With the development of big data technology and the improvement of deep learning technology, data-driven and machine learning application have been widely employed. By adopting the data-driven machine learning method, with the help of clustering processing of data sets, a recurrent neural network (RNN) model based on Keras framework is proposed to predict the injury severity in urban areas. First, with crash data from 2014 to 2017 in Nevada, OPTICS clustering algorithm is employed to extract the crash injury in Las Vegas. Next, by virtue of Keras' high efficiency and strong scalability, the parameters of loss function, activation function and optimizer of the deep learning model are determined to realize the training of the model and the visualization of the training results, and the RNN model is constructed. Finally, on the basis of training and testing data, the model can predict the injury severity with high accuracy and high training speed. The results provide an alternative and some potential insights on the injury severity prediction.
Despite statistics indicating that China has the world's largest taxi industry, there exists limited research about the relationship between workplace health hazards and taxi driver occupational crashes. In this paper, a cross-sectional survey of taxi drivers in four typical Chinese cities was conducted, and data on their self-reported job stress, health status, and daily risky driving behaviours, together with crash involvement experience in the two years before the survey was collected. Three hypotheses were then developed, and they were verified via multivariate analysis of variance (MANOVA) that the seriousness of drivers' health problems and the frequency of their daily risky driving behaviours could be the accurate predictor of their crash risk of taxi drivers. These factors were subsequently substituted in a bivariate negative binomial (BNB) distribution model to determine the joint rate of at-fault taxi drivers' involvement in property-damage-only (PDO) and personal-injury (PI) crashes. The results offer some useful advice for policy development to decrease and prevent professional taxi drivers from causing severe traffic crashes.
This research examines the injury severity of single-vehicle large-truck crashes in Florida while exploring the role of heterogeneity. A random parameter ordered logit (RPOL) model was applied to 27,505 single-vehicle large-truck crashes from 2007 to 2016 in Florida, and the contributing factors were identified. Random parameters and interaction effects were introduced to the model to determine the heterogeneity and its potential sources. The results suggested that driving speed of 76-120 mph and defective tires were the most influential factors in crash injury severity, increasing the probability of severe crashes. Regarding truckers' attributes, asleep or fatigued conditions and driving under the influence were correlated with a higher possibility of severe crashes. Interestingly, the results showed that truckers from outside the state of Florida were less likely to cause severe single-vehicle large-truck crashes compared to their Floridian counterparts. Y-intersections were also found as a high-risk location for single-vehicle large-truck crashes, leading to more severe outcomes. Regarding heterogeneity, the results indicated that the impacts of driving speed (26-50 mph) and light condition (dark - not lighted) significantly varied among the observations, and these variations could be attributed to driver action, vision obstruction, driver distraction, roadway type and roadway alignment.
The current work presented a comparative analysis of traffic demand and safety skills before and after control measures during the COVID-19 epidemic, acquired time-series change data curves, and constructed a prediction model after determining the trend of traffic demand over time. From a data analysis perspective, the paper draws some interesting conclusions about long span, coarse sampling studies. In terms of the study population, the paper did focus on the specificity of the global epidemic. Kuwait was selected as a case study. Traffic demand analysis was conducted using a Structural Equation Model (SEM), Auto-Regressive Integrated Moving Average (ARIMA), and safety skills questionnaire along with flow charts and demographic variables. These methods were utilized to study the impact of COVID-19 on traffic congestion and safety skills as well as to forecast the future traffic volumes. Results showed that traffic congestion had a significant reduction during COVID-19 as a result of the preventive safety measures taken to control the spread of the virus. Such reduced traffic volume was associated with a decrease in traffic violations and an increase in the safety skills and PM skills of drivers.
In China, bicycles are a popular mode of transportation for senior citizens. A disproportionate number of traffic-related fatalities and injuries involve cyclists. The violation of cycling laws is a significant cause of cyclist crashes. Few studies have analyzed the cycling violation behaviour of seniors. Therefore, it is essential to examine the factors that influence older individuals' intention to engage in cycling violation behaviours. In this study, the effects of social-demographic characteristics, the exogenous constructs in the health belief model (HBM), and the theory of planned behaviour (TPB) on senior cyclists' violation intention were investigated using hierarchical regression analysis. Interviews were conducted with older cyclists in urban areas of Wuhan City, all above 60 years of age. The results showed that very little variance in behavioural intention could be explained by social-demographic factors. The TPB has a significantly greater capacity than the HBM to explain variance in behavioural intention. Perceived susceptibility, perceived benefit, cues to action, subjective norm and attitude significantly impacted behavioural intention, whereas perceived severity, perceived barrier and self-efficacy did not.
Inadequate regional road safety studies have been conducted in developing countries like Iran. Regarding regional road safety indices (RSIs), a significant disparity between Iranian provinces was observed. Thus, it was aimed to evaluate the regional RSIs in Iran and identify their influencing factors and potential hot spots. Data on regional road crashes, fatalities, demographics, transportation, health institutions, economics, education, and fuel consumption rates were collected. The association between the variables was evaluated using correlation analysis. Using Moran's I and local Moran indices, provinces' spatial distributions were evaluated. Hot spot analysis was used to identify factors influencing RSIs. Significant correlations between the variables were detected. A vast local cluster in terms of fatality per injury (as a crash severity index) was identified in the country's southeast. The distribution patterns of provinces in terms of seven RSIs were cluster-like. Variable groups, including road length, demographic, income, education, and geographic, influence RSIs in hot or cold spot regions. Crashes were severe in underdeveloped and remote provinces. Increasing income and education levels make it possible to reduce crash severity indices in this country. A positive Moran's I index does not guarantee the existence of significant local cluster cores in a country.
Drivers traversing the horizontal curves are expected to complete the deceleration manoeuvre on the tangent and transition curve and maintain a constant speed upon reaching the curve. However, this may not be true for the horizontal curves constituting a two-lane undivided rural highway passing through mountainous terrain. The objective of this study is to investigate the speed variability on a two-lane rural highway passing through mountainous terrain and to identify its determinants. The continuous speed profiles of vehicles traversing the curves were extracted using the video image processing technique. Individual speed profiles, as well as the operating speed profiles obtained through quantile regression, indicate a significant speed variability on the horizontal curve. Speed variability on the curve was modelled in terms of the 85th percentile of maximum speed difference (MaxΔ85V) using the Robust Weighted Least Square (RWLS) Method. The findings indicate that the curvature change rate, length of the curve and the speed at the point of curvature affect the maximum speed difference on a curve. The findings also suggest that the operating speed estimated based on the spot speed data collected at the curve centre might lead to erroneous estimation of design and operating speed consistencies.
A variety of road hazard perception training programmes have been proposed recently, based on the assumption that these skills contribute to lower crash rates across different countries. However, the long-term effectiveness of suggested programmes has been under-investigated. The main objective of this study is to explore the long-term effectiveness of online hazard perception training for experienced drivers and examine the moderating role of driving self-efficacy. Fifty-six experienced drivers (21 males and 35 females) were assigned to the experimental (n = 31) or the control (n = 25) group. The experimental group received two 45 min session interventions; the control group received no intervention. The effectiveness of the programme was tested by the change in scores of Lithuanian hazard prediction test (HPT) LHP12 that was conducted before training (pre-test), immediately after training (post-test) and six months after training (follow-up). The twelve-item Adelaide Driving Self-Efficacy Scale (ADSES; George et al., 2007) was used to measure self-reported driving self-efficacy at the pre-test. The results revealed a significant increase in hazard prediction scores immediately after training, but the short-term effect of training decayed at follow-up. Experienced drivers with higher self-efficacy developed better hazard prediction skills during training. The results confirmed short-term effectiveness of the programme.
The number of deaths due to road accident is increasing day by day and has become an alarming global problem over the decades. India, with her rising motorization is no stranger to this global catastrophe. In this paper two relatively simple yet powerful and versatile techniques for forecasting time series data, autoregressive integrated moving average method (ARIMA) and exponential smoothing method are used to forecast the number of deaths due to road accidents in India from the year 2022-2031. The results based on the two methods are compared and it is found that they are in sync with each other and pre-existing literature. Furthermore, this is a unique attempt to use two time series analysis techniques on the same data and carry out a comparative analysis. The data was collected from the annual report of Ministry of Road Transport and Highways, India (2020) and Accidental Deaths & Suicides in India (ADSI) Report of National Crime Record Bureau (2021). After examining all the probable models, it is observed that ARIMA (2, 2, 2) model and exponential smoothing (M, A, N) model are suitable for the given data. Amongst the two, ARIMA (2, 2, 2) model has a lower AIC and BIC value. Thus, this comes out to be the best model as per our model selection criterion. Further, the study also reveals an upward trend of number of road accidental deaths for the upcoming 10 years in India.