Jinhwan Jang
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Jinhwan Jang has a total of 7 co-authors. Here are the top co-authors by number of shared publications:
- Dongmahn Seo
School of Computer Software Daegu Catholic University Gyeongsan Republic of Korea
1 shared publication - Hyeokgyu Kwon
Department of Computer Software, Daegu Catholic University, Gyeongsan-si, Republic of Korea
1 shared publication Jongsik Kim
Highway Research Division, Korea Institute of Civil Engineering and Building Technology, Goyang-si, Republic of Korea
1 shared publication- Nancy Huynh
University of Alberta
1 shared publication - Soobin Jeon
School of Computer Software, Daegu Catholic University, Gyeongsan-si, Republic of Korea
1 shared publication - Tae Jung Kwon
Univ. of Alberta
1 shared publication
Latest Publications
- Open Access
Preventing Black Ice Caused by Freezing Rain and Frost on Sensorless Rural Highways: Winter Maintenance in South Korea
Black ice poses a significant threat to drivers during winter due to its low visibility. Winter road maintenance personnel face continuous challenges in preventing its formation because of its high unpredictability. To address this issue, this paper proposes a practical strategy for effective winter road maintenance aimed at preventing black ice caused by freezing rain and frost. The strategy comprises three phases: black ice prediction, stakeholder notification, and anti-icing chemical application. The core of the strategy involves predicting black ice on rural highways that lack localized road weather sensors. Specifically, the prediction model relies exclusively on atmospheric data. The Extreme Gradient Boosting (XGBoost) algorithm was employed for prediction, achieving precision, recall, and F1 scores of 0.99 and outperforming Random Forest and Deep Neural Network models, which achieved F1 scores of 0.96 and 0.97, respectively. The XGBoost model’s hyperparameters were optimized using the DEPSO algorithm, improving its F1 score by approximately 0.03. Furthermore, a feature importance analysis was conducted to determine the relative contribution of various meteorological variables to black ice formation. To effectively disseminate predictive alerts to maintenance personnel and drivers, a smartphone-based system was developed. Finally, optimal spread rates for anti-icing chemicals, calibrated to pavement temperatures, are presented. The methodology proposed in this study can significantly enhance the efficiency of winter road maintenance on rural highways where localized road weather data are unavailable.
Open AccessPreventing Black Ice Caused by Freezing Rain and Frost on Sensorless Rural Highways: Winter Maintenance in South Korea
Black ice poses a significant threat to drivers during winter due to its low visibility. Winter road maintenance personnel face continuous challenges in preventing its formation because of its high unpredictability. To address this issue, this paper proposes a practical strategy for effective winter road maintenance aimed at preventing black ice caused by freezing rain and frost. The strategy comprises three phases: black ice prediction, stakeholder notification, and anti-icing chemical application. The core of the strategy involves predicting black ice on rural highways that lack localized road weather sensors. Specifically, the prediction model relies exclusively on atmospheric data. The Extreme Gradient Boosting (XGBoost) algorithm was employed for prediction, achieving precision, recall, and F1 scores of 0.99 and outperforming Random Forest and Deep Neural Network models, which achieved F1 scores of 0.96 and 0.97, respectively. The XGBoost model’s hyperparameters were optimized using the DEPSO algorithm, improving its F1 score by approximately 0.03. Furthermore, a feature importance analysis was conducted to determine the relative contribution of various meteorological variables to black ice formation. To effectively disseminate predictive alerts to maintenance personnel and drivers, a smartphone-based system was developed. Finally, optimal spread rates for anti-icing chemicals, calibrated to pavement temperatures, are presented. The methodology proposed in this study can significantly enhance the efficiency of winter road maintenance on rural highways where localized road weather data are unavailable.
- Open Access
- 28 April 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 13, 76109-76120
Implementation of a Real-Time Analysis Simulation System for Traffic Signal Control Algorithms to Reduce Vehicle Exhaust Emissions and Improve Traffic Flow
Traffic signal control (TSC) is a part of intelligent transportation systems to reduce traffic congestion and emissions. Recently, dynamic traffic signal control systems using artificial intelligence and reinforcement learning have been studied to achieve these goals. Although verification of TSC algorithms is ideally performed in a real environment, traffic simulation tools are widely used due to safety issues. However, existing simulation tools have limitations in real-time result analysis, making it difficult to analyze the strengths and weaknesses of each algorithm according to each traffic situation. In this paper, we propose RTASS, a real-time analysis simulation system for TSC algorithms that enables real-time comparison and result analysis, to improve the problems of such a simulation environments. To validate RTASS, we experiment using actual intersection data. As a result of comparing the traffic volume of the actual intersection data and the simulation, we confirmed that the error rate was maintained within 2% on average. In addition, we evaluate the performance of each TSC algorithm using RTASS and verify the results by comparing two existing methods and two proposed TSC methods. We confirmed that real-time comparison of traffic analysis and algorithms is possible using the proposed simulation system.
- 28 April 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 13, 76109-76120
Open AccessImplementation of a Real-Time Analysis Simulation System for Traffic Signal Control Algorithms to Reduce Vehicle Exhaust Emissions and Improve Traffic Flow
Traffic signal control (TSC) is a part of intelligent transportation systems to reduce traffic congestion and emissions. Recently, dynamic traffic signal control systems using artificial intelligence and reinforcement learning have been studied to achieve these goals. Although verification of TSC algorithms is ideally performed in a real environment, traffic simulation tools are widely used due to safety issues. However, existing simulation tools have limitations in real-time result analysis, making it difficult to analyze the strengths and weaknesses of each algorithm according to each traffic situation. In this paper, we propose RTASS, a real-time analysis simulation system for TSC algorithms that enables real-time comparison and result analysis, to improve the problems of such a simulation environments. To validate RTASS, we experiment using actual intersection data. As a result of comparing the traffic volume of the actual intersection data and the simulation, we confirmed that the error rate was maintained within 2% on average. In addition, we evaluate the performance of each TSC algorithm using RTASS and verify the results by comparing two existing methods and two proposed TSC methods. We confirmed that real-time comparison of traffic analysis and algorithms is possible using the proposed simulation system.
- Open Access
Travel Time Information on Signalized Arterials
Travel time information has become an essential component of everyday commuting. Without such information, schedule delays would increase, leading to inevitable losses in traveler utility. In Korea, dedicated short-range communication transponders that identify vehicles have been installed along signalized arterials to collect travel time data. By matching vehicle identifications at consecutive points, travel times can be measured. However, for travel time information to be effective, two types of data processing techniques are required: outlier filtering and travel time prediction. This study proposes algorithms to address both challenges. An outlier filtering algorithm based on the median-based confidence interval was developed, taking into account the travel time characteristics on suburban arterials with frequent entry and exit points. Additionally, a travel time prediction algorithm that integrates Long Short-Term Memory (LSTM) and Convolutional Neural Networks (CNNs), referred to as LSTM-CNN, was developed to capture both long-term trends and local patterns in travel time data. The implementation of these algorithms resulted in a 2.2% reduction in error rates under congested conditions compared to current practices. At the 4 km study site, the annual benefits from this error reduction could amount to USD 135,200.
Open AccessTravel Time Information on Signalized Arterials
Travel time information has become an essential component of everyday commuting. Without such information, schedule delays would increase, leading to inevitable losses in traveler utility. In Korea, dedicated short-range communication transponders that identify vehicles have been installed along signalized arterials to collect travel time data. By matching vehicle identifications at consecutive points, travel times can be measured. However, for travel time information to be effective, two types of data processing techniques are required: outlier filtering and travel time prediction. This study proposes algorithms to address both challenges. An outlier filtering algorithm based on the median-based confidence interval was developed, taking into account the travel time characteristics on suburban arterials with frequent entry and exit points. Additionally, a travel time prediction algorithm that integrates Long Short-Term Memory (LSTM) and Convolutional Neural Networks (CNNs), referred to as LSTM-CNN, was developed to capture both long-term trends and local patterns in travel time data. The implementation of these algorithms resulted in a 2.2% reduction in error rates under congested conditions compared to current practices. At the 4 km study site, the annual benefits from this error reduction could amount to USD 135,200.
- Open Access
- 1 March 2025
- Published by MDPI AG in Future Transportation
- Vol. 5 (1)
Optimizing RWIS Locations with Wasserstein Distance and Geostatistics: A Case Study in South Korea
Road Weather Information Systems (RWISs) are essential components of modern Intelligent Transportation Systems (ITSs) deployed in cold regions to gather real-time data on winter weather and road surface conditions. Despite their benefits, the high cost associated with RWIS installations demands optimized placement strategies to maximize their utility and cost-effectiveness. Geostatistics-based RWIS location-allocation methods, particularly those involving semivariogram modeling to quantify underlying spatial characteristics, have gained international recognition. However, new locations require unique semivariogram models, a process that is time-consuming and constrained by the availability of comprehensive datasets, often rendering location analysis challenging or infeasible. Addressing these limitations, this study introduces an innovative approach using Wasserstein Distance (WD) to link semivariograms across different datasets. This method streamlines optimization by eliminating the need for repetitive semivariogram modeling in new study areas. Our findings demonstrate that WD-matched models replicate the location choices of original models with a high degree of similarity while ensuring that clean-slate locations remain proximate to those of original models, enhancing geographic equity in RWIS deployment. This validates the practicality of reusing developed semivariogram parameters for WD-matched highways, significantly reducing the need for new geostatistical analyses and enhancing the framework’s applicability and accessibility for RWIS deployment across diverse geographic regions.
- 1 March 2025
- Published by MDPI AG in Future Transportation
- Vol. 5 (1)
Open AccessOptimizing RWIS Locations with Wasserstein Distance and Geostatistics: A Case Study in South Korea
Road Weather Information Systems (RWISs) are essential components of modern Intelligent Transportation Systems (ITSs) deployed in cold regions to gather real-time data on winter weather and road surface conditions. Despite their benefits, the high cost associated with RWIS installations demands optimized placement strategies to maximize their utility and cost-effectiveness. Geostatistics-based RWIS location-allocation methods, particularly those involving semivariogram modeling to quantify underlying spatial characteristics, have gained international recognition. However, new locations require unique semivariogram models, a process that is time-consuming and constrained by the availability of comprehensive datasets, often rendering location analysis challenging or infeasible. Addressing these limitations, this study introduces an innovative approach using Wasserstein Distance (WD) to link semivariograms across different datasets. This method streamlines optimization by eliminating the need for repetitive semivariogram modeling in new study areas. Our findings demonstrate that WD-matched models replicate the location choices of original models with a high degree of similarity while ensuring that clean-slate locations remain proximate to those of original models, enhancing geographic equity in RWIS deployment. This validates the practicality of reusing developed semivariogram parameters for WD-matched highways, significantly reducing the need for new geostatistical analyses and enhancing the framework’s applicability and accessibility for RWIS deployment across diverse geographic regions.
- Open Access
- 12 February 2025
- Published by MDPI AG in Preprints.org
Travel Time Information on Signalized Arterials
Travel time information has become an essential component of everyday commuting. Without such information, schedule delays would increase, leading to inevitable losses in traveler utility. In Korea, dedicated short-range communication transponders that identify vehicles have been installed along signalized arterials to collect travel time data. By matching vehicle identifications at consecutive points, travel times can be measured. However, for travel time information to be effective, two types of data processing techniques are required: outlier filtering and travel time prediction. This study proposes algorithms to address both challenges. An outlier filtering algorithm based on the median-based confidence interval was developed, taking into account the travel time characteristics on suburban arterials with frequent entry and exit points. Additionally, a travel time prediction algorithm that integrates Long Short-Term Memory (LSTM) and Convolutional Neural Network (CNN), referred to as LSTM-CNN, was developed to capture both long-term trends and local patterns in travel time data. The implementation of these algorithms resulted in a 2.2% reduction in error rates under congested conditions compared to current practices. At the 4-kilometer study site, the annual benefits from this error reduction could amount to $135,200.
- 12 February 2025
- Published by MDPI AG in Preprints.org
Open AccessTravel Time Information on Signalized Arterials
Travel time information has become an essential component of everyday commuting. Without such information, schedule delays would increase, leading to inevitable losses in traveler utility. In Korea, dedicated short-range communication transponders that identify vehicles have been installed along signalized arterials to collect travel time data. By matching vehicle identifications at consecutive points, travel times can be measured. However, for travel time information to be effective, two types of data processing techniques are required: outlier filtering and travel time prediction. This study proposes algorithms to address both challenges. An outlier filtering algorithm based on the median-based confidence interval was developed, taking into account the travel time characteristics on suburban arterials with frequent entry and exit points. Additionally, a travel time prediction algorithm that integrates Long Short-Term Memory (LSTM) and Convolutional Neural Network (CNN), referred to as LSTM-CNN, was developed to capture both long-term trends and local patterns in travel time data. The implementation of these algorithms resulted in a 2.2% reduction in error rates under congested conditions compared to current practices. At the 4-kilometer study site, the annual benefits from this error reduction could amount to $135,200.
- Open Access
- 17 December 2024
- Published by Elsevier in Transportation Research Interdisciplinary Perspectives
- Vol. 29
Predicting nighttime black ice using atmospheric data for efficient winter road maintenance patrols
- 17 December 2024
- Published by Elsevier in Transportation Research Interdisciplinary Perspectives
- Vol. 29
Open AccessPredicting nighttime black ice using atmospheric data for efficient winter road maintenance patrols
- Open Access
- 15 November 2024
- Published by Bentham Science Publishers Ltd. in The Open Transportation Journal
- Vol. 18 (1)
Travel Time Prediction Using Machine Learning Algorithms: Focusing on k-NN, LSTM, and Transformer
Background: Travel time prediction is fundamentally important in advanced traveler information systems. Particularly, in the Dedicated Short-Range Communications (DSRC) traffic information system targeted by this study, travel times are obtained after vehicles terminate trips on the route, indicating that a time-lag phenomenon is inevitable. Therefore, travel time prediction is even more emphasized in DSRC systems. With advances in artificial intelligence technologies, many new machine learning algorithms have recently been introduced. Methods: In this study, three machine learning algorithms—k-Nearest Neighbor (k-NN), Long Short-Term Memory (LSTM), and Transformer—were applied to predict travel times collected on a DSRC-equipped national highway. Travel time outliers caused by entry and exit maneuvers at intersections were filtered using a robust outlier treatment algorithm. Results: From applying the prediction algorithms, errors of 12.4%, 11.7%, and 10.3% were revealed for k-NN, LSTM, and Transformer, respectively. To identify the statistical significance of the differences in prediction performance, paired t-tests were performed. Consequently, all the differences between the algorithms were proven to be statistically significant. The differences in performance were found to be generally more significant under congested conditions compared to uncongested conditions. Conclusion: The enhanced reliability of real-time travel time information can increase the effectiveness of DSRC traffic information systems by encouraging drivers to divert from congested routes or adjust their trip schedules to less congested periods.
- 15 November 2024
- Published by Bentham Science Publishers Ltd. in The Open Transportation Journal
- Vol. 18 (1)
Open AccessTravel Time Prediction Using Machine Learning Algorithms: Focusing on k-NN, LSTM, and Transformer
Background: Travel time prediction is fundamentally important in advanced traveler information systems. Particularly, in the Dedicated Short-Range Communications (DSRC) traffic information system targeted by this study, travel times are obtained after vehicles terminate trips on the route, indicating that a time-lag phenomenon is inevitable. Therefore, travel time prediction is even more emphasized in DSRC systems. With advances in artificial intelligence technologies, many new machine learning algorithms have recently been introduced. Methods: In this study, three machine learning algorithms—k-Nearest Neighbor (k-NN), Long Short-Term Memory (LSTM), and Transformer—were applied to predict travel times collected on a DSRC-equipped national highway. Travel time outliers caused by entry and exit maneuvers at intersections were filtered using a robust outlier treatment algorithm. Results: From applying the prediction algorithms, errors of 12.4%, 11.7%, and 10.3% were revealed for k-NN, LSTM, and Transformer, respectively. To identify the statistical significance of the differences in prediction performance, paired t-tests were performed. Consequently, all the differences between the algorithms were proven to be statistically significant. The differences in performance were found to be generally more significant under congested conditions compared to uncongested conditions. Conclusion: The enhanced reliability of real-time travel time information can increase the effectiveness of DSRC traffic information systems by encouraging drivers to divert from congested routes or adjust their trip schedules to less congested periods.
- Open Access
- 28 August 2024
- Published by Bentham Science Publishers Ltd. in The Open Transportation Journal
- Vol. 18 (1)
Forecasting Nighttime Frost-induced Black Ice Formation on Bridges Using Atmospheric Data
Background: Faced with the tragedies caused by black ice in winter, especially on bridges, it is imperative to forecast black ice for preventive maintenance and to notify drivers approaching the dangerous spots. Methods: In this study, three different machine learning algorithms—Deep Neural Network (DNN), Random Forest (RF), and Support Vector Machine (SVM)—were employed to predict nighttime black ice induced by frost on three bridges in Korea. Input data consisted of atmospheric data (air temperature, relative humidity, dew point, and differences between air temperature and relative humidity over two consecutive days) provided by the weather agency. Results: To assess the employed models, reference data were generated based on the physical principle that ice forms when the pavement temperature is lower than the dew point temperature and negative. The pavement temperature was obtained using an infrared surface temperature sensor mounted on a maintenance patrol vehicle. Consequently, DNN and RF showed higher performance with an accuracy of 95%, followed by SVM with an accuracy of 92.5%. Conclusion: Due to the use of easily obtainable atmospheric data, the findings of this study can be practically applied to preventive maintenance and driver information, thereby enhancing traffic safety.
- 28 August 2024
- Published by Bentham Science Publishers Ltd. in The Open Transportation Journal
- Vol. 18 (1)
Open AccessForecasting Nighttime Frost-induced Black Ice Formation on Bridges Using Atmospheric Data
Background: Faced with the tragedies caused by black ice in winter, especially on bridges, it is imperative to forecast black ice for preventive maintenance and to notify drivers approaching the dangerous spots. Methods: In this study, three different machine learning algorithms—Deep Neural Network (DNN), Random Forest (RF), and Support Vector Machine (SVM)—were employed to predict nighttime black ice induced by frost on three bridges in Korea. Input data consisted of atmospheric data (air temperature, relative humidity, dew point, and differences between air temperature and relative humidity over two consecutive days) provided by the weather agency. Results: To assess the employed models, reference data were generated based on the physical principle that ice forms when the pavement temperature is lower than the dew point temperature and negative. The pavement temperature was obtained using an infrared surface temperature sensor mounted on a maintenance patrol vehicle. Consequently, DNN and RF showed higher performance with an accuracy of 95%, followed by SVM with an accuracy of 92.5%. Conclusion: Due to the use of easily obtainable atmospheric data, the findings of this study can be practically applied to preventive maintenance and driver information, thereby enhancing traffic safety.
- Open Access
- 18 October 2023
- Published by Bentham Science Publishers Ltd. in The Open Transportation Journal
- Vol. 17 (1)
Bridge Frost Prediction Using K-nearest Neighbor Classifier
Background: Considering the frequent occurrence of accidents on icy bridges during winter nights, it would be advantageous to notify both road managers and drivers regarding the most perilous areas. This notification would allow road managers to address the icy conditions by applying de-icing substances, while drivers could be more adequately prepared for potential hazards. Methods: In this study, the focus was on investigating k-nearest neighbor algorithms to predict nighttime icing caused by frost on three distinct bridges located on the National Highways in Korea. The algorithms utilized atmospheric data as input, which was obtained from the weather agency's website through an open API. The input data included relative humidity, air temperature, and dew point temperature, as well as the disparities in air temperature and humidity between two consecutive days. Results: In order to assess the effectiveness of the prediction models, reference data were created using the fundamental principle that ice is formed when the temperature of the pavement is below freezing and lower than the dew point temperature. Consequently, the developed algorithm demonstrated favorable performance, achieving an accuracy of 95% when evaluated using a test dataset that occupies 30% of the entire data. Conclusion: Considering the increasing focus on preventive maintenance, these newly developed forecasting models can be employed proactively as a preventive measure against icing. This proactive approach will ultimately contribute to improving traffic safety on winter roads.
- 18 October 2023
- Published by Bentham Science Publishers Ltd. in The Open Transportation Journal
- Vol. 17 (1)
Open AccessBridge Frost Prediction Using K-nearest Neighbor Classifier
Background: Considering the frequent occurrence of accidents on icy bridges during winter nights, it would be advantageous to notify both road managers and drivers regarding the most perilous areas. This notification would allow road managers to address the icy conditions by applying de-icing substances, while drivers could be more adequately prepared for potential hazards. Methods: In this study, the focus was on investigating k-nearest neighbor algorithms to predict nighttime icing caused by frost on three distinct bridges located on the National Highways in Korea. The algorithms utilized atmospheric data as input, which was obtained from the weather agency's website through an open API. The input data included relative humidity, air temperature, and dew point temperature, as well as the disparities in air temperature and humidity between two consecutive days. Results: In order to assess the effectiveness of the prediction models, reference data were created using the fundamental principle that ice is formed when the temperature of the pavement is below freezing and lower than the dew point temperature. Consequently, the developed algorithm demonstrated favorable performance, achieving an accuracy of 95% when evaluated using a test dataset that occupies 30% of the entire data. Conclusion: Considering the increasing focus on preventive maintenance, these newly developed forecasting models can be employed proactively as a preventive measure against icing. This proactive approach will ultimately contribute to improving traffic safety on winter roads.
- Open Access
- 29 August 2023
- Published by Elsevier in Transportation Research Interdisciplinary Perspectives
- Vol. 21
Night icing prediction using vehicle-mounted pavement temperature sensor and atmospheric data
- 29 August 2023
- Published by Elsevier in Transportation Research Interdisciplinary Perspectives
- Vol. 21
Open AccessNight icing prediction using vehicle-mounted pavement temperature sensor and atmospheric data