Traffic Risk Research Articles

One should really be more worried about too little physical activity than injuries while walking or cycling to school. Parents’ perception of risk concerning active school transportation.

IntroductionActive school transportation (AST) can be a powerful strategy to promote physical activity and improve children's health. Parents' perceptions of traffic risks are an important factor in their decision concerning their child's transport mode to school and addressing discordances between the percieved risks and actual incidents might contribute to promote children's AST. This study adopted a game originally designed for the context of information security to resonate with the context of Swedish AST and the aim of the study was to describe parent's experience of playing the game. MethodsA qualitative approach was chosen and data was collected using focus groups with 16 parents for totally 29 children aged 7–12. The collected data were analyzed using qualitative content analysis. ResultsThe results formed one main theme “Playing promotes parents' awareness and consideration of active school transportation” and three subthemes; “Fast, fun, and functional; Experiences from playing the game” showing that the parents perceived the game to be fun, easy, and time efficient to play. “Then and now. Own childhood experience vs being a parent” showing that parents' perception of risk can sometimes be barriers to their children's opportunity to use AST. “Personal perceptions vs statistical risk” showing that playing the game addressed the discordance between the parents' perceived risks and actual incidents, as well as promoted parents' awareness and consideration of AST. ConclusionsIncluding an intervention component such as a game for parents and using schools as the arena for implementing projects promoting AST to improve children's physical activity levels, might be a fruitful avenue.

How effective is reducing traffic speed for safer work zones? Methodology and a case study in Pennsylvania

Transportation agencies post and enforce reduced speed limits in work zones to ensure work zone safety, since traffic speed is found to be associated with work zone crash risks. However, prior findings on the relationship between speed and crash rate in work zones are inconsistent. This may be attributed to the methods of statistical associations between traffic speed and crash risks that do not necessarily discover true causal relations. In fact, work zone presence could lead to the reduction of actual traffic speed that influences crash risks, where it may also directly impose effects on crash risks as a result of work zone configurations. The actual traffic speed (not posted speed limit) is also known as a “mediator” where work zones can indirectly impact the crash risks. It is challenging to rigorously separate the causal effect of traffic speed on work zone crash risk from that directly caused by work zones. The underlying causal relation could help to determine what reduced post speed limit (with enforcement) is necessary to ensure work zone safety under the most desired “actual traffic speed”. This study proposes to use the sequential g-estimation and the regression discontinuity design to estimate the controlled direct effect of traffic speed on work zone crashes. Two research gaps are identified and filled: inaccurate inferences of the effect of reduced speed limit in work zones as a result of ignoring (1) potential post-treatment bias since traffic speed is a mediator; and (2) potential confounding bias caused by unobservable roadway characteristics. The proposed methodology was applied to 4008 work zones in Pennsylvania from 2015 to 2017, and the results were validated through a series of robustness tests. The results indicate that the direct causal effect of the presence of work zones on crash risk is significantly positive when the traffic speed is relatively low (i.e., lower than 55 mph in this case study), while traffic speed has a positive causal effect on crash occurrences when the actual traffic speed is high (i.e., greater or equal to 55 mph). It suggests that strictly enforcing reduced posted speed limits in work zones is particularly effective when the actual traffic speed is greater than 55 mph. This is particularly true on roadways with high traffic volume (i.e., AADT > 20,000 vehicles per day), long, and daytime work zones (i.e., > 3000 m). On the other hand, the effect of enforcing reduced speed on work zone safety is unclear when the actual speed is already low. In this case, improving work zone configurations and driving behaviors may be more effective in reducing crash risks.

Investigation on the driver-victim pairs in pedestrian and bicyclist crashes by latent class clustering and random forest algorithm

Pedestrians and bicyclists from marginalized and underserved populations experienced disproportionate fatalities and injury rates due to traffic crashes in the US. This disparity among road users of different races and the increasing trend of traffic risk for underserved racial groups called for an urgent agenda for transportation policy making and research to ensure equity in roadway safety. Pedestrian and bicyclist crashes involved drivers and pedestrians/bicyclists; the latter were usually victims. Traditional safety studies did not account for the interaction between the two parties and assumed that they were independent from each other. In this study we paired the driver and pedestrian/bicyclist involved in the same crash to understand the socioeconomic and demographic make-up of the two parties involved in crashes and assessed the geographic distribution of these crashes and crash-contributing factors. For this purpose, we applied thelatent class clustering analysis (LCA) to classify different crash types and analyze the patterns of the crashes based on the income and ethnicity of both drivers and victims involved in pedestrian and bicyclist crashes. We then used random forest algorithms and partial dependence plots (PDPs) to model and interpreted the contributing factors of the clusters in both pedestrian and bicyclist models. The clustering results showed a pattern of social segregation in pedestrian and bicyclist crashes that drivers and victims with similar socioeconomic characteristics tend to be involved in one crash. Pedestrian/bicyclist exposure, driver’s age, victim’s age, year of the car in use, annual average daily traffic (AADT), speed limit, roadbed width, and lane width were the most influential factors contributing to this pattern. Crashes that involved drivers and victims with lower income and non-white ethnicity tended to happen in the location with higher pedestrian/bicyclist exposure, higher speed limit, and wider road. The findings of this research can help to inform the decision-making process for improving safety to ensure equitable and sustainable safety for all road users and communities.

Analysis of Perception Accuracy of Roadside Millimeter-Wave Radar for Traffic Risk Assessment and Early Warning Systems.

Millimeter-wave (MMW) radar is essential in roadside traffic perception scenarios and traffic safety control. For traffic risk assessment and early warning systems, MMW radar provides real-time position and velocity measurements as a crucial source of dynamic risk information. However, due to MMW radar's measuring principle and hardware limitations, vehicle positioning errors are unavoidable, potentially causing misperception of the vehicle motion and interaction behavior. This paper analyzes the factors influencing the MMW radar positioning accuracy that are of major concern in the application of transportation systems. An analysis of the radar measuring principle and the distributions of the radar point cloud on the vehicle body under different scenarios are provided to determine the causes of the positioning error. Qualitative analyses of the radar positioning accuracy regarding radar installation height, radar sampling frequency, vehicle location, posture, and size are performed. The analyses are verified through simulated experiments. Based on the results, a general guideline for radar data processing in traffic risk assessment and early warning systems is proposed.

Gender differences in traffic risk behavior among motorcyclists in Indonesian city

Several factors, such as humans, vehicles, roads, and the environment, can cause traffic accidents. However, studies have shown that the crucial problem in traffic accidents is the human factor, in particular motorcycle users, or motorcyclists. This study aimed to investigate the traffic risk behavior of motorcyclists' considering gender as a factor to reduce traffic accidents. The data was obtained through a self-administered questionnaire completed by 155 motorcyclists in Aceh Province, Indonesia, specifically in Banda Aceh City and the neighboring areas. The questionnaire included information about the characteristics of the respondents as well as their perceptions of traffic risks. The data collected were analyzed using Chi-Square test/Fisher's Exact Test, and Odds Ratio (OR). According to the findings, gender had no significant effect on motorcyclists' traffic risk behavior among the eight variables considered (overtaking, speeding, the use of cell phone, turning without signal alert, tailgating, the use of helmet, helmet strapping, and the use of protective cloth). Subsequently, it was observed that not using helmets while riding put men at greater risk than women. The odds ratio of 3.930 implies that men who frequently drive without helmets have a 3.930 times higher risk of traffic accidents than women. The study revealed that, although gender did not affect traffic risk behavior, men were more at risk than women in almost all categories except tailgating. As a result, male motorcyclists are required to be more responsible when driving, and local governments must play a key role in law enforcement to decrease traffic accidents.

Training road users for future mixed automated traffic contexts: a practical framework for creating evidence-based education and awareness schemes

Training and awareness programs are developed in an H2020 Research & Innovation project on safety of road user groups in future automated traffic. With reducing crashes by reducing human error, increasing penetration of AVs is also expected to bring disruptions requiring public education and awareness on new traffic risks, safety technology and behaviours. We combine practices of Knowledge Translation of research with Constructive Alignment in educational design for a flexible framework meeting best practice requirements: knowledge users engaged to benefit from their expertise and networks; results translated as messages tailored to diverse audiences; learning outcomes and strategies are user-centred, evidence-based, targeting higher cognitive functions; iterative evaluation, updates. The framework is proposed to facilitate and enhance planning for uptake and impact of road safety R&I.

Trajectory Planning and Control of Autonomous Vehicles for Static Vehicle Avoidance in Dynamic Traffic Environments

This paper presents a trajectory planning and control algorithm of autonomous vehicles for static traffic agent avoidance in multi vehicle urban environments. In urban autonomous driving, the subject vehicle encounters diverse traffic scenes including lane changing, intersection driving, and illegally parked static vehicle avoidance. Among these, dealing with illegally parked static target vehicle is a major challenge to urban autonomous driving due to large velocity difference between ego and target vehicles and interactions with surrounding vehicles. In order to tackle this problem, we introduce a decision making and motion planning framework for static vehicle avoidance considering both the preceding static vehicles and surrounding vehicles. Among the surrounding vehicles, the set of objects with potential collision risk is selected based on the lane boundaries and road geometry. Then, the driving status of the selected target vehicles are classified as normal driving vehicles or parked vehicles based on their longitudinal speed, lateral position and lateral space occupancy. For the preceding parked vehicles, the motion planner generates lateral and longitudinal evasive motion, by taking side lane traffic flow and risk into account. The desired motion is executed by applying optimized control inputs computed by lateral and longitudinal model predictive controllers. The performance validation of the proposed algorithm has been conducted with actual autonomous test vehicles. The test results confirmed that the proposed algorithm can successfully perform evasive maneuvers on urban roads to ensure safety and mitigate collision risk with the surrounding traffic agents.

Operational Design Domain-Driven Coverage for the Safety Argumentation of Automated Vehicles

Automated Vehicles aim to increase road safety as automated driving systems (ADS) take over the human driving task in the operational design domain (ODD), introducing severe challenges for safety validation. Pure driving over many kilometers to gather enough evidence for a safety argument is not feasible. Scenario-based testing is an approach to overcome this, but challenges like parameter discretization still prevail, hindering safety assurance. This work proposes contributions towards a traceable and efficient safety argumentation for ADS built upon ODD coverage. First, ODD coverage is thoroughly quantified across all scenario levels, assuming distribution functions’ availability for the scenario parameters. Secondly, a sampling method for n-dimensional scenario parameter distributions is proposed. The provided algorithms adapt an initial k-means clustering using pre-defined boundary conditions requiring significantly fewer scenarios. Furthermore, a risk metric for urban intersections is presented for scenario evaluation. The risk metric consists of two parts, scene prediction of traffic participants (TPs) and risk assessment. The scene prediction uses a manoeuvre-based motion model with a data-driven approach towards trajectory prediction, increasing the validity. For the risk assessment, a probabilistic risk prediction for the TPs is performed for each scenario scene. The risk metric shows a reasonable tradeoff between sensitivity and specificity, outperforming time-to-collision. These contributions are exemplarily applied at an intersection using a simplified setup for generating TPs and ego vehicle trajectories. The results indicate that an increased safety argumentation is enabled using the proposed methods alongside a coverage process, facilitating further research.

A trajectory outlier detection method based on variational auto-encoder.

Trajectory outlier detection can identify abnormal phenomena from a large number of trajectory data, which is helpful to discover or predict potential traffic risks. In this work, we proposed a trajectory outlier detection model based on variational auto-encoder. First, the model encodes the trajectory data as parameters of distribution functions based on the statistical characteristics of urban traffic. Then, an auto-encoder network is built and trained. The training goal of the auto-encoder network is to maximize the generation probability of original trajectories when decoding. Once the model training is completed, we can detect the trajectory outlier by the difference between a trajectory and the trajectory generated by the model. The advantage of the proposed model is that it only needs to compute the difference between the original trajectory and the trajectory generated by the model when detecting the trajectory outlier, which greatly reduces the amount of calculation and makes the model very suitable for real-time detection scenarios. In addition, the distance threshold between the abnormal trajectory and the normal trajectory can be set by referring to the proportion of the abnormal trajectory in the training data set, which eliminates the difficulty of setting the threshold manually and makes the model more convenient to be applied in different actual scenes. In terms of effect, the proposed model has achieved more than 95% in accuracy, which is better than the two typical density-based and classification-based detection methods, and also better than the methods based on machine learning in recent years. In terms of efficiency, the model has good convergence in the training phase and the training time increases slowly with the data scale, which is better than or as the same as the comparison methods.

On the Interest of Extending Internet of Behaviors to Internet of Habits: A Preliminary Exploration

We are investigating in this paper the issue of managing driving behaviors. In spite of the intensive research works that have addressed this issue, there is still a need for further investigations concerning the changing magnitude of the behaviors, their impact, as well as their inter-relations. Within this context, we are proposing in this paper a new approach for driving behavior management via the concepts of the Internet of Driving Behaviors (IoDB) and the Internet of the Driving Habits (IoDH). We are particularly aiming to inspect driving behavior scores and identify those repetitive aggressive conducts that are becoming long-lasting habits. In order to prevent and/or heal any driving habits that would increase road traffic risks and crashes, we are proposing the Invest-Prevent-Heal-Accompany process that uses the concepts of Confusion Driving Matrix and Confusion Behavior Matrix to investigate the reasons and causes of traffic homogeneity and driving scores. We explore the interest of the proposed approach using a dataset of 942 trips of 252 drivers collected from a governmental organization at Muscat city, Oman.

Enhancing the feasibility of airport PPP projects with hybrid funding

In transport infrastructure concessions, the sources of revenue to the private partner (or concessionaire) may include (i) the infrastructure users (e.g., landing fees, in the case of airports), (ii) the government (e.g. through availability payments), and (iii) both users and government, which might be called a hybrid concession. An example of the latter is a highway concession where the concessionaire charges tolls to the road users but, because of relatively low revenues, the government agency complements the toll revenue with availability payments.Focusing on airports, this paper summarizes the cases where it may be justified for the government to complement users’ revenues and describes a model developed for the financial assessment of airport concessions involving payments by both the government and airport users, through the collection of several charges. The methodology described in the paper is also used to review the flexibility in new or ongoing airport concessions to mitigate traffic risks, which have been aggravated by the COVID-19 pandemic. The methodology can also be applied to other forms of transport infrastructure.A practical application of the model is demonstrated in the paper, using publicly available information, as well as basic assumptions, to build case studies for the Larnaca and Paphos airports in Cyprus. The model can also be used to carry out sensitivity analyses of the impact of key input parameters on outputs such as the investor's return on equity and annual debt service cover ratio.

Simplified model of children-pedestrian crossing speed at signalized crosswalks

Road accidents are one of the main causes of child mortality, and research results show that children do not have an adequate perception of traffic risks and are prone to insecure and impulsive reactions in traffic situations. The data clearly show that educational and restrictive measures need to be complemented by engineering measures that shape pedestrian transport infrastructure (especially crosswalks) according to the needs of vulnerable traffic users. The influence of children's age, gender, group movement, supervision, distractions and the influence of traffic infrastructure has been widely studied.In articles previously published by the authors, on the basis of field research involving child pedestrians at zebra crosswalks in the vicinity of primary schools in two Croatian and one Italian city, several parameters were identified as significant for the development of the linear regression model for the speed of child pedestrians. Statistical analysis showed that the statistically significant parameters are related to the design of the conflict zone of the transport infrastructure, as well as traffic regulation. As the developed model is considered complex with regard to the number of parameters, a simplified model will be developed in the following of this article. In the first phase, a simplified model with exclusively infrastructural parameters will be developed that would allow a simple application in solving various engineering and planning problems in traffic infrastructure, and its accuracy will be tested. In subsequent phases, models including other main/simple parameters will be developed and tested and the models will be compared. The simplified models thus developed could be used in the early stages of planning by professionals to test or plan pedestrian traffic infrastructure or traffic signals in corridors that are regularly used by children on the basis of easily definable parameters - considering the infrastructure elements and basic data on children.

Driver Profile and Driving Pattern Recognition for Road Safety Assessment: Main Challenges and Future Directions

This study reviews the Artificial Intelligence and Machine Learning approaches developed thus far for driver profile and driving pattern recognition, representing a set of macroscopic and microscopic behaviors respectively, to enhance the understanding of human factors in road safety, and therefore reduce the number of crashes. It provides a definition of the two scientific fields in terms of safety, and identifies the most efficient approaches used regarding methodology, data collection and driving metrics. Results show that K-means and Neural Networks are the most commonly used methodologies for driver profile identification, and Dynamic Time Warping for driving pattern detection. Most studies discovered driver profiles related to aggressiveness, considering mainly speed and acceleration as driving metrics. Based on the gaps and challenges identified, this paper provides a new framework for combining microscopic and macroscopic driving behavior analysis, instead of examining them separately as is the state-of-theart. Such combined results can potentially improve the development of traffic risk models, which could be exploited in applications that monitor drivers in real-time and provide feedback. These models will represent human behavior more accurately, which can eventually lead to the recognition of “optimal” human driving patterns that Automated Vehicles (AV) could ‘mimic’ to become safer.

A novel approach for fine-grained traffic risk characterization and evaluation of urban road intersections

An urban road intersection environment with low traffic risk is an essential component of sustainable cities. Various risk analysis and simulation techniques have been developed to measure and evaluate traffic risks. However, flexible assessment of intersection risk at a fine resolution remains a great challenge. This study proposes a novel approach for characterizing the field distribution of traffic risks at an urban road intersection and assessing traffic risks in multiple dimensions to increase the flexibility of risk assessment. First, we define a new traffic risk metric to quantify the risk of intersection traffic conflict points by combining their frequency and severity. Second, the plume diffusion model is introduced to model the non-linear relationship between traffic conflict points by considering the propagation of the traffic risk in the intersection. In this way, the entire intersection risk distribution map is derived for real-time risk analysis. Finally, a multidimensional hierarchical evaluation system of traffic risk at intersections is designed, including the traffic risk indicators at the intersection, path, and turn mode levels. The results of a case study in Wuhan, China, demonstrate the continuous characterization results of the risk field and provide multidimensional risk assessment index results, which are highly consistent with the cognitive results of manual risk scoring (up to 90% similarity). It also allows for risk assessment in a single path and turn modes through the risk field characterization and reveals the spatiotemporal dynamic risk distribution, thus providing an alternative tool for risk decision-making at urban road intersections.

Mining of risk perception dimensions of Chinese tourists' outbound tourism based on word vector method.

Safety is the lifeline of tourism development. The article's goal is to examin how Chinese tourists perceive risk when travelling aboard. In order to create the initial corpus, this study first uses "outbound tourism"as the key word to crawl the question and answer (hereinafter referred to as "Q & A") data from 4 Chinese online travel operator platforms, then preprocesses the "Q & A" data in Python. Secondly, after being extracted, the feature words are converted into the word vector model using the word vector method based on neural network language model. Finally, the word vectors are clustered and classified. It is found that there are six dimensions of risk perception of Chinese tourists' outbound tourism, namely traffic risk, planning risk, service risk, communication risk, financial risk and functional risk. Important and practical information for government and tourism enterprises is provided to accurately grasp the risk perception of Chinese tourists' outbound tourism and continuously improve the supply of tourism risk information.

Spatiotemporal modeling of traffic risk mapping: A study of urban road networks in Barcelona, Spain

Accidents on the road have always been a major concern in modern society. According to the World Health Organization, globally road traffic collisions are one of the leading and fastest growing causes of disability and death. The present research work is conducted on ten years of traffic accident data in an urban environment to explore and analyze spatial and temporal variation in the accidents and related injuries. The proposed spatiotemporal model can make predictions regarding the number of injuries incurred on individual road segments. Bayesian methodology using Integrated Nested Laplace Approximation (INLA) with Stochastic Partial Differential Equations (SPDE) has been applied to generate a predicted risk map for the entire road network. The current study introduces INLA- SPDE modeling to perform spatiotemporal predictive analysis on selected areas, precisely on road networks instead of traditional continuous regions. Additionally, the result risk maps act as a baseline to identify the safe routes in a spatiotemporal context. The methodology can be adapted and applied to enhanced INLA-SPDE modeling of spatial point processes precisely on road networks.

EVALUATION OF TWO-LANE ROAD SECTIONS IN TERMS OF TRAFFIC RISK USING AN INTEGRATED MCDM MODEL

The impact of geometric characteristics on traffic risk is reflected through identifying conflict points on roads,traffic accidents, and any other unforeseen situation that is inherently hazardous for traffic participants. In order to identify the road sections with the highest risk, it is necessary to consider a number of criteria that affect risk, and conduct extensive empirical research, analysis and data synthesis. This paper evaluates 9 sections of two-lane roads in the territory of Bosnia and Herzegovina (the Republic of Srpska) using an integrated Multi-Criteria Decision-Making (MCDM) model.To determine the significance of 8 criteria for the evaluation of the sections, it was applied a subjective–objective model consisting of 3 methods: (1) CRiteria Importance Through Inter-criteria Correlation (CRITIC), (2) FUll COnsistency Method (FUCOM) and (3) fuzzy PIvot Pairwise RElative Criteria Importance Assessment (PIPRECIA). The aggregation of the criterion values obtained using the methods yielded the final criterion values. Measurement Alternatives and Ranking according to COmpromise Solution (MARCOS) method was used to evaluate the sections and determine their objective diversity. The obtained results identified one location as extremely hazardous by most of analysed input parameters. The section with the highest risk is the Rudanka – Doboj section (A4), which represents a section of the road infrastructure of the 105 road. The validation of the results obtained by applying the integrated MCDM model was performed through an extensive sensitivity analysis. The weights of criteria were observed through initially individual methods implemented in the MARCOS method. Then, a comparative analysis was performed with 6 other MCDM methods and Spearman’s Correlation Coefficient (SCC) was calculated as a statistical indicator of rank correlation in a sensitivity analysis. In addition,the Standard Deviation (STDEV) of the obtained results was determined.

Prescribed anti-glaucoma medication consumption and road traffic crash.

Glaucoma is one of the leading causes of visual impairment worldwide. Influence of visual defects associated with this condition, as well as potential side effects of anti-glaucoma medications on driving may be a relevant traffic safety concern. This study therefore aimed to investigate whether and/or to what extent prescribed anti-glaucoma medicine consumption is associated with increased likelihood of crash risk, and traffic crash responsibility among drivers involved in road traffic crashes. Data from three French national databases were extracted and matched as part of the CESIR (a combination of studies on health and road safety) project. The sample included 201 497 drivers involved in an injurious road crash in France from July 1, 2005 to December 31, 2015, and an age- and sex-matched control group (113 357 drivers) that was randomly drawn from the general population. Exposure to anti-glaucoma medications were compared between responsible and non-responsible drivers involved in a crash and between drivers involved in a crash and people from the control group. The proportion of drivers with prescribed anti-glaucoma medicine markedly increased with age. One type (OR=0.79, 95% CI: 0.72-0.86) and two or more types (OR=0.82, 95% CI: 0.68-0.98) anti-glaucoma medicine prescriptions were less frequent in crash-involved drivers than in controls. One type (OR=0.99, 95% CI: 0.88-1.12) and two or more types (OR=1.04, 95% CI: 0.82-1.33) anti-glaucoma medicine prescriptions were not associated with crash responsibility. Our findings are reassuring as regard to existing guidelines for safe driving for individuals using anti-glaucoma medications. Our results also suggest that driving behavior adaptation is effective mitigating potential traffic crash risks for people diagnosed with glaucoma.

Weibo users and Academia's foci on tourism safety: Implications from institutional differences and digital divide

Tourism safety is essential for tourists and tourism practitioners. This study conducted a bibliometric analysis using VOSviewer and CiteSpace for 2018 articles indexed on the Web of Science (WoS). It also analysed 7293 Weibo posts between 1977 and 2022 using Python, MYSQL, AI sentiment, and Tableau. The first tourism safety publication on WoS appeared in 1977, while the first Weibo microblog dated was dated back to 2011. Compared to the information posted on Weibo, the annual publications about tourism safety on WoS recorded a stable increment. On Web of Science (WoS), the academic staff and universities produced the largest number of tourism safety posts. On the flip side, the most productive organisations on Weibo are government agencies in popular tourism destinations. “Accident”, “medical tourism”, “environment”, “mediating role”, and “hospitality” were important burst nodes in tourism safety on WoS. “Quality”, “accident”, and health-related words were the foci on both Weibo and WoS. On Web of Science, the top 10 most popular keywords of tourism safety-related articles could be classified into two groups: health (“Covid-19”, “restoration”, “pandemics”, “Sars-Cov-2”, “Sars”, “mental health”) and IT terminologies (“big data”, “artificial intelligence”). It has been concluded that “artificial intelligence (AI)” is more likely to be included in the keywords on tourism researched by academia. In contrast, the public may not know about or use AI in the tourism industry. Besides, the top 10 most popular keywords on Weibo related to tourism risks and hazards were drowning and traffic risks and hazards, such as drowning and traffic risks. The digital divide may explain such a difference: the academic circle benefits more from the digital age than laypersons. It may also be the result of institutional differences and information asymmetry.

A Multi-Level Risk Framework for Driving Safety Assessment Based on Vehicle Trajectory

Few existing research studies have explored the relationship of road section level, local area level and vehicle level risks within the highway traffic safety system, which can be important to the formation of an effective risk event prediction. This paper proposes a framework of multi-level risks described by a set of carefully selected or designed indicators. The interrelationship among these latent multi-level risks and their observable indicators are explored based on vehicle trajectory data using the structural equation model (SEM). The results show that there exists significant positive correlation between the latent risk constructs that each have adequate convergent validity, and it is difficult to completely separate the local traffic level risk from both the road section level risk and vehicle level risk. The local and road level indicators are also found to be of more importance when risk prediction time gets earlier based on feature importance scoring of the LightGBM. The proposed conceptual multi-level indicator based latent risk framework generally fits with the observed results and emphasises the importance of including multi-level indicators for risk event prediction in the future.