Analysis of pedestrian accident injury-severities at road junctions and crossings using an advanced random parameter modelling framework: The case of Scotland

Innovative prediction and causal analysis of accident vehicle towing probability using advanced gradient boosting techniques on extensive road traffic scene data

Accurate prediction and causal analysis of road crashes are crucial for improving road safety. One critical indicator of road crash severity is whether the involved vehicles require towing. Despite its importance, limited research has utilized this factor for predicting vehicle towing probability and analyzing its causal factors. This study addresses this gap by predicting the probability of vehicle towing in road crashes based on road scene features and identifying key causal factors. Utilizing the Transportation Injury Mapping System (TIMS) dataset from California, USA, encompassing 12 years, 14 relevant features, and over 2 million road crash records, research team developed a prediction model using advanced gradient boosting techniques. Our model outperforms Random Forest, GBDT, and XGBoost in predictive accuracy. Employing the Shapley Additive Explanation (SHAP) method, researchers elucidate seven key factors influencing towing necessity. These findings introduce a novel predictive approach and offer valuable insights for road crash risk assessment and road safety planning.

Investigating vehicle-vehicle and vehicle-pedestrian crash severity at street intersections with the latent class parameterized correlation bivariate generalized ordered probit

Street intersection crashes often involve two parties: either two vehicles hitting each other (i.e., a vehicle-vehicle crash) or a vehicle colliding with a pedestrian (i.e., a vehicle-pedestrian crash). In such crashes, the severity of injuries can vary considerably between the parties involved. It is necessary to understand the injuries of both parties simultaneously to identify the causality of a vehicle-pedestrian or two-vehicle crash. While the latent class ordinal model has been used in crash severity studies to capture heterogeneity in crash propensity, most of these studies are univariate, which is inappropriate for crashes involving two parties. This study proposes a latent class parameterized correlation bivariate generalized ordered probit (LCp-BGOP) model to examine 32,308 vehicle-vehicle and vehicle-pedestrian crashes at intersections in Taipei City, Taiwan. The model parameterizes thresholds and within-crash correlations of crash severity involving two parties and classifies these crashes into two distinct risk groups: the "Ordinary Crash Severity" (OCS) group and the "High Crash Severity" (HCS) group. The OCS group is mainly two-vehicle crashes involving motorcycles. The HCS group comprises vulnerable road users such as pedestrians and cyclists, mainly in mixed traffic with heavy volumes. The results also show that the effects of party-specific factors contributing to injury severity are greater than those of generic factors. Our study provides invaluable insight into intersection crashes, helping to reduce the severity of injuries in vehicle-vehicle and vehicle-pedestrian crashes.

Investigating pedestrian-vehicle crashes on interstate highways: Applying random parameter binary logit model with heterogeneity in means

In the U.S., the interstate highway system is categorized as a controlled-access or limited-access route, and it is unlawful for pedestrians to enter or cross this type of highway. However, pedestrian-vehicle crashes on the interstate highway system pose a distinctive safety concern. Most of these crashes involve 'unintended pedestrians', drivers who come out of their disabled vehicles, or due to the involvement in previous crashes on the interstate. Because these are not 'typical pedestrians', a separate investigation is required to better understand the pedestrian crash problem on interstate highways and identify the high-risk scenarios. This study explored 531 KABC (K = Fatal, A = Severe, B = Moderate, C = Complaint) pedestrian injury crashes on Louisiana interstate highways during the 2014-2018 period. Pedestrian injury severity was categorized into two levels: FS (fatal/severe) and IN (moderate/complaint). The random parameter binary logit with heterogeneity in means (RPBL-HM) model was utilized to address the unobserved heterogeneity (i.e., variations in the effect of crash contributing factors across the sample population) in the crash data. Some of the factors were found to increase the likelihood of pedestrian's FS injury in crashes on interstate highways, including pedestrian impairment, pedestrian action, weekend, driver aged 35-44 years, and spring season. The interaction of 'pedestrian impairment' and 'weekend' was found significant, suggesting that alcohol-involved pedestrians were more likely to be involved in FS crashes during weekends on the interstate. The obtained results can help the 'unintended pedestrians' about the crash scenarios on the interstate and reduce these unexpected incidents.

Predicting pedestrian-involved crash severity using inception-v3 deep learning model

This research leverages a novel deep learning model, Inception-v3, to predict pedestrian crash severity using data collected over five years (2016-2021) from Louisiana. The final dataset incorporates forty different variables related to pedestrian attributes, environmental conditions, and vehicular specifics. Crash severity was classified into three categories: fatal, injury, and no injury. The Boruta algorithm was applied to determine the importance of variables and investigate contributing factors to pedestrian crash severity, revealing several associated aspects, including pedestrian gender, pedestrian and driver impairment, posted speed limits, alcohol involvement, pedestrian age, visibility obstruction, roadway lighting conditions, and both pedestrian and driver conditions, including distraction and inattentiveness. To address data imbalance, the study employed Random Under Sampling (RUS) and the Synthetic Minority Oversampling Technique (SMOTE). The DeepInsight technique transformed numeric data into images. Subsequently, five crash severity prediction models were developed with Inception-v3, considering various scenarios, including original, under-sampled, over-sampled, a combination of under and over-sampled data, and the top twenty-five important variables. Results indicated that the model applying both over and under sampling outperforms models based on other data balancing techniques in terms of several performance metrics, including accuracy, sensitivity, precision, specificity, false negative ratio (FNR), false positive ratio (FPR), and F1-score. This model achieved prediction accuracies of 93.5%, 77.5%, and 85.9% for fatal, injury, and no injury categories, respectively. Additionally, comparative analysis based on several performance metrics and McNemar's tests demonstrated that the predictive performance of the Inception-v3 deep learning model is statistically superior compared to traditional machine learning and statistical models. The insights from this research can be effectively harnessed by safety professionals, emergency service providers, traffic management centers, and vehicle manufacturers to enhance their safety measures and applications.

Evasive actions to prevent pedestrian collisions in varying space/time contexts in diverse urban and non-urban areas

This study aims to identify driver-safe evasive actions associated with pedestrian crash risk in diverse urban and non-urban areas. The research focuses on the integration of quantitative methods and granular naturalistic data to examine the impacts of different driving contexts on transportation system performance, safety, and reliability. The data is derived from real-life driving encounters between pedestrians and drivers in various settings, including urban areas (UAs), suburban areas (SUAs), marked crossing areas (MCAs), and unmarked crossing areas (UMCAs). By determining critical thresholds of spatial/temporal proximity-based safety surrogate techniques, vehicle-pedestrian conflicts are clustered through a K-means algorithm into different risk levels based on drivers' evasive actions in different areas. The results of the data analysis indicate that changing lanes is the key evasive action employed by drivers to avoid pedestrian crashes in SUAs and UMCAs, while in UAs and MCAs, drivers rely on soft evasive actions, such as deceleration. Moreover, critical thresholds for several Safety Surrogate Measures (SSMs) reveal similar conflict patterns between SUAs and UMCAs, as well as between UAs and MCAs. Furthermore, this study develops and delivers a pseudo-code algorithm that utilizes the critical thresholds of SSMs to provide tangible guidance on the appropriate evasive actions for drivers in different space/time contexts, aiming to prevent collisions with pedestrians. The developed research methodology as well as the outputs of this study could be potentially useful for the development of a driver support and assistance system in the future.

A hybrid approach of random forest and random parameters logit model of injury severity modeling of vulnerable road users involved crashes

Vulnerable road users (VRUs) involved crashes are a major road safety concern due to the high likelihood of fatal and severe injury. The use of data-driven methods and heterogeneity models separately have limitations in crash data analysis. This study develops a hybrid approach of Random Forest based SHAP algorithm (RF-SHAP) and random parameters logit modeling framework to explore significant factors and identify the underlying interaction effects on injury severity of VRUs-involved crashes in Shenyang (China) from 2015 to 2017. The results show that the hybrid approach can uncover more underlying causality, which not only quantifies the impact of individual factors on injury severity, but also finds the interaction effects between the factors with random parameters and fixed parameters. Seven factors are found to have significant effect on crash injury severity. Two factors, including primary roads and rural areas produce random parameters. The interaction effects reveal interesting combination features. For example, even though rural areas and primary roads increase the likelihood of fatal crash occurrence individually, the interaction effect of the two factors decreases the likelihood of being fatal. The findings form the foundation for developing safety countermeasures targeted at specific crash groups for reducing fatalities in future crashes.

An Exploratory Assessment of Self-Reported Satisfaction with Infrastructure and Out-of-Home Activities for People with Vision Impairments

PURPOSE

The purpose of this study is to assess the satisfaction levels of people with VI with regard to infrastructure and outdoor activities. Furthermore, this study aims to develop an assessment model for the levels of difficulty in using public transport.

METHODS

Participants in a standardized survey questionnaire included 74 participants with VI. Three assessment-ordered probit models were estimated based on self-reported responses.

RESULTS

Estimation results revealed that the use of public transport is extremely difficult for 83.47% of older participants. In addition, 84.2% of people with albinism have extreme difficulty using public transport. Furthermore, 53.98% of people with restricted horizontal and vertical fields face extreme difficulty using public transport. There was dissatisfaction with outdoor activities among 97.40% of people with macular disease. The results show that 51.70% of people with normal or near-normal horizontal visual fields and restricted vertical planes are satisfied with their level of outdoor activity while 72.65% of people with retinal diseases expressed dissatisfaction with the existing infrastructure.

CONCLUSION

This study revealed that the experiences of people with VI are heterogeneous and depend on their eye condition, access to assistive technology, and socioeconomic characteristics. Results clearly show evidence of heterogeneity among individuals with VI. The combination of horizontal and vertical restrictions yields random parameters, underscoring the heterogeneous experiences of people with VI, influenced by their eye condition and access to assistive devices. Our results have important implications for developing targeted interventions to enhance the mobility of people with VI.

Sex, age, role and geographic differences in traumatic spinal fractures caused by motor vehicle collisions: a multicentre retrospective study

To investigate the sex, age, role and geographic differences in traumatic spinal fractures (TSFs) caused by motor vehicle collisions (MVCs) in adults (≥ 18 years old). This was a multicentre retrospective observational study. In total, 798 patients with TSFs caused by MVCs admitted to our hospitals from January 2013 to December 2019 were enrolled. The patterns were summarized with respect to different sexes (male and female), age group (18-60 and ≥ 60), role (driver, passenger and pedestrian) and geographic location (Chongqing and Shenyang). Significant differences in distribution related to district (p = 0.018), role (p < 0.01), motorcycle (p = 0.011), battery electric vehicle (p = 0.045), bicycle (p = 0.027), coma after injury (p = 0.002), pelvic fracture (p = 0.021), craniocerebral injury (p = 0.008) and fracture location (p < 0.01) were observed between the male and female groups. Significant differences in distribution related to district (p < 0.01), role (p < 0.01), car (p = 0.013), coma after injury (p = 0.003), lower limb fracture (p = 0.016), fracture location (p = 0.001) and spinal cord injury (p < 0.01) were observed between the young adult and elderly groups. Significant differences in distribution related to sex ratio (p  <  0.01), age (p  <  0.01), district (p  <  0.01), most vehicles involved (P  <  0.01), lower limb fracture (p  <  0.01), pelvic fracture (p < 0.01), fracture location (p < 0.01), complications (p < 0.01), and spinal cord injury (p < 0.01) were observed between the three different groups of pedestrian, passenger, and driver. Significant differences in distribution related to sex ratio (p = 0.018), age (p < 0.01), role (p < 0.01), most vehicles involved (p < 0.01), coma after injury (p = 0.030), LLF (P = 0.002), pelvic fracture (p < 0.01), craniocerebral injury (p = 0.011), intrathoracic injury (p < 0.01), intra-abdominal injury (p < 0.01), complications (p = 0.033) and spinal cord injury (p < 0.01) were observed between the Chongqing and Shenyang groups. This study demonstrates the age-, gender-, role- and geographic-specific clinical characteristics of TSFs resulting from MVCs and reveals a significant relationship between different ages, sexes, roles, geographic locations and associated injuries, complications and spinal cord injuries.

Assessing School Travel Safety in Scotland: An Empirical Analysis of Injury Severities for Accidents in the School Commute

School travel has been a significant source of safety concerns for children, parents, and public authorities. It will continue to be a source of concerns as long as severe accidents continue to emerge during pupils’ commute to school. This study provides an empirical analysis of the factors influencing the injury severities of the accidents that occurred on trips to or from school in Scotland. Using 9-year data from the STATS19 public database, random parameter binary logit models with allowances for heterogeneity in the means were estimated in order to investigate injury severities in urban and rural areas. The results suggested that factors such as the road type, lighting conditions, vehicle type, and age of the driver or casualty constitute the common determinants of injury severities in both urban and rural areas. Single carriageways and vehicles running on heavy oil engines were found to induce opposite effects in urban and rural areas, whereas the involvement of a passenger car in the accident decomposed various layers of unobserved heterogeneity for both area types. The findings of this study can inform future policy interventions with a focus on traffic calming in the proximity of schools.