Low Traffic Density Research Articles

Tree-based approaches to understanding factors influencing crash severity across roadway classes: A Thailand case study

Existing studies often overlook the nuanced differences between various road classifications and their respective crash dynamics, hindering the development of targeted interventions to mitigate crash severity. To address this gap, this study investigates factors influencing the likelihood of fatality in road crashes across highways, collector roads, and local roads in Thailand using crash data from 2015 to 2021. Highways connect regions with high-speed traffic and large volumes, collector roads link smaller communities with lower traffic density but allow higher speeds, and local roads primarily pass through villages, with narrow pathways, two traffic lanes, and frequent motorcycle use. The study employs machine learning methodologies utilizing tree-based algorithms, including Decision Trees, Random Forest, Gradient Boosting, AdaBoost, Extra Trees, XGBoost, LightGBM, and CatBoost. The XGBoost model delivered superior performance for highways, while Gradient Boosting slightly outperformed XGBoost for local and collector roads. Both models consistently achieved a test accuracy of 0.70, with precision between 0.66 and 0.67, recall ranging from 0.59 to 0.61, and F1-scores from 0.58 to 0.61. The AUC values also consistently ranged from 0.59 to 0.61. SHAP values reveal key factors influencing fatality risk across road types, including speeding, gender disparities, driving under the influence of alcohol, inadequate lighting, and elderly drivers. Specific concerns include reversing on highways, collisions in poorly lit areas on collector roads, and helmet non-use on local roads. The findings support policy recommendations to address speeding, target male and older drivers, prevent reversing incidents, enhance lighting, and promote helmet use. This research deepens our understanding of factors affecting road crash severity and offers valuable insights for improving road safety across various environments.

Effects of route guidance strategies on traffic emissions in urban traffic networks

Urban traffic emissions have significant environmental and health implications. Diverging from traditional research that primarily aims to improve traffic flow and efficiency, this study specifically focuses on the environmental impact on traffic emissions, conducting a comprehensive analysis within the Manhattan urban network through four route guidance strategies. The performance of these strategies is examined across various vehicle densities, and their impact on four traffic pollutant emissions (Carbon-dioxide, Nitrogen Oxides, Volatile Organic Compounds and Particulate Matter) is assessed. Moreover, our innovative approach analyzes emissions from the perspectives of both travel distance and trip frequency, placing special emphasis on trip frequency to provide practical insights with high real-world applicability. The results highlight the potential and limitation of the Congestion Coefficient Strategy. Under equal travel demands, the Congestion Coefficient Strategy showed promise in reducing carbon emissions. However, at lower vehicle densities, it led to a significant increase in emissions. This revelation pointed to the need for modifications to the strategy when applied in scenarios with lower traffic density. Recognizing this limitation, we introduced a modified strategy that achieved remarkable reductions in emissions across diverse vehicle densities, effectively overcoming the challenges posed by the original Congestion Coefficient Strategy. These findings offer valuable insights for policymakers and transportation planners in selecting optimal route guidance strategies to reduce pollutant emissions. Future studies will explore the efficacy of these strategies in road networks characterized by different topological configurations.

Safe Hybrid-Action Reinforcement Learning-Based Decision and Control for Discretionary Lane Change

Autonomous lane-change, a key feature of advanced driver-assistance systems, can enhance traffic efficiency and reduce the incidence of accidents. However, safe driving of autonomous vehicles remains challenging in complex environments. How to perform safe and appropriate lane change is a popular topic of research in the field of autonomous driving. Currently, few papers consider the safety of reinforcement learning in discretionary lane-change scenarios. We introduce safe hybrid-action reinforcement learning into discretionary lane change for the first time and propose the Parameterized Soft Actor–Critic with PID Lagrangian (PASAC-PIDLag) algorithm. Furthermore, we conduct a comparative analysis with Parameterized Soft Actor–Critic (PASAC), which is an unsafe version of PASAC-PIDLag. Both algorithms are employed to train the lane-change strategy to output both discrete lane-change decisions and continuous longitudinal vehicle acceleration. Our simulation results indicate that at a traffic density of 15 vehicles per kilometer (15 veh/km), the PASAC-PIDLag algorithm exhibits superior safety with a collision rate of 0%, outperforming the PASAC algorithm, which has a collision rate of 1%. The generalization assessments reveal that at low traffic density levels, both the PASAC-PIDLag and PASAC algorithms are proficient in attaining zero collision rates. However, at high traffic density levels, although both algorithms result in collisions, PASAC-PIDLag has a much lower collision rate than PASAC.

Object Detection, Recognition, and Tracking Algorithms for ADASs-A Study on Recent Trends.

Advanced driver assistance systems (ADASs) are becoming increasingly common in modern-day vehicles, as they not only improve safety and reduce accidents but also aid in smoother and easier driving. ADASs rely on a variety of sensors such as cameras, radars, lidars, and a combination of sensors, to perceive their surroundings and identify and track objects on the road. The key components of ADASs are object detection, recognition, and tracking algorithms that allow vehicles to identify and track other objects on the road, such as other vehicles, pedestrians, cyclists, obstacles, traffic signs, traffic lights, etc. This information is then used to warn the driver of potential hazards or used by the ADAS itself to take corrective actions to avoid an accident. This paper provides a review of prominent state-of-the-art object detection, recognition, and tracking algorithms used in different functionalities of ADASs. The paper begins by introducing the history and fundamentals of ADASs followed by reviewing recent trends in various ADAS algorithms and their functionalities, along with the datasets employed. The paper concludes by discussing the future of object detection, recognition, and tracking algorithms for ADASs. The paper also discusses the need for more research on object detection, recognition, and tracking in challenging environments, such as those with low visibility or high traffic density.

The monitoring requests on young driver’s fatigue and take-over performance in prolonged conditional automated driving

Introduction: L3 automated vehicles can perform all dynamic driving tasks unless a take-over occurs due to operational limits. This issue is potentially important for young drivers who are vulnerable road users since they have skill deficits and easily evolve into aberrant driving. However, drivers lacking active involvement may be fatigued and drowsy. Previous research indicated that performing a voluntary non-driving-related task (NDRT) could keep drivers alert, but there was no difference in take-over performance with or without NDRT. Providing a monitoring request (MR) before a possible take-over request (TOR) exhibited better take-over performance in temporary automated driving. Therefore, the study aimed to investigate the effects of MR and voluntary NDRT on young drivers’ fatigue and performance. Method: Twenty-five young drivers experienced 60 min automated driving on a highway with low traffic density and a TOR prompted due to a collision event. A within-subjects was designed that comprised three conditions: NONE, TOR-only, and MR + TOR. Drivers were allowed to perform a self-paced phone NDRT during automated driving. Results: The PERCLOS and blink frequency data showed that playing phones could keep drivers vigilant. The take-over performance on whether taking phone had no difference, but with MRs condition exhibited better take-over performance including the shorter reaction time and the longer TTC. Subjective evaluations also showed the advantages of MRs with more safety, trust, acceptance, and lower workload. Conclusions: Taking MRs had a positive effect on relieving fatigue and improving take-over performance. Furthermore, MRs could potentially improve the safety and acceptance of automated driving. Practical Applications: The MR design can be used in the automotive industry to ensure the safest interfaces between fatigue drivers and automation systems.

6 Back-To-Drive Project. Assessment of Fitness-To-Drive in Older and Cognitively Impaired Adult Drivers: Adaptation of the on-Road Driving Observation Schedule to Simulation-Based Environments

Objective:Cognitive, motor and sensory deficits associated with aging, and with some neurological conditions such as acquired brain injury, may lead to severe driving performance impairment. While rehabilitation and driver assistance technologies may improve driving performance, the assessment of the actual fitness-to-drive of these people is challenging. Office-based neuropsychological/physical tests are considered insufficient to understand one’s ability to drive. The gold standard is the on-road assessment with dual control cars, superior in ecological validity, but expensive, stressful, and potentially unsafe. Valid, more cost-effective solutions for a safer, more accurate, standardized assessment of fitness-to-drive are currently needed. Modern and sensorized driving simulators offer key advantages, such as the possibility of exposing drivers to several relevant driving scenarios, including hazard situations, and of assessing their driving performance without being physically at risk. However, the extraction and direct interpretation of existing simulator-produced data may require specialized data processing skills or simulation expertise. To overcome this, we have developed an easy-to-use, pencil-and-paper observational instrument. The Sim-DOS is an adaptation of the widely used instrument to assess “natural driving”, the Driving Observation Schedule (DOS; Vlahodimitrakou et al., 2013).Participants and Methods:Via expert consensus, DOS targeted behaviors were adapted to a simulated-based environment (signaling, observation of environment, speed regulation, slow or unsafe reaction, distance interpretation, vehicle/lane positioning), and the Sim-DOS scores calculation (based on errors while doing such behaviors) was adapted from DOS to include hazard situations (HS, 0-100) and free driving (FD, 0-°°) scores. The instrument was then piloted with a sample of 35 older adults, along with the collection of simulator-produced data on number of harsh events and driving speed. Participants drove two consecutive 20-minutes long scenarios, with low and high traffic density (LTD, HTD). In each scenario, there were periods with and without potentially hazard situations.Results:Assessments were performed by two independent trained observers, producing substantial inter-rater reliability (intra-class correlation coefficients above 0.94). Participants (70.7±4.1 years old, 60% male, 46.1±6.7 years of driving experience) were mostly regular drivers (74%). However, psychomotor skills of the majority were compromised, with only one participant being above the 80th percentile in the reaction times test of the national mandatory driving assessment. When exposed to hazard situations, most of the participants (94.1%) did not perform well, independently of traffic density, with average Sim-DOS-HS scores of 87.1±9.7 (out of 100, t-values>7.3, p-values<.05).Compared to LTD scenarios, in HTD scenarios participants drove less smoothly (HTD:0.97±1.24 vs. LTD:0.33±0.58 of harsh events, Z=3.1, p<.05). However, they also drove slower (HTD:82.41±27.43 vs. LTD:103.55±14.61 km/h, t=5.2, p<.05), improving their ability to manage hazard situations, and therefore producing higher than expected Sim-DOS scores (HTD:87.05±10.28). During free driving, participants performed worse under LTD conditions (Sim-DOS-FD scores: HTD:11.68±6.20 vs. LTD:14.40±9.58, t=2.15, p<.05) as they drove at higher speed (HTD:85.01±24.28 vs. LTD:104.70±11.94 km/h, t=5.8, p<.05), although they did it more smoothly (HTD:1.94±3.74 vs. LTD:0.45±0.74 harsh events, Z=2.65, p<.05).Conclusions:Our study provides a validated driving assessment tool for use in driving simulators that will allow a safer, more ecologic, holistic and informative evaluation of the fitness-to-drive of older adults and neurological patients.

Edge-independent spanning trees in folded crossed cubes

Edge-independent spanning trees (for short EISTs) have widespread applications in fault-tolerance to enhance stability and security of networks, as well as in IP fast rerouting to prevent network breakdown caused by link failure. Then, the algorithms for constructing EISTs on many classes of graphs have been investigated. The folded crossed cube was proposed based on the folded cube and the crossed cube, which possesses such appealing properties as short diameter, short mean internode distance and very low message traffic density. In this paper, we study the existence and construction of EISTs with the same root r in the n-dimensional folded crossed cube (for short FCQn). For v∈V(FCQn)∖{r} and i∈{0,1,⋯,n−1}, we first propose two algorithms to obtain the sequence Sv,i and the set Fv, respectively. Then, based on them, an algorithm with time complexity O(n2) by using N processors is proposed to construct n+1 EISTs rooted at any vertex r in FCQn, where N=2n. And the corresponding theoretical proof and simulation experiments are presented to verify its validity. Since FCQn is (n+1)-regular, the result is optimal with respect to the number of EISTs constructed. Moreover, the performance of the proposed algorithm is evaluated experimentally in terms of average distance and average distance-diameter ratio of resulting EISTs.

Multidimensional framework for analyzing the distribution patterns of traditional villages in the karst landscape regions of China

The habitat environment of karst regions is facing challenges, such as rocky desertification and soil erosion. Traditional villages are among the most important components for the long-term survival and development of the habitat in karst areas, so the distribution patterns of traditional villages need to be understood for the conservation of karst habitats. In this study, a multidimensional framework was constructed to explore the distribution patterns of traditional villages in karst areas of China from planimetric, topographic, humanistic, socioeconomic, and hydrogeologic dimensions. We used geographical spatial analysis indicators and related analysis methods to analyze the main distribution patterns of traditional villages in karst landscape areas. Results showed that in the planimetric dimension, the distribution of traditional villages exhibited a cohesive pattern of “one core and many points.” In the topographic dimension, the number of traditional villages showed an inverted-U-shape change with rising altitude, and the villages were concentrated in the area with an altitude of 600–800 m a.s.l and a slope <5°. In the humanistic dimension, the distribution was dominated by Han (37.32%), Miao (21.3%), and Dong (10.61%) ethnic minorities, which were mostly found in the low-population-density area with 100–200 people/km2. In the socioeconomic dimension, the traditional villages had low traffic density and GDP per capita and were far from cities. In the hydrogeological dimension, the traditional villages were mainly located in areas with a groundwater runoff modulus >30 L/s-km2. This study expands our understanding of the characteristics of traditional settlements in karst areas and provides effective guidance for the future planning and heritage management of settlements in karst areas.

Development and techno-economic analysis of hybrid annuity model for highway projects of India

The hybrid annuity model is relatively new in the Indian economy as the private investors are still reluctant to adopt this public–private partnership model. This paper aims at developing and proposing a suitable hybrid annuity model for a highway project in Gujarat, India. A detailed financial analysis has been carried out to compute the net present value and internal rate of return. It has been observed that the breakeven and payback period is obtained after about 8 to 9 years. The corresponding net present value is about 28.253 million Indian rupee. The corresponding internal rate of return as observed from the analysis is about 20%. Thereby, the HAM model has been found to be techno-economically suitable and feasible for the development of new highways in Indian subcontinent particularly in the stretches of low traffic density.

PATRIC: A high performance parallel urban transport simulation framework based on traffic clustering

Parallel traffic simulation requires partitioning the road network into several components that can be assigned to different computing nodes (CPNs). Existing studies focus more on reducing edge-cuts (message-passing pipes between CPNs) to decrease synchronization message amongst CPNs for efficiency improvement. However, even reducing edge-cuts drastically, the volume of messages transmitted might still be high, which does not significantly improve performance. Based on observation that some traffic clusters (TCs) exist during simulation, i.e., areas with high internal and low external traffic density. For high-performance urban transport simulation, we propose a data-driven parallel approach named PATRIC, which can generate parallel partitions automatically based on traffic clustering. Specifically, the TC-based automatic partitioner (TAP) is designed to automatically identify TCs and then construct partitions in parallel. We present a partition-growing algorithm that prevents traffic-intensive TCs being split across multiple CPNs when distributing computing workloads, resulting in more balanced load and fewer synchronization operations. Unlike prior work using fixed thresholds for load balancing, we develop the adaptive partition updater (APU) to fit the dynamic traffic in the road network, which achieves a better trade-off between balancing workload and lowering communication for higher efficiency. Experiments on real-world datasets demonstrate that our approach outperforms the state-of-the-art methods.

Cyclical Trends of Network Load Fluctuations in Traffic Jamming

The transport of information packets in complex networks is a prototype system for the study of traffic jamming, a nonlinear dynamic phenomenon that arises with increased traffic load and limited network capacity. The underlying mathematical framework helps to reveal how the macroscopic jams build-up from microscopic dynamics, depending on the posting rate, navigation rules, and network structure. We investigate the time series of traffic loads before congestion occurs on two networks with structures that support efficient transport at low traffic or higher traffic density, respectively. Each node has a fixed finite queue length and uses next-nearest-neighbour search to navigate the packets toward their destination nodes and the LIFO queueing rule. We find that when approaching the respective congestion thresholds in these networks, the traffic load fluctuations show a similar temporal pattern; it is described by dominant cyclical trends with multifractal features and the broadening of the singularity spectrum regarding small-scale fluctuations. The long-range correlations captured by the power spectra show a power-law decay with network-dependent exponents. Meanwhile, the short-range correlations dominate at the onset of congestion. These findings reveal inherent characteristics of traffic jams inferred from traffic load time series as warning signs of congestion, complementing statistical indicators such as increased travel time and prolonged queuing in different transportation networks.

Road traffic density and recurrent asthma emergency department visits among Medicaid enrollees in New York State 2005–2015

BackgroundEnvironmental exposures such as traffic may contribute to asthma morbidity including recurrent emergency department (ED) visits. However, these associations are often confounded by socioeconomic status and health care access.ObjectiveThis study aims to assess the association between traffic density and recurrence of asthma ED visits in the primarily low income Medicaid population in New York State (NYS) between 2005 and 2015.MethodsThe primary outcome of interest was a recurrent asthma ED visit within 1-year of index visit. Traffic densities (weighted for truck traffic) were spatially linked based on home addresses. Bivariate and multivariate logistic regression analyses were conducted to identify factors predicting recurrent asthma ED visits.ResultsIn a multivariate model, Medicaid recipients living within 300-m of a high traffic density area were at a statistically significant risk of a recurrent asthma ED visit compared to those in a low traffic density area (OR = 1.31; 95% CI:1.24,1.38). Additionally, we evaluated effect measure modification for risk of recurrent asthma visits associated with traffic exposure by socio-demographic factors. The highest risk was found for those exposed to high traffic and being male (OR = 1.87; 95% CI:1.46,2.39), receiving cash assistance (OR = 2.11; 95% CI:1.65,2.72), receiving supplemental security income (OR = 2.21; 95% CI:1.66,2.96) and being in the 18.44 age group (OR = 1.59;95% CI 1.48,1.70) was associated with the highest risk of recurrent asthma ED visit. Black non-Hispanics (OR = 2.35; 95% CI:1.70,3.24), Hispanics (OR = 2.13; 95% CI:1.49,3.04) and those with race listed as “Other” (OR = 1.89 95% CI:1.13,3.16) in high traffic areas had higher risk of recurrent asthma ED visits as compared to White non-Hispanics in low traffic areas.ConclusionWe observed significant persistent disparities in asthma morbidity related to traffic exposure and race/ethnicity in a low-income population. Our findings suggest that even within a primarily low-income study population, socioeconomic differences persist. These differences in susceptibility in the extremely low-income group may not be apparent in health studies that use Medicaid enrollment as a proxy for low SES.

A comparison of the application potential of waterborne platooning for the Danube and the Rhine corridors

ObjectiveThis study investigates the implementation of the waterborne platooning transport concept in two of the largest European inland navigation corridors, the Rhine and the Danube region. Each region has different geo-economic and environmental features. These features are compared, and their effects on implementing a waterborne platooning transport concept are studied. The waterborne platooning concept, referred to as the Vessel Train, aims to reduce crew cost by automating the navigation tasks and moving the navigational responsibility to the leading vessel of the platoon, which is fully manned.MethodsThe implementation of the Vessel Train is assessed by making use of a developed model, which allows the assessment of the concept's viability by comparing the annual cost per transported ton of a reference vessel that sails individually to a vessel that sails as a part of a VT on the same route.ResultsThe results conclude that the application of waterborne platooning on the Rhine is more promising than on the Danube. The low wages hamper the implementation of the concept on the Danube in the region, the low traffic density on the waterway, and the common use of large push tows instead of self-propelled vessels.Implications for researchAs determined in the analysis for the Rhine case, a reduction in transport cost would make waterborne transport more attractive. However, other factors, such as the further integration of the VT in the overall supply chain, play a role in the successful implementation of this IWT transport concept. Applying the VT concept in the Danube case requires more potential cargo flows, which can be obtained by adding push convoys into the vessel train. This way of transport is more numerous on the Danube than self-propelled vessels. Both of these aspects should be studied further.

Cooperative-driving control for mixed fleets at wireless charging sections for lane changing behaviour

Recently, electric vehicles (EVs) have become an important traffic tool in connected and automated environments due to their own advantages. However, the optimal driving control strategies of EVs and charging modes of wireless charging lanes are complex since wireless charging technology is limited when EVs are moving and traffic conditions are random. This paper constructs a scenario in which wireless charging facilities are deployed in part of a single lane and proposes a cooperative control method of mixed fleets including EVs and fuel vehicles based on the four objectives of driving safety, traffic efficiency, passenger comfort and energy management. Combined with the particle swarm optimization algorithm, the average driving velocity, lane-changing frequency and location of vehicles under different deployment lengths of charging facilities and mixing ratios are compared to discuss the influence of deployment length on the control effect of energy consumption and traffic efficiency. The numerical results indicate that when the lane length is 1500 m, with the increasing proportion of EVs, the optimal deployment lengths are approximately 700 m, 550 m and 400 m in the case of low traffic density and approximately 550 m, 500 m and 450 m in the case of high traffic density.

Influence of Vehicle Traffic Emissions on Spatial Variation of Ozone and its Precursors in Air of Port Harcourt City, Nigeria

This study aimed at assessing the influence of vehicle traffic on spatial variation of ozone (O3) and its vehicular emission precursors in the air of Port Harcourt city. Sampling was carried out in ten (10) sites, eight (8) located within the high traffic density area (study sites) and two (2) located within the very low traffic density area (reference sites). The precursor pollutants measured were nitrogen dioxide (NO2), carbon monoxide (CO) and volatile organic compounds (VOCSs). Ozone and the precursor pollutants were measured in situ using AeroQUAL 500 series portable ambient air analyzer while traffic flow survey was achieved by direct counting. Measurements were carried out at morning, evening and off-peak traffic periods respectively. The mean concentrations of ozone and the precursor pollutants were significantly higher (p = 0.05) in the study sites than in the reference sites. Mean concentrations were higher at peak traffic periods than at off-peak traffic periods except for ozone that was higher at off-peak than at morning peak. There was significant correlation between traffic density and each of the pollutants including ozone. The spatial variability in concentration of pollutants was influenced by vehicular traffic. VOCs and NO2 levels were higher than the National Ambient Air Quality Standard (NAAQS) limit of 0.05 ppm and 0.04 – 0.06 ppm respectively, O3 concentration was below the standard limit (0.06 ppm) but was at the verge of exceeding. Traffic emission within the city was significant and could be mitigated through regular monitoring and control.

A Collaborative Road Traffic Regulation Approach Using a Wireless Sensor Network

In this paper, we detail and evaluate a coordinated approach to determining the sequence and duration of green lights at several intersections as part of the Intelligent Transportation System. We present the architecture of a wireless network used to track variations in adjacent intersections. Our algorithm exploits the collected data to determine the sequence of the green lights based on three objectives: (i) reduce the length of queues in the intersection, (ii) prioritize sending vehicle flows to intersections with lower traffic density than the most congested, (iii) synchronize traffic signals between adjacent intersections to create green waves. Traffic simulations have been simulated by the SUMO traffic simulator, they show that our solution manages to react to traffic change and reduce waiting time compared to isolated control strategies.

A web-based control system for traditional street lighting that uses high-pressure sodium lamps

Street lighting is a critical component of any city's infrastructure. On the other hand, the street lighting system consumes a significant amount of electricity. As a result, many technologies and studies are being developed to reduce the energy cost of street lighting. While the majority of the proposed ideas for reducing the energy cost of the street lighting system are based on light emitting diode lamps, they are not suitable for high-pressure sodium lamps, which continue to dominate in developing countries. Moreover, the high initial cost, difficulty of installation and maintenance, reliability, and service lifetime are all significant barriers to the practical implementation of these ideas. This paper presents a web-based control system for traditional street lighting systems that still employs high-pressure sodium lamps. The proposed idea converts existing modules of the conventional controller, which are photo switches, into IoT devices. The web application on the server then manages and controls the devices. The web application allows users to create a schedule for turning off the lights during the late-night hours to save energy. The system's advantages include its low cost, ease of installation, and maintenance. The proposed system is useful for roads or areas with low traffic density at late night. This system has been validated at Walailak University, Thailand.

Road Traffic Safety Before and During COVID-19: Have Restrictions Reduced Fatalities?

During COVID-19 pandemic, many restrictions have been applied to reduce social interaction, which might be related with road traffic density. The aim of the current study was to examine the relationship between COVID-19 related restrictions and road traffic accidents by comparing the monthly rates between 2019 and 2020 (traffic fatalities, traffic injuries, traffic accidents with injuries and fatalities, and traffic accidents with property damage only). For the analyses, the traffic accident data were computed per 1000 people in province for all 81 cities in Turkey to make more accurate comparisons. The traffic accident-related injury rates decreased in all months between March and December. Similarly rate of accidents with injuries and fatalities decreased in all months except for July. However, the road traffic fatality rates decreased only in May and October and, the rates of accidents with property damage only decreased in March, April, May, June, and December. According to results, although the rates of accidents declined in general, the fatality rates did not show a similar pattern. The low traffic density, which is due to the restrictions, might be related with higher speeds on the roads. The findings were discussed based on timeline of restrictions that have been applied by the government. It can be suggested that speed related enforcements might be applied even there is lockdown in a province not just to decrease number of accidents but also fatalities.

A systematic analysis for maritime accidents causation in Chinese coastal waters using machine learning approaches

Maritime safety is critical as many maritime accidents involve catastrophic consequences, including both fatalities and financial loss. To identify the factors which caused maritime accidents and provide a comprehensive suggestion for maritime administrators and mariners, a data-driven Bayesian Network (BN) is developed to analyse the major accidents records in the Chinese coastal waters by using an advanced machine learning approach. For this aim, the statistical interactions among causation factors identified from major accidents records are paired to construct the BN. Then the obtained BN is validated through a two-step validation process and a comparison analysis to demonstrate its superiorities in reliability and efficiency. The results revealed the importance of different risk influencing factors and the critical scenarios in the coastal waters. The small general cargo ships are the riskiest of in the coastal waters of China. While, bad weather conditions often lead to catastrophic accidents and minor accidents often happen in waters with low traffic density. The research findings could provide useful guidance to support risk preventions, and advance the maritime safety management system in coastal waters.

Which accuracy levels of positioning technologies do drivers really need in connected vehicle settings for safety?

Based on Global Navigation Satellite Systems (GNSS) and Internet of Vehicles (IoV), collision warning systems can provide drivers specific warning information such as the position of hazardous vehicles. However, positioning technologies vary in accuracy. Highly accurate positioning technologies are theoretically favorable for driving safety but are relatively difficult to use widely due to the high cost. While many researchers compared accuracies of different positioning technologies or explored methods to improve accuracy, few directly investigated the effect of different positioning accuracies for hazardous vehicles on driving safety. The present study conducted a laboratory experiment to explore the effect of four positioning accuracies of hazardous vehicles (i.e., highest: 1.5 m, high: 5 m, medium: 8.5 m, and low: 20 m, each value represented the radius of a circle and the estimated positions of hazardous vehicle were within this circle) and different traffic densities (high: one car per 150 m and low: one car per 1500 m) on driver behaviors. Thirty participants were distributed into high and low traffic density groups. Each participant received auditory warning information, which provided positions of hazardous vehicles relative to the ego vehicle concerning four accuracy levels. Driving safety (i.e., number of collisions and min TTC), driving performance (i.e., brake onset reaction time, steer onset reaction time, and mean deceleration), and subjective workload for warnings with different positioning accuracies of hazardous vehicles were recorded and analyzed. Results suggested the following: 1) Under low traffic density, four positioning accuracies showed no significant differences on driving safety and performance as well as subjective workload. 2) Under high traffic density, the medium positioning accuracy (8.5 m) exerted no significant differences on driving safety and performance compared with the highest positioning accuracy (1.5 m), which had the fastest brake onset reaction time, the least number of collisions, and the largest mean deceleration. Moreover, when traffic density was high, warnings with the highest (1.5 m), high (5 m), and medium (8.5 m) positioning accuracies generated a significantly lower workload than the warning with low (20 m) positioning accuracy. In conclusion, low positioning accuracy (20 m), which was a general accuracy of A-GNSS (no added cost) or GNSS (SPP) (relatively low cost) in dense urban area, was feasible for providing warnings under low traffic density. Medium positioning accuracy (8.5 m), which was the accuracy of A-GNSS (no added cost) in open-sky area, was acceptable for supporting warning systems under high traffic density. These findings have implications for promoting the application of positioning technologies to realize warnings in the near future.