Pub Date : 2025-02-27DOI: 10.1109/OJITS.2025.3544262
Rodolfo Valiente;Jiejun Xu;Alireza Esna Ashari
Autonomous vehicles rely on accurate traffic sign classification, which is typically achieved through supervised learning. However, the diversity and complexity of traffic signs make it impractical to rely solely on large labeled datasets. While abundant data exists for common signs such as stop and yield signs, less common signs often lack sufficient representation in existing datasets. Few-shot learning has been proposed as an alternative solution for such cases in which there is not enough training data, but its effectiveness decreases as the number of classes increases. To address these challenges, our research introduces an innovative adaptive hierarchical framework with contrastive aggregation (HF-CA). This framework strategically reduces class dimensionality and enriches the dataset with more examples per category through contrastive aggregation. We validated our approach using modified versions of the GTSRB and Mapillary datasets, demonstrating that our method consistently outperforms existing baselines. By simplifying the classification process, our solution enhances classification accuracy and provides a scalable approach for scenarios with numerous classes but limited labels.
{"title":"An Adaptive Hierarchical Framework With Contrastive Aggregation for Traffic Sign Classification","authors":"Rodolfo Valiente;Jiejun Xu;Alireza Esna Ashari","doi":"10.1109/OJITS.2025.3544262","DOIUrl":"https://doi.org/10.1109/OJITS.2025.3544262","url":null,"abstract":"Autonomous vehicles rely on accurate traffic sign classification, which is typically achieved through supervised learning. However, the diversity and complexity of traffic signs make it impractical to rely solely on large labeled datasets. While abundant data exists for common signs such as stop and yield signs, less common signs often lack sufficient representation in existing datasets. Few-shot learning has been proposed as an alternative solution for such cases in which there is not enough training data, but its effectiveness decreases as the number of classes increases. To address these challenges, our research introduces an innovative adaptive hierarchical framework with contrastive aggregation (HF-CA). This framework strategically reduces class dimensionality and enriches the dataset with more examples per category through contrastive aggregation. We validated our approach using modified versions of the GTSRB and Mapillary datasets, demonstrating that our method consistently outperforms existing baselines. By simplifying the classification process, our solution enhances classification accuracy and provides a scalable approach for scenarios with numerous classes but limited labels.","PeriodicalId":100631,"journal":{"name":"IEEE Open Journal of Intelligent Transportation Systems","volume":"6 ","pages":"230-243"},"PeriodicalIF":4.6,"publicationDate":"2025-02-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10907807","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143611809","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-02-27DOI: 10.1109/OJITS.2025.3546685
Stijn Harbers;Jens Kalkkuhl;Tom van der Sande
The current state-of-the-art vehicle egomotion state estimation systems are limited in their usage for advanced driver-assistance systems (ADAS), as the estimation accuracy is limited in driving scenarios where large amounts of wheel slip occur. Addressing the fundamental limitations of vehicle egomotion state estimation, this study investigates the usage of RADAR in vehicle egomotion state estimation through a tight-coupling between an Inertial Measurement Unit (IMU) and RADAR. The limitations of the state-of-the-art are caused by the usage of automotive grade sensors, which provide limited accuracy. RADAR is a type of sensor, which is already used extensively in ADAS, however, not yet in egomotion estimation. A reason for not using RADAR in this context is that it requires knowledge of the motion of the detected targets. In literature statistical methods are suggested to reject moving detections, but these are per definition not robust. This research, therefore, answers the question: How can RADAR be used in vehicle egomotion state estimation to improve performance and expand the capabilities of the system in a robust way? A new method is developed, which through IMU-RADAR tight-coupling, is able to reject moving detections. These stationary detections are then integrated into a Kalman filter to obtain the vehicle motion. This new method is compared to the state-of-the-art methods and the results are validated on a real data set of a vehicle driving in urban and highway setting. The findings demonstrate that the newly introduced method enhances the accuracy of vehicle egomotion state estimation.
{"title":"Vehicle Egomotion Estimation Through IMU-RADAR Tight-Coupling","authors":"Stijn Harbers;Jens Kalkkuhl;Tom van der Sande","doi":"10.1109/OJITS.2025.3546685","DOIUrl":"https://doi.org/10.1109/OJITS.2025.3546685","url":null,"abstract":"The current state-of-the-art vehicle egomotion state estimation systems are limited in their usage for advanced driver-assistance systems (ADAS), as the estimation accuracy is limited in driving scenarios where large amounts of wheel slip occur. Addressing the fundamental limitations of vehicle egomotion state estimation, this study investigates the usage of RADAR in vehicle egomotion state estimation through a tight-coupling between an Inertial Measurement Unit (IMU) and RADAR. The limitations of the state-of-the-art are caused by the usage of automotive grade sensors, which provide limited accuracy. RADAR is a type of sensor, which is already used extensively in ADAS, however, not yet in egomotion estimation. A reason for not using RADAR in this context is that it requires knowledge of the motion of the detected targets. In literature statistical methods are suggested to reject moving detections, but these are per definition not robust. This research, therefore, answers the question: How can RADAR be used in vehicle egomotion state estimation to improve performance and expand the capabilities of the system in a robust way? A new method is developed, which through IMU-RADAR tight-coupling, is able to reject moving detections. These stationary detections are then integrated into a Kalman filter to obtain the vehicle motion. This new method is compared to the state-of-the-art methods and the results are validated on a real data set of a vehicle driving in urban and highway setting. The findings demonstrate that the newly introduced method enhances the accuracy of vehicle egomotion state estimation.","PeriodicalId":100631,"journal":{"name":"IEEE Open Journal of Intelligent Transportation Systems","volume":"6 ","pages":"244-255"},"PeriodicalIF":4.6,"publicationDate":"2025-02-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10907934","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143611808","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-02-20DOI: 10.1109/OJITS.2025.3543831
Jingxiong Meng;Junfeng Zhao
As infrastructure equipment development matures, leveraging these assets to enhance automated vehicle perception becomes increasingly valuable for more accurate and broader 3D object detection. This paper proposes a straightforward and scalable framework to incorporate infrastructure and vehicle onboard sensors to perform 3D object detection on Bird’s Eye View(BEV) images. And a cross-attention based block is involved in utilizing the interacted information among the sensors for sensor information fusion. Our model gets validated on the online V2X-Sim dataset under two scenarios: the short-range case and the long-range case. Our model demonstrates superior accuracy and broader detection capabilities compared to the baseline model from the experiment results.
{"title":"VI-BEV: Vehicle-Infrastructure Collaborative Perception for 3-D Object Detection on Bird’s-Eye View","authors":"Jingxiong Meng;Junfeng Zhao","doi":"10.1109/OJITS.2025.3543831","DOIUrl":"https://doi.org/10.1109/OJITS.2025.3543831","url":null,"abstract":"As infrastructure equipment development matures, leveraging these assets to enhance automated vehicle perception becomes increasingly valuable for more accurate and broader 3D object detection. This paper proposes a straightforward and scalable framework to incorporate infrastructure and vehicle onboard sensors to perform 3D object detection on Bird’s Eye View(BEV) images. And a cross-attention based block is involved in utilizing the interacted information among the sensors for sensor information fusion. Our model gets validated on the online V2X-Sim dataset under two scenarios: the short-range case and the long-range case. Our model demonstrates superior accuracy and broader detection capabilities compared to the baseline model from the experiment results.","PeriodicalId":100631,"journal":{"name":"IEEE Open Journal of Intelligent Transportation Systems","volume":"6 ","pages":"256-265"},"PeriodicalIF":4.6,"publicationDate":"2025-02-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10896690","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143611896","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-02-20DOI: 10.1109/OJITS.2025.3544374
Joseba Gorospe;Shahriar Hasan;Arrate Alonso Gómez;Elisabeth Uhlemann
Cooperative Adaptive Cruise Control (CACC) utilizes Vehicle-to-Vehicle (V2V) communications and onboard sensors to facilitate cooperative maneuvering among a group of automated vehicles called a vehicle string. Such string formation of automated vehicles enables improved safety, fuel efficiency, traffic flow, and road capacity. A vehicle using CACC computes its acceleration through information obtained from its preceding vehicle and/or the Leading Vehicle (LV) of the string through V2V communications. However, wireless communication is susceptible to inevitable transient outages due to irregular packet losses, which has severe consequences on the safety and stability of a vehicle string. To address this problem, this paper proposes an enhancement to an existing CACC algorithm; the idea is that when a vehicle does not receive information from its intended sources, i.e., the LV and the predecessor, for a certain duration, it uses information from the closest available longitudinal neighbors to the intended sources to compute its desired acceleration. Furthermore, we also investigate the possibility of using such information for training Machine Learning (ML) models and making predictions on the desired accelerations of the intended sources. Rigorous simulation studies demonstrate that when information from alternative sources is utilized during transient outages, a significant improvement in terms of safety, string stability, and fuel efficiency can be observed compared to the existing CACC. Moreover, the proposed approach can handle transient outages without requiring changes in the CACC communication topology, increasing the number of transmitted messages, or degrading string performance, as proposed by many works in the literature.
{"title":"Toward Resilient CACC Systems for Automated Vehicles","authors":"Joseba Gorospe;Shahriar Hasan;Arrate Alonso Gómez;Elisabeth Uhlemann","doi":"10.1109/OJITS.2025.3544374","DOIUrl":"https://doi.org/10.1109/OJITS.2025.3544374","url":null,"abstract":"Cooperative Adaptive Cruise Control (CACC) utilizes Vehicle-to-Vehicle (V2V) communications and onboard sensors to facilitate cooperative maneuvering among a group of automated vehicles called a vehicle string. Such string formation of automated vehicles enables improved safety, fuel efficiency, traffic flow, and road capacity. A vehicle using CACC computes its acceleration through information obtained from its preceding vehicle and/or the Leading Vehicle (LV) of the string through V2V communications. However, wireless communication is susceptible to inevitable transient outages due to irregular packet losses, which has severe consequences on the safety and stability of a vehicle string. To address this problem, this paper proposes an enhancement to an existing CACC algorithm; the idea is that when a vehicle does not receive information from its intended sources, i.e., the LV and the predecessor, for a certain duration, it uses information from the closest available longitudinal neighbors to the intended sources to compute its desired acceleration. Furthermore, we also investigate the possibility of using such information for training Machine Learning (ML) models and making predictions on the desired accelerations of the intended sources. Rigorous simulation studies demonstrate that when information from alternative sources is utilized during transient outages, a significant improvement in terms of safety, string stability, and fuel efficiency can be observed compared to the existing CACC. Moreover, the proposed approach can handle transient outages without requiring changes in the CACC communication topology, increasing the number of transmitted messages, or degrading string performance, as proposed by many works in the literature.","PeriodicalId":100631,"journal":{"name":"IEEE Open Journal of Intelligent Transportation Systems","volume":"6 ","pages":"276-293"},"PeriodicalIF":4.6,"publicationDate":"2025-02-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10897819","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143654900","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-02-20DOI: 10.1109/OJITS.2025.3544301
Ekin Ugurel;Gaoang Wang
Congestion due to morning and evening traffic peaks has caused economic losses amounting to billions of dollars annually. Thus, accurately predicting the departure time of commuters is of interest to transportation planners, engineers, and elected officials alike. We develop a statistically-informed deep learning approach to improve commuter departure time prediction models. Specifically, we leverage elements of the proportional hazards model, a class of time-to-event prediction approaches, to augment vanilla deep neural network (DNN) architectures. The proposed approach also employs collaborative filtering to segment the commuter population into distinct behavioral classes, allowing tailored predictions for specific commuter profiles. We find that our class of survival analysis-enhanced DNNs outperforms conventional neural networks in predicting trip departure times, while also offering more interpretability through the hazard coefficients.
{"title":"Beat the Morning Rush: Survival Analysis-Informed DNNs With Collaborative Filtering to Predict Departure Times","authors":"Ekin Ugurel;Gaoang Wang","doi":"10.1109/OJITS.2025.3544301","DOIUrl":"https://doi.org/10.1109/OJITS.2025.3544301","url":null,"abstract":"Congestion due to morning and evening traffic peaks has caused economic losses amounting to billions of dollars annually. Thus, accurately predicting the departure time of commuters is of interest to transportation planners, engineers, and elected officials alike. We develop a statistically-informed deep learning approach to improve commuter departure time prediction models. Specifically, we leverage elements of the proportional hazards model, a class of time-to-event prediction approaches, to augment vanilla deep neural network (DNN) architectures. The proposed approach also employs collaborative filtering to segment the commuter population into distinct behavioral classes, allowing tailored predictions for specific commuter profiles. We find that our class of survival analysis-enhanced DNNs outperforms conventional neural networks in predicting trip departure times, while also offering more interpretability through the hazard coefficients.","PeriodicalId":100631,"journal":{"name":"IEEE Open Journal of Intelligent Transportation Systems","volume":"6 ","pages":"266-275"},"PeriodicalIF":4.6,"publicationDate":"2025-02-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10897825","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143637940","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Traditional radar perception often rely on point clouds derived from radar heatmap using CFAR filtering, which can result in the loss of valuable information, especially weaker signals crucial for accurate perception. To address this, we present a novel approach for representation learning directly from pre-CFAR heatmaps, specifically for place recognition using a high-resolution MIMO radar sensor. By avoiding CFAR filtering, our method preserves richer contextual data, capturing finer details essential for identifying and matching distinctive features across locations. Pre-CFAR heatmaps, however, contain inherent noise and clutter, complicating their application in radar perception tasks. To overcome this, we propose a self-supervised network that learns robust latent features from noisy heatmaps. The architecture consists of two identical U-Net encoders that extract features from the pair of radar scans, which are then processed by a transformer encoder to estimate ego-motion. Ground truth ego-motion trajectories guide the network training using a weighted mean-square error loss. The latent feature representations from the trained encoders are used to create a database of feature vectors for previously visited locations. During runtime, for place recognition and loop closure detection, cosine similarity is applied to query scan feature representation and the database to find the closest matches. We also introduce data augmentation techniques to handle limited training data, enhancing the model’s generalization capability. Our approach, tested on the publicly available Coloradar dataset and our own, outperforms existing methods, showing significant improvements in place recognition accuracy, particularly in noisy and cluttered environments.
{"title":"Representation Learning for Place Recognition Using MIMO Radar","authors":"Prashant Kumar Rai;Nataliya Strokina;Reza Ghabcheloo","doi":"10.1109/OJITS.2025.3543286","DOIUrl":"https://doi.org/10.1109/OJITS.2025.3543286","url":null,"abstract":"Traditional radar perception often rely on point clouds derived from radar heatmap using CFAR filtering, which can result in the loss of valuable information, especially weaker signals crucial for accurate perception. To address this, we present a novel approach for representation learning directly from pre-CFAR heatmaps, specifically for place recognition using a high-resolution MIMO radar sensor. By avoiding CFAR filtering, our method preserves richer contextual data, capturing finer details essential for identifying and matching distinctive features across locations. Pre-CFAR heatmaps, however, contain inherent noise and clutter, complicating their application in radar perception tasks. To overcome this, we propose a self-supervised network that learns robust latent features from noisy heatmaps. The architecture consists of two identical U-Net encoders that extract features from the pair of radar scans, which are then processed by a transformer encoder to estimate ego-motion. Ground truth ego-motion trajectories guide the network training using a weighted mean-square error loss. The latent feature representations from the trained encoders are used to create a database of feature vectors for previously visited locations. During runtime, for place recognition and loop closure detection, cosine similarity is applied to query scan feature representation and the database to find the closest matches. We also introduce data augmentation techniques to handle limited training data, enhancing the model’s generalization capability. Our approach, tested on the publicly available Coloradar dataset and our own, outperforms existing methods, showing significant improvements in place recognition accuracy, particularly in noisy and cluttered environments.","PeriodicalId":100631,"journal":{"name":"IEEE Open Journal of Intelligent Transportation Systems","volume":"6 ","pages":"144-154"},"PeriodicalIF":4.6,"publicationDate":"2025-02-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10891700","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143535499","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-02-17DOI: 10.1109/OJITS.2025.3542193
Wouter Scholte;Tom van der Sande;Peter Zegelaar;Henk Nijmeijer
Road safety and traffic congestion are amongst the main challenges in current transportation systems. Literature shows that these challenges can be tackled using cooperative platoons, which is a technique in which vehicles use communication to drive closely behind each other in a string. An important topic of research regarding cooperative platoons is merging vehicles into a platoon at highway on-ramps. In previous work, the authors proposed a control strategy for the merging of a single cooperative automated vehicle into a platoon of vehicles at highway on-ramps. In this paper, the performance of this controller is demonstrated using experiments with two full-scale vehicles. During these experiments a two-vehicle platoon is formed in a limited distance. Several scenarios with different trajectories of the preceding vehicle are investigated. The preceding vehicle can be accelerating or decelerating simulating disturbances encountered in a larger platoon. In all experiments, the maneuver is successfully executed. The results of these experiments show the large potential of the proposed merging controller.
{"title":"Experimental Demonstration of Platoon Formation Using a Cooperative Merging Controller","authors":"Wouter Scholte;Tom van der Sande;Peter Zegelaar;Henk Nijmeijer","doi":"10.1109/OJITS.2025.3542193","DOIUrl":"https://doi.org/10.1109/OJITS.2025.3542193","url":null,"abstract":"Road safety and traffic congestion are amongst the main challenges in current transportation systems. Literature shows that these challenges can be tackled using cooperative platoons, which is a technique in which vehicles use communication to drive closely behind each other in a string. An important topic of research regarding cooperative platoons is merging vehicles into a platoon at highway on-ramps. In previous work, the authors proposed a control strategy for the merging of a single cooperative automated vehicle into a platoon of vehicles at highway on-ramps. In this paper, the performance of this controller is demonstrated using experiments with two full-scale vehicles. During these experiments a two-vehicle platoon is formed in a limited distance. Several scenarios with different trajectories of the preceding vehicle are investigated. The preceding vehicle can be accelerating or decelerating simulating disturbances encountered in a larger platoon. In all experiments, the maneuver is successfully executed. The results of these experiments show the large potential of the proposed merging controller.","PeriodicalId":100631,"journal":{"name":"IEEE Open Journal of Intelligent Transportation Systems","volume":"6 ","pages":"131-143"},"PeriodicalIF":4.6,"publicationDate":"2025-02-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10891586","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143535413","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-02-05DOI: 10.1109/OJITS.2025.3539419
Nikos Kougiatsos;Evelien L. Scheffers;Marcel C. van Benten;Dingena L. Schott;Peter de Vos;Rudy R. Negenborn;Vasso Reppa
Waterborne transport is very important for moving freight and passengers globally. To make this transport more efficient, vessel design must adapt to changing missions, regulations and the occurrence of malfunctions. This paper presents the design of an intelligent decision-support framework to assist marine engineers and vessel operators in updating the system and control architecture of marine vessels before and during a mission. The connection between the system architecture and control design perspectives is enabled using a semantics-based technique. To this end, the multi-level vessel control system is described by a semantic database, a knowledge graph used to connect the components automatically, and quantitative service criteria. Considering the system architecture, the optimal modification is deduced using modularity and complexity criteria, originating from the field of network theory. On the control side, an intelligent automation supervisor is designed to make offline and online decisions regarding the energy deficit to execute a new mission and the active automation configuration during operation. For offline decisions, system architecture modifications are requested by the vessel designers to cover the energy deficit. During operation, switching between hardware and virtual sensors as well as switching between energy management controllers is implemented to handle the effects of sensor faults. The framework is successfully applied to a case study of a tugboat used to adapt to missions with different power requirements, while simulation results are used to indicate its application in supporting the decisions of vessel designers and human vessel operators.
{"title":"An Intelligent Agent-Based Resilient Framework for Marine Vessel Mission Adaptations","authors":"Nikos Kougiatsos;Evelien L. Scheffers;Marcel C. van Benten;Dingena L. Schott;Peter de Vos;Rudy R. Negenborn;Vasso Reppa","doi":"10.1109/OJITS.2025.3539419","DOIUrl":"https://doi.org/10.1109/OJITS.2025.3539419","url":null,"abstract":"Waterborne transport is very important for moving freight and passengers globally. To make this transport more efficient, vessel design must adapt to changing missions, regulations and the occurrence of malfunctions. This paper presents the design of an intelligent decision-support framework to assist marine engineers and vessel operators in updating the system and control architecture of marine vessels before and during a mission. The connection between the system architecture and control design perspectives is enabled using a semantics-based technique. To this end, the multi-level vessel control system is described by a semantic database, a knowledge graph used to connect the components automatically, and quantitative service criteria. Considering the system architecture, the optimal modification is deduced using modularity and complexity criteria, originating from the field of network theory. On the control side, an intelligent automation supervisor is designed to make offline and online decisions regarding the energy deficit to execute a new mission and the active automation configuration during operation. For offline decisions, system architecture modifications are requested by the vessel designers to cover the energy deficit. During operation, switching between hardware and virtual sensors as well as switching between energy management controllers is implemented to handle the effects of sensor faults. The framework is successfully applied to a case study of a tugboat used to adapt to missions with different power requirements, while simulation results are used to indicate its application in supporting the decisions of vessel designers and human vessel operators.","PeriodicalId":100631,"journal":{"name":"IEEE Open Journal of Intelligent Transportation Systems","volume":"6 ","pages":"184-203"},"PeriodicalIF":4.6,"publicationDate":"2025-02-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10876184","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143564008","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-02-03DOI: 10.1109/OJITS.2025.3538037
Ashutosh Holla B.;Manohara M. M. Pai;Ujjwal Verma;Radhika M. Pai
Vehicle re-identification and tracking play a vital role in intelligent transportation systems as they enhance traffic management, improve safety, and optimize flow by precisely monitoring and analyzing vehicle movements across various locations. This technology enables the collecting of data in real-time, which allows for effective identification of incidents, enforcement of laws, and decision-making in urban planning. Deep learning techniques used in vehicle re-identification extract distinct characteristics to identify and match a vehicle across different camera perspectives. This bridges the non-overlapping field of camera views and forms a relationship between the detected vehicles. Tracking enhances this process by assigning a distinct identifier to the recognized vehicle, allowing for the creation of a continuous trajectory across the network for further analysis. Vehicle re-identification and tracking have made substantial progress in recent years as a result of the accelerated development of deep learning. Consequently, it is imperative to conduct a thorough examination of these chores. To provide a detailed picture of the research towards vehicle re-identification and tracking, this study provides the recent advancements of various datasets, and frameworks and strategies undertaken to perform these tasks. Specifically, the paper provides a comprehensive review of the different modes of re-identification of vehicles and further analysis. The paper also discusses the challenges and directions that can be taken in future for vehicle re-identification and tracking.
{"title":"Vehicle Re-Identification and Tracking: Algorithmic Approach, Challenges and Future Directions","authors":"Ashutosh Holla B.;Manohara M. M. Pai;Ujjwal Verma;Radhika M. Pai","doi":"10.1109/OJITS.2025.3538037","DOIUrl":"https://doi.org/10.1109/OJITS.2025.3538037","url":null,"abstract":"Vehicle re-identification and tracking play a vital role in intelligent transportation systems as they enhance traffic management, improve safety, and optimize flow by precisely monitoring and analyzing vehicle movements across various locations. This technology enables the collecting of data in real-time, which allows for effective identification of incidents, enforcement of laws, and decision-making in urban planning. Deep learning techniques used in vehicle re-identification extract distinct characteristics to identify and match a vehicle across different camera perspectives. This bridges the non-overlapping field of camera views and forms a relationship between the detected vehicles. Tracking enhances this process by assigning a distinct identifier to the recognized vehicle, allowing for the creation of a continuous trajectory across the network for further analysis. Vehicle re-identification and tracking have made substantial progress in recent years as a result of the accelerated development of deep learning. Consequently, it is imperative to conduct a thorough examination of these chores. To provide a detailed picture of the research towards vehicle re-identification and tracking, this study provides the recent advancements of various datasets, and frameworks and strategies undertaken to perform these tasks. Specifically, the paper provides a comprehensive review of the different modes of re-identification of vehicles and further analysis. The paper also discusses the challenges and directions that can be taken in future for vehicle re-identification and tracking.","PeriodicalId":100631,"journal":{"name":"IEEE Open Journal of Intelligent Transportation Systems","volume":"6 ","pages":"155-183"},"PeriodicalIF":4.6,"publicationDate":"2025-02-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10870125","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143564006","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-01-31DOI: 10.1109/OJITS.2025.3532796
Kleona Binjaku;C. Pasquale;E. K. Meçe;S. Sacone
Effective traffic management and control are essential for mitigating congestion and minimizing environmental impacts on road transportation systems. In this paper, we propose a novel approach for traffic modeling that integrates physics-based dynamics with machine learning techniques. Our method leverages Gaussian Processes (GPs) and a multi-class second-order discrete traffic model known as METANET to develop a Physics-Regularized Machine Learning framework. Furthermore, the proposed approach includes for the first time multi-class on/off ramps within the modeling framework, enhancing the realism of the predictive model. We systematically evaluate the performance of the hybrid model across varying dataset sizes to determine optimal data requirements for accurate traffic predictions. Experimental results indicate the improved predictive performance of the proposed approach compared to traditional machine learning and physics-based models. Our findings underscore the potential of Physics-Regularized Machine Learning for enhancing traffic management and control strategies in real-world scenarios.
{"title":"Freeway Traffic Modeling by Physics-Regularized Gaussian Processes","authors":"Kleona Binjaku;C. Pasquale;E. K. Meçe;S. Sacone","doi":"10.1109/OJITS.2025.3532796","DOIUrl":"https://doi.org/10.1109/OJITS.2025.3532796","url":null,"abstract":"Effective traffic management and control are essential for mitigating congestion and minimizing environmental impacts on road transportation systems. In this paper, we propose a novel approach for traffic modeling that integrates physics-based dynamics with machine learning techniques. Our method leverages Gaussian Processes (GPs) and a multi-class second-order discrete traffic model known as METANET to develop a Physics-Regularized Machine Learning framework. Furthermore, the proposed approach includes for the first time multi-class on/off ramps within the modeling framework, enhancing the realism of the predictive model. We systematically evaluate the performance of the hybrid model across varying dataset sizes to determine optimal data requirements for accurate traffic predictions. Experimental results indicate the improved predictive performance of the proposed approach compared to traditional machine learning and physics-based models. Our findings underscore the potential of Physics-Regularized Machine Learning for enhancing traffic management and control strategies in real-world scenarios.","PeriodicalId":100631,"journal":{"name":"IEEE Open Journal of Intelligent Transportation Systems","volume":"6 ","pages":"116-130"},"PeriodicalIF":4.6,"publicationDate":"2025-01-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10859260","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143512935","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：