532 s) but declined substantially during the final transition (set 3→4: 143 s), consistent with goal-gradient acceleration. Higher previous-day activity (HR = 0.28), higher platform-level engagement (HR = 0.98), and longer user tenure (HR = 1.09) were also significantly associated with stopping risk. CONCLUSIONS Salient goal structure and contingent micro-incentives systematically shape motivation within a single exercise session. These findings advance understanding of within-session engagement dynamics in mHealth apps and suggest that graduated reinforcement strategies and optional continuation missions may help sustain engagement beyond minimum prescribed exercise doses.
- 7 April 2026
- Published by JMIR Publications Inc. in JMIR Preprints
Open AccessGoal Distance and Reward Contingency Predict Within-Session Cessation: Evidence from Large-Scale App Logs (Preprint)
BACKGROUND The mobile health (mHealth) and fitness app market has grown substantially, with global downloads of health and fitness applications reaching 3.6 billion in 2024. Despite this growth, the motivational dynamics that shape within-session engagement in mission-based fitness apps remain poorly understood. Although goal-gradient theory predicts effort acceleration as users approach a reward-contingent endpoint, it is unclear whether nonlinear motivational patterns emerge across the sequential stages of a single structured exercise session. OBJECTIVE This study aimed to examine whether within-session engagement in a mission-based fitness app exhibits (a) nonlinear motivational dynamics across sequential sets and (b) goal-proximal acceleration as users approach a reward-contingent completion boundary. METHODS We analyzed large-scale behavioral log data from Daysquat, a mission-based squat fitness app in which users complete four sets of ten squats per day to earn redeemable points. Stopping behavior across the structured four-set daily mission was modeled using a discrete-time proportional hazard model with covariates including set position, previous-day activity, user tenure, and platform-level engagement. A total of 221,461 user-day observations with at least one completed set, drawn from 10,725 registered users over a 187-day observation period, were included. Motivational intensity was additionally indexed by inter-set completion intervals among sessions in which all four sets were completed (n = 132,688). RESULTS The results revealed a non-monotonic, reverse-J-shaped engagement pattern: stopping risk declined across the first three sets (HR = 0.58 at set 3 vs. set 2) but increased sharply at the reward-contingent completion boundary (HR = 49.74 at set 4). Timing analyses converged with these findings; mean inter-set intervals peaked at the midpoint transition (set 2→3: 532 s) but declined substantially during the final transition (set 3→4: 143 s), consistent with goal-gradient acceleration. Higher previous-day activity (HR = 0.28), higher platform-level engagement (HR = 0.98), and longer user tenure (HR = 1.09) were also significantly associated with stopping risk. CONCLUSIONS Salient goal structure and contingent micro-incentives systematically shape motivation within a single exercise session. These findings advance understanding of within-session engagement dynamics in mHealth apps and suggest that graduated reinforcement strategies and optional continuation missions may help sustain engagement beyond minimum prescribed exercise doses.
- Open Access
- 1 January 2025
- Published by International Journal of Precision Engineering and Manufacturing-Smart Technology of Korean Society for Precision Engineering in International Journal of Precision Engineering and Manufacturing-Smart Technology
- Vol. 3 (1) , 17-30
Advanced Machine Learning Techniques for Predicting Z-Axis Belt Wear in Wafer Transfer Robots
Wafer-transfer robots (WTRs) are critical to semiconductor manufacturing, where precision and efficiency are essential to ensure high production yields and minimize defects. A key challenge is the wear and degradation of the Z-axis belt, which can cause misalignments, operational disruptions, and costly downtimes. This study presents a data-driven approach leveraging machine learning to predict Z-axis belt wear, enhancing WTR reliability and performance. High-frequency acceleration sensor data from the Upper Blade axis of WTRs were utilized to develop and test various machine learning classification models, including K-Nearest Neighbors (KNN), Logistic Regression, Naive Bayes, Decision Tree, and Random Forest. The Random Forest model achieved the highest predictive accuracy at 99.8%, significantly surpassing traditional maintenance methods. These findings highlight the potential of machine learning to transform predictive maintenance in semiconductor manufacturing. By anticipating and preventing faults with exceptional precision, machine learning-based predictive maintenance enables more reliable and efficient operations, reducing costs and enhancing system longevity. This study not only demonstrates the effectiveness of machine learning in predictive maintenance but also sets the foundation for future applications aimed at optimizing the performance and lifespan of critical manufacturing equipment.
- 1 January 2025
- Published by International Journal of Precision Engineering and Manufacturing-Smart Technology of Korean Society for Precision Engineering in International Journal of Precision Engineering and Manufacturing-Smart Technology
- Vol. 3 (1) , 17-30
Open AccessAdvanced Machine Learning Techniques for Predicting Z-Axis Belt Wear in Wafer Transfer Robots
Wafer-transfer robots (WTRs) are critical to semiconductor manufacturing, where precision and efficiency are essential to ensure high production yields and minimize defects. A key challenge is the wear and degradation of the Z-axis belt, which can cause misalignments, operational disruptions, and costly downtimes. This study presents a data-driven approach leveraging machine learning to predict Z-axis belt wear, enhancing WTR reliability and performance. High-frequency acceleration sensor data from the Upper Blade axis of WTRs were utilized to develop and test various machine learning classification models, including K-Nearest Neighbors (KNN), Logistic Regression, Naive Bayes, Decision Tree, and Random Forest. The Random Forest model achieved the highest predictive accuracy at 99.8%, significantly surpassing traditional maintenance methods. These findings highlight the potential of machine learning to transform predictive maintenance in semiconductor manufacturing. By anticipating and preventing faults with exceptional precision, machine learning-based predictive maintenance enables more reliable and efficient operations, reducing costs and enhancing system longevity. This study not only demonstrates the effectiveness of machine learning in predictive maintenance but also sets the foundation for future applications aimed at optimizing the performance and lifespan of critical manufacturing equipment.
- 20 July 2024
- Published by Springer Nature in International Journal of Precision Engineering and Manufacturing
- Vol. 25 (9) , 1851-1863
Improving Path Accuracy and Vibration Character of Industrial Robot Arms with Iterative Learning Control Method
- 20 July 2024
- Published by Springer Nature in International Journal of Precision Engineering and Manufacturing
- Vol. 25 (9) , 1851-1863
Improving Path Accuracy and Vibration Character of Industrial Robot Arms with Iterative Learning Control Method
- Open Access
- 11 June 2024
- Published by Springer Nature in International Journal of Precision Engineering and Manufacturing
- Vol. 25 (8) , 1755
Correction: An Active Geophone Sensor with Optimized State Variable Filter for Measuring Low-Band Frequencies
- 11 June 2024
- Published by Springer Nature in International Journal of Precision Engineering and Manufacturing
- Vol. 25 (8) , 1755
Open AccessCorrection: An Active Geophone Sensor with Optimized State Variable Filter for Measuring Low-Band Frequencies
- 12 March 2024
- Published by Springer Nature in Journal of Mechanical Science and Technology
- Vol. 38 (3) , 1403-1410
Development of an ultrasonic-assisted soldering system for Fe-based lead frame surface coating
- 12 March 2024
- Published by Springer Nature in Journal of Mechanical Science and Technology
- Vol. 38 (3) , 1403-1410
Development of an ultrasonic-assisted soldering system for Fe-based lead frame surface coating
- Open Access
- 11 March 2024
- Published by Springer Nature in International Journal of Precision Engineering and Manufacturing
- Vol. 25 (5) , 1125
Correction to: An Active Geophone Sensor with Optimized State Variable Filter for Measuring Low‑Band Frequencies
- 11 March 2024
- Published by Springer Nature in International Journal of Precision Engineering and Manufacturing
- Vol. 25 (5) , 1125
Open AccessCorrection to: An Active Geophone Sensor with Optimized State Variable Filter for Measuring Low‑Band Frequencies
- 6 February 2024
- Published by Springer Nature in International Journal of Precision Engineering and Manufacturing
- Vol. 25 (5) , 981-993
An Active Geophone Sensor with Optimized State Variable Filter for Measuring Low-Band Frequencies
- 6 February 2024
- Published by Springer Nature in International Journal of Precision Engineering and Manufacturing
- Vol. 25 (5) , 981-993
An Active Geophone Sensor with Optimized State Variable Filter for Measuring Low-Band Frequencies