Article
Biotechnology & Applied Microbiology
M. Griffith, R. Akkem, J. Maheshwari, T. Seacrist, K. B. Arbogast, V. Graci
Summary: In highly autonomous driving scenarios, it is important to accelerate driver reaction times during the takeover of control. Previous studies have shown that an Acoustic Startling Pre-Stimulus (ASPS) can accelerate reaction times in simple ankle flexion exercises. This study examined whether an ASPS warning can lead to shorter takeover reaction times in a simulated evasive swerving maneuver. The results suggest that the ASPS can accelerate driver reaction times during correction maneuvers, and that warnings for autonomous vehicles may need to be tailored to the driver's age and sex.
FRONTIERS IN BIOENGINEERING AND BIOTECHNOLOGY
(2023)
Article
Psychology, Multidisciplinary
Koki Muto, Shoko Oikawa, Yasuhiro Matsui, Toshiya Hirose
Summary: This study investigated the effect of driver posture on driving control after a takeover request in autonomous vehicles, revealing that both upper-body and foot postures impact steering and braking reaction time.
BEHAVIORAL SCIENCES
(2022)
Article
Green & Sustainable Science & Technology
Qiuhong Wang, Haolin Chen, Jianguo Gong, Xiaohua Zhao, Zhenlong Li
Summary: This study analyzes the change in drivers' perception level and its influence on takeover performance in autonomous driving. It finds that after a takeover request, the driver's gaze duration is shortened and pupil area is enlarged, which helps in quicker extraction and understanding of road information. Male drivers have higher perception levels and prioritize leisure tasks. Age is found to decrease drivers' perception level. Shorter gaze duration and larger pupil area lead to shorter takeover response time. Drivers' perception level has a positive effect on takeover performance.
Article
Engineering, Civil
Rui Wang, Jinfeng Xu, Jia Liu, Di Wu, Yixue Hao, Xianzhi Li, Min Chen
Summary: This study proposes an intelligent fabric space enabled by multi-sensing sensors, and presents a behavior analysis pipeline to capture crowd information from micro-level and macro-level. By fusing multiple trajectories, it achieves precise crowd segmentation and motion description.
IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS
(2022)
Article
Ergonomics
Qian Liu, Xuesong Wang, Shikun Liu, Chunjun Yu, Yi Glaser
Summary: Intersections are high-risk locations for autonomous vehicles (AVs). Analyzing the pre-crash scenarios and contributing factors of AV crashes at intersections using the association rule method revealed that rear-end and lane change crashes were the most frequently occurring scenarios for AVs. The main contributing factors of these scenarios were identified, such as the location outside the intersection, traffic signal control, autonomous engaged mode, mixed-use or public land, and weekdays. Inadequate stop and deceleration decisions by the AV's automated driving system (ADS) and insufficient collision avoidance decisions in lane change crashes were important causes of these AV crashes.
ACCIDENT ANALYSIS AND PREVENTION
(2024)
Article
Public, Environmental & Occupational Health
Chao Huang, Bo Yang, Kimihiko Nakano
Summary: This study aims to extract drivers' patterns of gaze behaviors and maneuvers during takeovers from crash data and utilize them to provide guidance for human-machine-interface design to enhance safety and acceptability of automated driving. The study identified five typical patterns of unsafe behaviors and emphasized the importance of drivers' gaze behaviors and maneuvers in emergent situations.
TRAFFIC INJURY PREVENTION
(2023)
Article
Chemistry, Multidisciplinary
Okkeun Lee, Hyunmin Kang
Summary: In the context of partial autonomy, bringing drivers back into the loop is a challenge. This study investigates visual, auditory, and tactile stimuli in the autonomous environment, examining their variations on age, gender, and individual backgrounds. The research reveals correlations between age and reaction times for auditory and tactile signals, and explores subjective evaluations of the signals. By evaluating individual differences in perception, this study contributes to future research in the field.
APPLIED SCIENCES-BASEL
(2023)
Article
Psychology, Applied
Jianguo Gong, Xiucheng Guo, Cong Qi, Xiaoxi Liang, Qiuhong Wang
Summary: This study investigated the factors influencing automated vehicle takeover performance, including takeover request time, non-driving-related tasks, and driving scenarios. The results showed that these factors had significant effects on takeover behavior, and takeover request time, driving scenario complexity, and secondary work state tasks were correlated with takeover and control time.
TRANSPORTATION RESEARCH PART F-TRAFFIC PSYCHOLOGY AND BEHAVIOUR
(2023)
Article
Ergonomics
Xinghua Wang, Yong Peng, Tuo Xu, Qian Xu, Xianhui Wu, Guoliang Xiang, Shengen Yi, Honggang Wang
Summary: This study aims to develop autonomous driving testing scenarios for China by analyzing actual crash cases. The researchers identified six functional scenarios and their corresponding logical scenarios based on clustering and distribution analysis of dynamic parameters. They also proposed a virtual crash generation approach to obtain concrete testing scenarios. The results showed that the statistical characteristics of virtual crashes were consistent with those of original crashes. This study provides a theoretical basis and data support for establishing autonomous driving testing schemes tailored to the Chinese traffic environment.
ACCIDENT ANALYSIS AND PREVENTION
(2022)
Article
Chemistry, Analytical
Zhizhong Ding, Chao Sun, Momiao Zhou, Zhengqiong Liu, Congzhong Wu
Summary: Current research on autonomous driving vehicles mainly focuses on scenarios where manual and autonomous vehicles share the road, but there is potential for a future where all vehicles are autonomous. To improve driving safety and production cost expectations, adjustments need to be made to methods of environment sensing, traffic instruction, and vehicle control.
Article
Engineering, Civil
Songyi Zhang, Zhiqiang Jian, Xiaodong Deng, Shitao Chen, Zhixiong Nan, Nanning Zheng
Summary: The study introduces a hybrid A* motion planning method based on hierarchical search spatial scales, capable of generating smooth and safe paths in large-scale complex scenarios. It covers two stages, generating local goals on a coarse scale and exploring paths on a fine scale with novel heuristic functions and strategies. The use of clothoid improves path smoothness to G(2)-continuous, meeting vehicle kinematic constraints effectively.
IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS
(2022)
Article
Engineering, Civil
Wonteak Lim, Seongjin Lee, Myoungho Sunwoo, Kichun Jo
Summary: This paper proposes a hybrid trajectory planning scheme that integrates the strength of sampling and optimization methods to solve the problem of safe trajectory planning in dynamic driving environments. The sampling method is used for lateral movement while the numerical optimization method is used for longitudinal movement, helping the planner generate adaptive trajectories in various situations.
IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS
(2021)
Article
Engineering, Civil
Canming Xia, Mali Xing, Shenghuang He
Summary: This article introduces a planning framework based on partially observable Markov decision process (POMDP) to ensure social compliance and optimize motion response for autonomous vehicles. By designing a novel POMDP model and using a deterministic planner, the framework is able to effectively predict behavioral intentions and decision-making, with good adaptability to real-world scenarios.
IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS
(2022)
Article
Engineering, Civil
Luca Anzalone, Paola Barra, Silvio Barra, Aniello Castiglione, Michele Nappi
Summary: This work combines Curriculum Learning with Deep Reinforcement Learning to learn a competitive driving policy without prior domain knowledge in the CARLA autonomous driving simulator. The approach divides the reinforcement learning phase into multiple stages of increasing difficulty, guiding the agent towards an increasingly better driving policy. The agent architecture includes various neural networks and novel value decomposition scheme and gradient size normalization function. Quantitative and qualitative results of the learned driving policy are presented.
IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS
(2022)
Article
Transportation
Qingchao Liu, Ruohan Yu, Yingfeng Cai, Long Chen
Summary: This study uses the XGBoost algorithm to predict the crash risk of autonomous vehicles in different highway sections. By selecting appropriate variables and threshold values, the prediction accuracy can be improved and costs can be reduced, providing reference for designing safer and more economical autonomous vehicles.
TRANSPORTATION LETTERS-THE INTERNATIONAL JOURNAL OF TRANSPORTATION RESEARCH
(2023)
Article
Engineering, Electrical & Electronic
Haijian Li, Guoqiang Zhao, Lingqiao Qin, Yanfang Yang
IEEE SENSORS JOURNAL
(2020)
Article
Computer Science, Interdisciplinary Applications
Shuo Zheng, Haijian Li, Zhufei Huang, Keyi Li, Lingqiao Qin
SIMULATION MODELLING PRACTICE AND THEORY
(2020)
Article
Engineering, Civil
Haijian Li, Zhufei Huang, Xiaofang Zou, Shuo Zheng, Yanfang Yang
JOURNAL OF ADVANCED TRANSPORTATION
(2020)
Article
Engineering, Civil
Tong Mo, Keyi Li, Junjie Zhang, Lingqiao Qin, Zhufei Huang, Haijian Li
JOURNAL OF ADVANCED TRANSPORTATION
(2020)
Article
Engineering, Electrical & Electronic
Xin Chang, Jian Rong, Haijian Li, Yiping Wu, Xiaohua Zhao
Summary: This study evaluates the impacts of in-vehicle human-machine interfaces (HMI) on driving performance, traffic safety, and eco-driving behavior using a driving simulator. Results show that HMI can be beneficial for speed adjustment and may affect time headway, but has no significant effects on driving comfort and eco-driving behavior. Future research is needed to develop more efficient technology and warning strategies to enhance eco-safe driving in different zones. The findings are useful for active safety management and evaluating the effectiveness of connected vehicle systems.
IET INTELLIGENT TRANSPORT SYSTEMS
(2021)
Article
Engineering, Civil
Haijian Li, Guoqiang Zhao, Lingqiao Qin, Hanimaiti Aizeke, Xiaohua Zhao, Yanfang Yang
Summary: Motor vehicle collisions result in significant deaths, disabilities, and economic losses globally. This article highlights the importance of timely driver warnings in reducing accident frequency and improving traffic safety. Research on safety warnings in connected vehicle environments has made progress and can serve as a reference for future studies in various aspects.
IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS
(2021)
Article
Engineering, Civil
Wei Guan, Haolin Chen, Xuewei Li, Haijian Li, Xin You
Summary: This study aims to assess the impact of different levels of connected vehicle fog warning systems on driving behavior and safety. Results show that the connected vehicle fog warning system can significantly reduce driving speed and improve driving safety. Furthermore, drivers' responses are more pronounced in the group with a human-machine interface, where they maintain a lower speed.
JOURNAL OF ADVANCED TRANSPORTATION
(2022)
Article
Engineering, Civil
Haijian Li, Junjie Zhang, Zhonghua Liu, Jianguo Gong, Xiaohua Zhao
Summary: This study investigates the influence of secondary task immersion duration on the take-over process through driving simulation experiments. The results show that the response time for take-over is positively correlated with immersion duration, and the take-over performance varies with different immersion durations. Furthermore, the immersion duration of secondary tasks also affects the efficiency of section traffic flow.
TRANSPORTATION RESEARCH RECORD
(2022)
Article
Engineering, Civil
Dongwei Xu, Hang Peng, Chenchen Wei, Xuetian Shang, Haijian Li
Summary: Road traffic state estimation is crucial for ITSs, with incomplete data being a common issue. The GA-GAN model proposed in this study utilizes GraphSAGE and GAN to impute missing road traffic state data by reconstructing the road network structure based on correlation coefficients of historical data. Experimental results on California and Seattle traffic data demonstrate the model's superior performance compared to state-of-the-art methods.
IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS
(2022)
Article
Green & Sustainable Science & Technology
Jun Niu, Shan Lin, Erlong Lou, Zongdian Li, Kaiqun Chen, Haijian Li
Summary: A variable speed limit system suitable for freeway bottleneck areas was constructed and tested using VISSIM microscopic traffic simulation software. The research showed that reasonable speed limits could effectively reduce roadway delays and improve operational efficiency in certain traffic flow ranges.
Article
Engineering, Civil
Dongwei Xu, Xuetian Shang, Hang Peng, Haijian Li
Summary: The future trajectory prediction of heterogeneous traffic-agents for autonomous vehicles in mixed traffic scene is of great significance for safe and reliable driving. Thus, we propose the Multi-View Adaptive Hierarchical Spatial Graph Convolution Network (MVHGN) to predict the future trajectories of heterogeneous traffic-agents. Our proposed method outperforms the comparison models in the Apolloscape trajectory data set.
IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS
(2023)
Article
Computer Science, Artificial Intelligence
Dongwei Xu, Xuetian Shang, Yewanze Liu, Hang Peng, Haijian Li
Summary: Vehicle trajectory prediction is a challenging task in autonomous driving and it plays a crucial role in ensuring safety. To address the interaction between vehicles and their own trajectory, a novel graph network model is proposed. This model utilizes complex network methodology to construct a correlation network for vehicles at each time, and an adaptive parameter matrix is introduced to optimize the global spatio-temporal graph. By extracting global features through stacked graph convolution module, the model achieves accurate trajectory prediction for road vehicles in the future. Experimental results demonstrate the superiority of our model compared to other advanced schemes, achieving higher accuracy in predicting future vehicle trajectories.
IEEE TRANSACTIONS ON INTELLIGENT VEHICLES
(2023)
Article
Engineering, Civil
Jianhua Zhang, Xiaohua Zhao, Haijian Li, Jianyu Qi, Guanyang Xing
Summary: This study compared the effects of traditional driving environment and connected vehicle environment on driving performance and traffic safety through a driving simulation experiment. The results showed that in the connected vehicle environment, the driver's response efficiency and deceleration behavior were both improved. The warning information provided by the in-vehicle human-machine interface enabled drivers to have earlier awareness of road conditions ahead and adjust their speed more gently, resulting in a more stable vehicle speed when entering the tunnel. In addition, there were significant differences in brake-movement time among drivers of different driving ages, with brake-movement time increasing with driving age. This suggests that experienced drivers usually adopt a softer way to adjust speed, ensuring driving safety and avoiding potential dangers of drastic speed changes. This study proposed a validity evaluation method based on survival analysis for tunnel entrance scenarios, providing a reference for connected vehicle evaluation.
IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS
(2023)
Article
Ergonomics
Xiaohua Zhao, Wenxiang Xu, Jianming Ma, Haijian Li, Yufei Chen, Jian Rong
ACCIDENT ANALYSIS AND PREVENTION
(2019)