Pub Date : 2025-06-18DOI: 10.1109/OJITS.2025.3580802
Annarita Carianni;Andrea Gemma
Forecasting traffic conditions is critical for modern mobility management. With urbanization and motorization rates rising globally, accurate traffic flow prediction plays a vital role in mitigating congestion, optimizing traffic strategies, and reducing environmental impacts. This paper provides a comprehensive review of traffic forecasting methods, bridging traditional techniques and innovative approaches driven by computational intelligence and abundant data. The study classifies forecasting methods into four categories: naïve techniques, parametric methods, simulation-based approaches, and nonparametric models such as machine learning and deep learning. Each category is analyzed for its historical development, theoretical foundations, and practical applications, with special emphasis on artificial intelligence’s transformative role in enabling dynamic and accurate predictions. The review evaluates traditional models like ARIMA and Kalman filters, alongside nonparametric techniques such as neural networks, and explores hybrid approaches that integrate multiple forecasting methods. It also assesses the complementary role of traffic simulation, from macroscopic to microscopic scales, in capturing complex traffic dynamics. The methodology synthesizes insights from foundational works and recent influential studies, examining metrics for prediction accuracy and identifying contextual factors shaping method effectiveness. The paper highlights strengths, limitations, and opportunities for advancement across forecasting approaches. Concluding with a forward-looking perspective, the review underscores trends such as spatiotemporal modeling and real-time data integration, which promise smarter, more adaptive traffic management solutions. This survey serves as a valuable resource for researchers, policymakers, and practitioners in navigating the evolving field of traffic flow forecasting.
{"title":"Overview of Traffic Flow Forecasting Techniques","authors":"Annarita Carianni;Andrea Gemma","doi":"10.1109/OJITS.2025.3580802","DOIUrl":"https://doi.org/10.1109/OJITS.2025.3580802","url":null,"abstract":"Forecasting traffic conditions is critical for modern mobility management. With urbanization and motorization rates rising globally, accurate traffic flow prediction plays a vital role in mitigating congestion, optimizing traffic strategies, and reducing environmental impacts. This paper provides a comprehensive review of traffic forecasting methods, bridging traditional techniques and innovative approaches driven by computational intelligence and abundant data. The study classifies forecasting methods into four categories: naïve techniques, parametric methods, simulation-based approaches, and nonparametric models such as machine learning and deep learning. Each category is analyzed for its historical development, theoretical foundations, and practical applications, with special emphasis on artificial intelligence’s transformative role in enabling dynamic and accurate predictions. The review evaluates traditional models like ARIMA and Kalman filters, alongside nonparametric techniques such as neural networks, and explores hybrid approaches that integrate multiple forecasting methods. It also assesses the complementary role of traffic simulation, from macroscopic to microscopic scales, in capturing complex traffic dynamics. The methodology synthesizes insights from foundational works and recent influential studies, examining metrics for prediction accuracy and identifying contextual factors shaping method effectiveness. The paper highlights strengths, limitations, and opportunities for advancement across forecasting approaches. Concluding with a forward-looking perspective, the review underscores trends such as spatiotemporal modeling and real-time data integration, which promise smarter, more adaptive traffic management solutions. This survey serves as a valuable resource for researchers, policymakers, and practitioners in navigating the evolving field of traffic flow forecasting.","PeriodicalId":100631,"journal":{"name":"IEEE Open Journal of Intelligent Transportation Systems","volume":"6 ","pages":"848-882"},"PeriodicalIF":4.6,"publicationDate":"2025-06-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11042911","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144623889","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}
Platooning or cooperative adaptive cruise control (CACC) has been investigated for decades, but debate about its lasting impact is still ongoing. While the benefits of platooning and the formation of platoons are well understood for trucks, they are less clear for passenger cars, which have a higher heterogeneity in trips and drivers’ preferences. Most importantly, it remains unclear how to form platoons of passenger cars in order to optimize the personal benefit for the individual driver. To this end, in this paper, we propose a novel platoon formation algorithm that optimizes the personal benefit for drivers of individual passenger cars. For computing vehicle-to-platoon assignments, the algorithm utilizes a new metric that we propose to evaluate the personal benefits of various driving systems, including platooning. By combining fuel and travel time costs into a single monetary value, drivers can estimate overall trip costs according to a personal monetary value for time spent. This provides an intuitive way for drivers to understand and compare the benefits of driving systems like human driving, adaptive cruise control (ACC), and, of course, platooning. Unlike previous similarity-based methods, our proposed algorithm forms platoons only when beneficial for the driver, rather than solely for platooning. We demonstrate the new metric for the total trip cost in a numerical analysis and explain its interpretation. Results of a large-scale simulation study demonstrate that our proposed platoon formation algorithm outperforms normal ACC as well as previous similarity-based platooning approaches by balancing fuel savings and travel time, independent of traffic and drivers’ time cost.
{"title":"Incentive-Based Platoon Formation: Optimizing the Personal Benefit for Drivers","authors":"Julian Heinovski;Doğanalp Ergenç;Kirsten Thommes;Falko Dressler","doi":"10.1109/OJITS.2025.3580464","DOIUrl":"https://doi.org/10.1109/OJITS.2025.3580464","url":null,"abstract":"Platooning or cooperative adaptive cruise control (CACC) has been investigated for decades, but debate about its lasting impact is still ongoing. While the benefits of platooning and the formation of platoons are well understood for trucks, they are less clear for passenger cars, which have a higher heterogeneity in trips and drivers’ preferences. Most importantly, it remains unclear how to form platoons of passenger cars in order to optimize the personal benefit for the individual driver. To this end, in this paper, we propose a novel platoon formation algorithm that optimizes the personal benefit for drivers of individual passenger cars. For computing vehicle-to-platoon assignments, the algorithm utilizes a new metric that we propose to evaluate the personal benefits of various driving systems, including platooning. By combining fuel and travel time costs into a single monetary value, drivers can estimate overall trip costs according to a personal monetary value for time spent. This provides an intuitive way for drivers to understand and compare the benefits of driving systems like human driving, adaptive cruise control (ACC), and, of course, platooning. Unlike previous similarity-based methods, our proposed algorithm forms platoons only when beneficial for the driver, rather than solely for platooning. We demonstrate the new metric for the total trip cost in a numerical analysis and explain its interpretation. Results of a large-scale simulation study demonstrate that our proposed platoon formation algorithm outperforms normal ACC as well as previous similarity-based platooning approaches by balancing fuel savings and travel time, independent of traffic and drivers’ time cost.","PeriodicalId":100631,"journal":{"name":"IEEE Open Journal of Intelligent Transportation Systems","volume":"6 ","pages":"813-831"},"PeriodicalIF":4.6,"publicationDate":"2025-06-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11039020","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144597691","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}
This paper presents a bio-inspired, generative Multi-Modal Sensor Fusion (MSF) framework to effectively detecting novel and dynamic situations in the surroundings of Autonomous Vehicle (AV). The MSF framework fuses both proprioceptive (wheel odometry) and exteroceptive (LiDAR point-clouds) sensory inputs. A novel 3-Dimensional Dynamic Variational Auto-Encoder (3D-DVAE) model is employed to learn attention-focused distributions from point-clouds in an unsupervised manner. By fusing the distributions of both modalities (wheel and lidar), modality-specific experts’ distributions are learned, capturing both proprioceptive and exteroceptive information from the surroundings. Bayesian Filtering is then applied to detect novel situations/dynamics by probabilistically inferring future states. The proposed method is validated using the KITTI dataset across diverse and complex urban environments. Both quantitative and qualitative results demonstrate the effectiveness of the proposed approach in detecting novelties through multi-modal fusion.
{"title":"Novelty Detection in Autonomous Driving: A Generative Multi-Modal Sensor Fusion Approach","authors":"Hafsa Iqbal;Haleema Sadia;Abdulla Al-Kaff;Fernando Garcié","doi":"10.1109/OJITS.2025.3580271","DOIUrl":"https://doi.org/10.1109/OJITS.2025.3580271","url":null,"abstract":"This paper presents a bio-inspired, generative Multi-Modal Sensor Fusion (MSF) framework to effectively detecting novel and dynamic situations in the surroundings of Autonomous Vehicle (AV). The MSF framework fuses both proprioceptive (wheel odometry) and exteroceptive (LiDAR point-clouds) sensory inputs. A novel 3-Dimensional Dynamic Variational Auto-Encoder (3D-DVAE) model is employed to learn attention-focused distributions from point-clouds in an unsupervised manner. By fusing the distributions of both modalities (wheel and lidar), modality-specific experts’ distributions are learned, capturing both proprioceptive and exteroceptive information from the surroundings. Bayesian Filtering is then applied to detect novel situations/dynamics by probabilistically inferring future states. The proposed method is validated using the KITTI dataset across diverse and complex urban environments. Both quantitative and qualitative results demonstrate the effectiveness of the proposed approach in detecting novelties through multi-modal fusion.","PeriodicalId":100631,"journal":{"name":"IEEE Open Journal of Intelligent Transportation Systems","volume":"6 ","pages":"799-812"},"PeriodicalIF":4.6,"publicationDate":"2025-06-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11037519","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144557713","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 : 2025-06-13DOI: 10.1109/OJITS.2025.3579653
Guido Perboli;Antonio Tota;Filippo Velardocchia
Heavy vehicle rollover plays a pivotal role in road safety scenarios. Numerous researchers addressed the topic, with particular focus on drivers related injuries. Considering the same and other connected implications, the necessity for techniques able to estimate and predict overturning eventualities appears evident. Different methodologies were explored, with notable achievements obtained by neural network-based algorithms. At the same time, their heavy requirements in terms of data needs to be addressed to allow practical applications in terms of time and costs. Consequently, exploring the interaction between simulation and experimental data becomes extremely important, motivating the methodology proposed by this paper. In details, an heavy vehicle model was designed in IPG Carmaker®, while experimental data on its physical alter ego were acquired. This led to the generation of a synthetic dataset and the collection of an empirical one. Both were used to define a Long Short-Term Memory architecture, with a dual purpose. First, as typical rollover indicator, estimate the vehicle roll angle. Second, compare the performance of the neural networks, aiming to obtain at least the same order of magnitude in terms of RMSE, MSE and MAE. The goal was to demonstrate that synthetic data can not only be used in combination with real data, but also as substitutes able to address time and cost constraints inevitably linked to the latter, allowing more efficient experiments for overtipping prevention.
{"title":"Long-Short Term Memory Networks and Synthetic Data for Heavy Vehicle Rollover Prevention","authors":"Guido Perboli;Antonio Tota;Filippo Velardocchia","doi":"10.1109/OJITS.2025.3579653","DOIUrl":"https://doi.org/10.1109/OJITS.2025.3579653","url":null,"abstract":"Heavy vehicle rollover plays a pivotal role in road safety scenarios. Numerous researchers addressed the topic, with particular focus on drivers related injuries. Considering the same and other connected implications, the necessity for techniques able to estimate and predict overturning eventualities appears evident. Different methodologies were explored, with notable achievements obtained by neural network-based algorithms. At the same time, their heavy requirements in terms of data needs to be addressed to allow practical applications in terms of time and costs. Consequently, exploring the interaction between simulation and experimental data becomes extremely important, motivating the methodology proposed by this paper. In details, an heavy vehicle model was designed in IPG Carmaker®, while experimental data on its physical alter ego were acquired. This led to the generation of a synthetic dataset and the collection of an empirical one. Both were used to define a Long Short-Term Memory architecture, with a dual purpose. First, as typical rollover indicator, estimate the vehicle roll angle. Second, compare the performance of the neural networks, aiming to obtain at least the same order of magnitude in terms of RMSE, MSE and MAE. The goal was to demonstrate that synthetic data can not only be used in combination with real data, but also as substitutes able to address time and cost constraints inevitably linked to the latter, allowing more efficient experiments for overtipping prevention.","PeriodicalId":100631,"journal":{"name":"IEEE Open Journal of Intelligent Transportation Systems","volume":"6 ","pages":"792-798"},"PeriodicalIF":4.6,"publicationDate":"2025-06-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11036550","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144536565","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 : 2025-06-11DOI: 10.1109/OJITS.2025.3578872
Mariam Nour;Mohamed H. Zaki;Mohamed Abdel-Aty
Connected and automated vehicle (CAV) technology has the potential to enhance lane change safety in work zones, especially during lane closures. However, the safety implications of vehicle-to-vehicle (V2V) communication under realistic operating conditions remain insufficiently understood. This study investigates the impact of V2V communication on lane change safety in work zone scenarios using a calibrated co-simulation framework that integrates both traffic and communication networks. The framework simulates a range of realistic conditions—including varying market penetration rates (MPRs), communication ranges, and merge strategies (early and late)—and evaluates lane change safety using the time-to-collision (TTC) metric. A data dissemination algorithm is incorporated to coordinate V2V messaging and enable CAVs to initiate safe lane changes. Unlike prior studies that assume ideal communication conditions, this work simulates realistic V2V communication by incorporating metrics such as packet loss and packet delivery ratio to examine their impact on lane change safety. Findings indicate that higher MPRs and extended communication ranges generally enhance safety; however, limitations in communication quality can significantly reduce these benefits—particularly in late merge scenarios, where degraded data exchange decreases safety. Sensitivity analyses further reveal that lane-change timing and communication range are critical factors influencing safety outcomes, emphasizing the need to account for communication reliability when designing and evaluating CAV-based safety interventions.
{"title":"Assessing the Impact of Vehicle-to-Vehicle Communication on Lane Change Safety in Work Zones","authors":"Mariam Nour;Mohamed H. Zaki;Mohamed Abdel-Aty","doi":"10.1109/OJITS.2025.3578872","DOIUrl":"https://doi.org/10.1109/OJITS.2025.3578872","url":null,"abstract":"Connected and automated vehicle (CAV) technology has the potential to enhance lane change safety in work zones, especially during lane closures. However, the safety implications of vehicle-to-vehicle (V2V) communication under realistic operating conditions remain insufficiently understood. This study investigates the impact of V2V communication on lane change safety in work zone scenarios using a calibrated co-simulation framework that integrates both traffic and communication networks. The framework simulates a range of realistic conditions—including varying market penetration rates (MPRs), communication ranges, and merge strategies (early and late)—and evaluates lane change safety using the time-to-collision (TTC) metric. A data dissemination algorithm is incorporated to coordinate V2V messaging and enable CAVs to initiate safe lane changes. Unlike prior studies that assume ideal communication conditions, this work simulates realistic V2V communication by incorporating metrics such as packet loss and packet delivery ratio to examine their impact on lane change safety. Findings indicate that higher MPRs and extended communication ranges generally enhance safety; however, limitations in communication quality can significantly reduce these benefits—particularly in late merge scenarios, where degraded data exchange decreases safety. Sensitivity analyses further reveal that lane-change timing and communication range are critical factors influencing safety outcomes, emphasizing the need to account for communication reliability when designing and evaluating CAV-based safety interventions.","PeriodicalId":100631,"journal":{"name":"IEEE Open Journal of Intelligent Transportation Systems","volume":"6 ","pages":"832-847"},"PeriodicalIF":4.6,"publicationDate":"2025-06-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11030855","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144623869","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 : 2025-06-02DOI: 10.1109/OJITS.2025.3575808
Yiqiao Li;Andre Y. C. Tok;Stephen G. Ritchie
Inductive loop sensors are widely deployed across the U.S. and can provide vehicle classification data with comparable accuracy to the current axle-based sensor systems when they are enhanced with the inductive signature technology and advanced machine learning models. However, the existing truck population is expected to turnover and be replaced with newer models that may generate distinct inductive signature characteristics. Consequently, legacy inductive signature-based models may not perform optimally in classifying newer trucks operating on the highways over time. To enhance the resilience of the signature-based classification system, this paper investigated a self-learning framework to address the classification system obsolescence through the integration of two complementary sensor technologies: Inductive loop sensors and Light Detection and Ranging (LiDAR) sensors. In this framework, the LiDAR-based Federal Highway Administration (FHWA) classification model served as a data labeling platform to generate class labels for validating and updating the legacy signature-based model. Next, an adaptive transfer learning framework was implemented to improve the performance of a legacy inductive signature-based classification model without compromising computation efficiency. This framework demonstrates the resilience enhancement of the inductive signature-based FHWA classification model with an intelligent system update to accommodate vehicle transition over time while retaining legacy knowledge of the pre-existing population using a methodology that significantly reduces the overall burden of periodic model calibration by utilizing the information stored in the legacy model. The experiment demonstrates that this adaptive self-learning framework achieves an overall correct classification rate of 0.89 on a dataset with distinctively different truck configurations.
{"title":"Adaptive Self-Learning Framework for Resilient Vehicle Classification Through the Integration of Inductive Loops and LiDAR Sensors","authors":"Yiqiao Li;Andre Y. C. Tok;Stephen G. Ritchie","doi":"10.1109/OJITS.2025.3575808","DOIUrl":"https://doi.org/10.1109/OJITS.2025.3575808","url":null,"abstract":"Inductive loop sensors are widely deployed across the U.S. and can provide vehicle classification data with comparable accuracy to the current axle-based sensor systems when they are enhanced with the inductive signature technology and advanced machine learning models. However, the existing truck population is expected to turnover and be replaced with newer models that may generate distinct inductive signature characteristics. Consequently, legacy inductive signature-based models may not perform optimally in classifying newer trucks operating on the highways over time. To enhance the resilience of the signature-based classification system, this paper investigated a self-learning framework to address the classification system obsolescence through the integration of two complementary sensor technologies: Inductive loop sensors and Light Detection and Ranging (LiDAR) sensors. In this framework, the LiDAR-based Federal Highway Administration (FHWA) classification model served as a data labeling platform to generate class labels for validating and updating the legacy signature-based model. Next, an adaptive transfer learning framework was implemented to improve the performance of a legacy inductive signature-based classification model without compromising computation efficiency. This framework demonstrates the resilience enhancement of the inductive signature-based FHWA classification model with an intelligent system update to accommodate vehicle transition over time while retaining legacy knowledge of the pre-existing population using a methodology that significantly reduces the overall burden of periodic model calibration by utilizing the information stored in the legacy model. The experiment demonstrates that this adaptive self-learning framework achieves an overall correct classification rate of 0.89 on a dataset with distinctively different truck configurations.","PeriodicalId":100631,"journal":{"name":"IEEE Open Journal of Intelligent Transportation Systems","volume":"6 ","pages":"768-780"},"PeriodicalIF":4.6,"publicationDate":"2025-06-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11021459","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144367055","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 : 2025-04-29DOI: 10.1109/OJITS.2025.3565209
Kassem Danach;Hassan Harb;Badih Baz;Abbass Nasser
Efficient logistics management is critical in the modern global supply chain, and this study introduces an advanced hyperheuristic approach to the Multi-Echelon Hub and Routing Optimization (MEHRO) problem. The MEHRO problem encompasses optimizing hub locations and vehicle routes while balancing cost efficiency, service quality, and environmental sustainability. A novel mathematical model integrates transportation, hub setup, and inventory costs, strengthened by valid inequalities to enhance computational efficiency. The hyperheuristic framework dynamically selects from a pool of low-level heuristics, adapting strategies to varying problem instances. A real-world case study validates the model’s effectiveness, demonstrating significant cost reductions, improved service levels, and minimized environmental impact compared to traditional methods. This work sets a foundation for scalable and adaptive solutions in logistics and combinatorial optimization, catering to the evolving demands of global supply chain management.
{"title":"A Hyperheuristic Approach to Multi-Echelon Hub and Routing Optimization: Model, Valid Inequalities, and Case Study","authors":"Kassem Danach;Hassan Harb;Badih Baz;Abbass Nasser","doi":"10.1109/OJITS.2025.3565209","DOIUrl":"https://doi.org/10.1109/OJITS.2025.3565209","url":null,"abstract":"Efficient logistics management is critical in the modern global supply chain, and this study introduces an advanced hyperheuristic approach to the Multi-Echelon Hub and Routing Optimization (MEHRO) problem. The MEHRO problem encompasses optimizing hub locations and vehicle routes while balancing cost efficiency, service quality, and environmental sustainability. A novel mathematical model integrates transportation, hub setup, and inventory costs, strengthened by valid inequalities to enhance computational efficiency. The hyperheuristic framework dynamically selects from a pool of low-level heuristics, adapting strategies to varying problem instances. A real-world case study validates the model’s effectiveness, demonstrating significant cost reductions, improved service levels, and minimized environmental impact compared to traditional methods. This work sets a foundation for scalable and adaptive solutions in logistics and combinatorial optimization, catering to the evolving demands of global supply chain management.","PeriodicalId":100631,"journal":{"name":"IEEE Open Journal of Intelligent Transportation Systems","volume":"6 ","pages":"781-791"},"PeriodicalIF":4.6,"publicationDate":"2025-04-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10979948","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144481830","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}
Adapting to dynamically changing situations remains a pivotal challenge for automated driving systems, which demand robust and efficient solutions. Occasional perception errors inherent in artificial intelligence further complicate the task. Whereas traditional motion planning algorithms address this challenge by replanning the entire trajectory, a significantly more efficient strategy is to repair only the flawed segments. Our paper introduces a groundbreaking approach by formulating an optimal trajectory repairing problem and proposing an innovative and efficient framework for critical timing detection and trajectory repairing. This trajectory repairing specifically employs Bernstein basis polynomials in both 2D distance-time and 3D spatiotemporal spaces. A distinctive feature of our method is the use of an anytime grid search to determine a sub-optimal time-to-repair, which contrasts with previous methods that relied on manually tuned or fixed repair times, limiting both flexibility and robustness. A statistical analysis of 100 scenarios demonstrates that our trajectory-repairing framework outperforms the path-speed decoupled repairing framework in terms of scenario success rate. Furthermore, we introduce a novel algorithm for driving corridor generation that more accurately approximates the collision-free space than state-of-the-art work. The proposed approach has broad potential for application in embedded systems across various autonomous platforms.
{"title":"Anytime Optimal Trajectory Repairing for Autonomous Vehicles","authors":"Kailin Tong;Martin Steinberger;Martin Horn;Selim Solmaz;Daniel Watzenig","doi":"10.1109/OJITS.2025.3563823","DOIUrl":"https://doi.org/10.1109/OJITS.2025.3563823","url":null,"abstract":"Adapting to dynamically changing situations remains a pivotal challenge for automated driving systems, which demand robust and efficient solutions. Occasional perception errors inherent in artificial intelligence further complicate the task. Whereas traditional motion planning algorithms address this challenge by replanning the entire trajectory, a significantly more efficient strategy is to repair only the flawed segments. Our paper introduces a groundbreaking approach by formulating an optimal trajectory repairing problem and proposing an innovative and efficient framework for critical timing detection and trajectory repairing. This trajectory repairing specifically employs Bernstein basis polynomials in both 2D distance-time and 3D spatiotemporal spaces. A distinctive feature of our method is the use of an anytime grid search to determine a sub-optimal time-to-repair, which contrasts with previous methods that relied on manually tuned or fixed repair times, limiting both flexibility and robustness. A statistical analysis of 100 scenarios demonstrates that our trajectory-repairing framework outperforms the path-speed decoupled repairing framework in terms of scenario success rate. Furthermore, we introduce a novel algorithm for driving corridor generation that more accurately approximates the collision-free space than state-of-the-art work. The proposed approach has broad potential for application in embedded systems across various autonomous platforms.","PeriodicalId":100631,"journal":{"name":"IEEE Open Journal of Intelligent Transportation Systems","volume":"6 ","pages":"537-553"},"PeriodicalIF":4.6,"publicationDate":"2025-04-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10979545","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143908427","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 : 2025-04-25DOI: 10.1109/OJITS.2025.3564361
Kristi Dulaj;Abdulraqeb Alhammadi;Ibraheem Shayea;Ayman A. El-Saleh;Mohammad Alnakhli
A revolutionary age in telecommunications is being ushered in by the confluence of machine learning (ML) with fifth-generation (5G) wireless communication technologies and beyond. This research investigates ML approaches in 5G networks for adaptive spectrum usage, quality of service (QoS) management, predictive maintenance, and network optimization. By leveraging ML algorithms, 5G networks can forecast user behavior, allocate resources optimally, and dynamically adjust to changing conditions, enhancing performance and dependability. Additionally, ML-driven methods improve cybersecurity in 5G settings. Furthermore, the integration of ML in 5G networks is pivotal for advancing intelligent transportation systems, enabling dynamic route optimization, adaptive traffic management, and enhanced vehicular communication. Intelligent networks will transform wireless communication by replacing traditional processing with end-to-end solutions, utilizing cognitive radio systems and deep reinforcement learning for optimized spectrum sharing and efficiency. Despite significant potential, challenges such as interoperability, security, scalability, and energy efficiency must be addressed. This paper discusses these challenges and highlights future trends beyond 5G, emphasizing ML's critical role in shaping the future of wireless communication systems.
{"title":"Harnessing Machine Learning for Intelligent Networking in 5G Technology and Beyond: Advancements, Applications and Challenges","authors":"Kristi Dulaj;Abdulraqeb Alhammadi;Ibraheem Shayea;Ayman A. El-Saleh;Mohammad Alnakhli","doi":"10.1109/OJITS.2025.3564361","DOIUrl":"https://doi.org/10.1109/OJITS.2025.3564361","url":null,"abstract":"A revolutionary age in telecommunications is being ushered in by the confluence of machine learning (ML) with fifth-generation (5G) wireless communication technologies and beyond. This research investigates ML approaches in 5G networks for adaptive spectrum usage, quality of service (QoS) management, predictive maintenance, and network optimization. By leveraging ML algorithms, 5G networks can forecast user behavior, allocate resources optimally, and dynamically adjust to changing conditions, enhancing performance and dependability. Additionally, ML-driven methods improve cybersecurity in 5G settings. Furthermore, the integration of ML in 5G networks is pivotal for advancing intelligent transportation systems, enabling dynamic route optimization, adaptive traffic management, and enhanced vehicular communication. Intelligent networks will transform wireless communication by replacing traditional processing with end-to-end solutions, utilizing cognitive radio systems and deep reinforcement learning for optimized spectrum sharing and efficiency. Despite significant potential, challenges such as interoperability, security, scalability, and energy efficiency must be addressed. This paper discusses these challenges and highlights future trends beyond 5G, emphasizing ML's critical role in shaping the future of wireless communication systems.","PeriodicalId":100631,"journal":{"name":"IEEE Open Journal of Intelligent Transportation Systems","volume":"6 ","pages":"605-633"},"PeriodicalIF":4.6,"publicationDate":"2025-04-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10977045","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144108343","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 : 2025-04-25DOI: 10.1109/OJITS.2025.3564558
Dewant Katare;David Solans Noguero;Souneil Park;Nicolas Kourtellis;Marijn Janssen;Aaron Yi Ding
Vulnerable road users (VRUs), including pedestrians, cyclists, and motorcyclists, account for approximately 50% of road traffic fatalities globally, as per the World Health Organization. In these scenarios, the accuracy and fairness of perception applications used in autonomous driving become critical to reduce such risks. For machine learning models, performing object classification and detection tasks, the focus has been on improving accuracy and enhancing model performance metrics; however, issues such as biases inherited in models, statistical imbalances and disparities within the datasets are often overlooked. Our research addresses these issues by exploring class imbalances among vulnerable road users by focusing on class distribution analysis, evaluating model performance, and bias impact assessment. Using popular CNN models and Vision Transformers (ViTs) with the nuScenes dataset, our performance evaluation shows detection disparities for underrepresented classes. Compared to related work, we focus on metric-specific and cost-sensitive learning for model optimization and bias mitigation, which includes data augmentation and resampling. Using the proposed mitigation approaches, we see improvement in IoU(%) and NDS(%) metrics from 71.3 to 75.6 and 80.6 to 83.7 for the CNN model. Similarly, for ViT, we observe improvement in IoU and NDS metrics from 74.9 to 79.2 and 83.8 to 87.1. This research contributes to developing reliable models while addressing inclusiveness for minority classes in datasets. Code can be accessed at: BiasDet.
{"title":"Analyzing and Mitigating Bias for Vulnerable Road Users by Addressing Class Imbalance in Datasets","authors":"Dewant Katare;David Solans Noguero;Souneil Park;Nicolas Kourtellis;Marijn Janssen;Aaron Yi Ding","doi":"10.1109/OJITS.2025.3564558","DOIUrl":"https://doi.org/10.1109/OJITS.2025.3564558","url":null,"abstract":"Vulnerable road users (VRUs), including pedestrians, cyclists, and motorcyclists, account for approximately 50% of road traffic fatalities globally, as per the World Health Organization. In these scenarios, the accuracy and fairness of perception applications used in autonomous driving become critical to reduce such risks. For machine learning models, performing object classification and detection tasks, the focus has been on improving accuracy and enhancing model performance metrics; however, issues such as biases inherited in models, statistical imbalances and disparities within the datasets are often overlooked. Our research addresses these issues by exploring class imbalances among vulnerable road users by focusing on class distribution analysis, evaluating model performance, and bias impact assessment. Using popular CNN models and Vision Transformers (ViTs) with the nuScenes dataset, our performance evaluation shows detection disparities for underrepresented classes. Compared to related work, we focus on metric-specific and cost-sensitive learning for model optimization and bias mitigation, which includes data augmentation and resampling. Using the proposed mitigation approaches, we see improvement in IoU(%) and NDS(%) metrics from 71.3 to 75.6 and 80.6 to 83.7 for the CNN model. Similarly, for ViT, we observe improvement in IoU and NDS metrics from 74.9 to 79.2 and 83.8 to 87.1. This research contributes to developing reliable models while addressing inclusiveness for minority classes in datasets. Code can be accessed at: BiasDet.","PeriodicalId":100631,"journal":{"name":"IEEE Open Journal of Intelligent Transportation Systems","volume":"6 ","pages":"590-604"},"PeriodicalIF":4.6,"publicationDate":"2025-04-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10977047","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144072774","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：