IATSS Research最新文献

This study aims to examine the differentiated impacts of frontage-lane, main-lane, and standard-lane motorcycle collisions on rider injury severity in Thailand. Using comprehensive crash data from 2016 to 2019, a random parameters model accommodating heterogeneity in means and variances is applied. Predictive comparisons between out-of-sample and within-sample predictions uncover differences between collision location sub-models. For frontage-lane collisions, severe or fatal injuries are positively associated with male riders, truck involvement, and speeding. In main-lane collisions, increased severe/fatal injury risk is associated with nighttime, holidays, weekends, single-motorcycle crashes, and collisions with trucks or pickups. In standard-lane collisions, positive severe/fatal injury associations include male riders, nighttime, non-peak hours, speeding, truck/pickup involvement, and pillion presence. Predictive comparisons show reorienting crashes to the frontage lane could reduce fatal injury probability by 0.1199 and 0.2233 versus the main and standard lanes respectively, preventing many fatalities. This research underscores accounting for unobserved heterogeneity and predictive simulation insights. The findings inform effective countermeasures and aid safety professionals, instructors, policymakers, law enforcement, and designers in motorcycle safety efforts.

Overtaking is one of the most complex maneuvers on the two-lane two-way (TLTW) roads, where the follower vehicles use the opposing lane to bypass the leader slower vehicles. Overtaking becomes more riskier with the presence of oncoming vehicles from the opposite direction. This paper presents the development of an overtaking model for the Egyptian TLTW roads under mixed traffic conditions. About 20-h of videotaped data from 6 different TLTW roads, located in the Delta region in Egypt are studied to develop the proposed model. This model considers key factors affecting model performance that are not considered in most international state-of-the-art models. These factors include (a) acceleration of the follower vehicle (FV) during all phases of overtaking, (b) deceleration of the leader vehicle (LV), (c) length of the LV, and (d) speed change of the LV during overtaking. The proposed overtaking model is divided into two parts. The first part includes two conditions as criteria for acceptance/rejection decision of overtaking while the second part computes the overtaking duration and the corresponding distance for successful overtakes. The model is calibrated for 4 sites and validated on the remaining 2 sites. The results show that the proposed overtaking model matches closely the observed accepted and rejected overtakes by 96.45% and 95.90%, respectively. Furthermore, the results showed that the proposed model was consistent with the observed overtaking distances and times. In addition, the results are compared with other international models such as the Ghods, Tang, TWO-PAS, and Cirianni models.

Vehicle-pedestrian hit-and-run crashes have grave consequences as the pedestrian casualty is left at the crash scene, leading to delays in emergency response. Research indicates that prevailing weather conditions are pivotal in influencing these crashes. Nonetheless, there exists a substantial gap in understanding the diverse associations of contributing factors that impact the severity of pedestrian injuries in vehicle-pedestrian hit-and-run crashes under varying weather conditions. This paper extensively examines the intricate patterns of critical factor associations influencing the pedestrian casualty injury severity in vehicle-pedestrian hit-and-run collisions in Ghana, differentiating between clear and inclement weather conditions. Leveraging historical crash data, this research employs the robust Association Rule Mining tool to unveil the relationships between key crash factors and their outcomes. The association rules extracted from the data reveal that specific factors consistently influence injury outcomes in vehicle-pedestrian hit-and-run crashes, regardless of weather conditions. However, there are distinct patterns of factors contributing to fatal and severe injuries in these crashes that vary depending on the prevailing weather conditions. Hit-and-run crashes resulting in fatal and severe pedestrian injuries are primarily associated with males and are notably correlated with factors like pedestrian crossings, late and early morning hours, the absence of traffic control measures on median-separated motorways, and good urban roads, regardless of prevailing weather conditions. During clear weather conditions, fatal/severe injuries tend to occur in daylight conditions, primarily on roadways devoid of shoulders, and involve older pedestrians crossing the road. Conversely, for inclement weather conditions, these crashes are more prone to occur during nighttime hours on unlit roads, often involving younger pedestrians walking along the road's edge. The reduction of fatal/severe injury hit-and-run crashes involving vehicles and pedestrians can be effectively tackled by eliminating or improving the critical factors pinpointed in the extracted rules. The study underscores the urgency of policy recommendations to mitigate fatalities resulting from vehicle-pedestrian hit-and-run crashes, emphasising the need for stricter legislation, heightened enforcement measures, increased awareness of pedestrian safety, enhanced driver training, and infrastructure upgrades. By acting on these insights, a significant stride can be made in curbing the alarming rates of vehicle-pedestrian hit-and-run fatalities.

Run-Off-Road (ROR) crashes are frequent and pose a significant risk of injury and fatality. Given the complexity of their mechanisms and the interaction of multiple factors, this study aims to comprehensively investigate the factors influencing the severity of ROR crashes, which have been understudied. Furthermore, addressing the current methodological challenge in machine learning (ML) crash modeling, this study proposes an approach to tackle unobserved heterogeneity in ML. An interpretable ML joint with prior latent class clustering is implemented. The significant risk factors and interactions are interpreted using SHAP (SHapley Additive exPlanations) method across clusters. This study utilizes ROR crash records, traffic, and geometric data from main suburban freeways in Iran collected over a 5-year period. The key interacting factors associated with severe ROR during adverse weather (cluster 1) are: co-occurrence of low congestion and higher speed variation; low congestion, nighttime darkness and rollover; roadway departure by buses and mini-buses and rollover occurrence; vehicle departure and collision with fixed objects. Moreover, the critical interactions for nighttime condition (cluster 2) are: curve sections combined with longitudinal slope; inside shoulder width <1.5 m and a hit median concrete NewJersey barrier. The risky interactions for crashes occurred in curves (cluster 3) are: departing in two-lane sections in low congestion conditions; vehicles collisions with the median concrete NewJersey barriers. The findings of this study enhance comprehension of the significant effects of interactions under various conditions, offering valuable insights for policymakers. Additionally, recommendations are offered to mitigate the risk of severe ROR crashes.

As part of a research project of the International Association for Traffic and Safety Sciences (IATSS), a series of surveys were conducted on the behavior of cyclists and the countermeasures. This paper outlines the mechanisms of bicycle accidents in Japan and their countermeasures. According to accident statistics, the two target groups for the prevention of bicycle accidents include the child population of elementary, junior high, and high school students, and the older population aged 65 years and over. Questionnaires and behavioral observation surveys of junior high school students revealed tendencies for students to violate rules, such as not stopping temporarily, not searching left and right at crossings, riding alongside, and looking aside when riding. A field experiment with junior high school students using gyro sensors revealed that their level of searching for safety was extremely low. Furthermore, clear differences existed in the manner in which each individual searched. Based on these results, several educational and practical activities were performed. In addition, a study was conducted with older cyclists at a driving school to measure behavioral characteristics, such as searching behavior at crossings and riding (walking) position. The license status factor significantly impacted the safety aspects of the behavior of older cyclists, with unlicensed cyclists lacking safety searching behavior when passing through intersections and crossing the road and having lower riding position scores.

The first edition of the Highway Safety Manual (HSM) contains a simplistic version of the crash prediction model for two-way stop-controlled intersections (TWSC) on rural two-lane highways. This model considers AADT on major and minor roads, with the base conditions defined as no intersection skewness, no turning lanes, and no lighting. A crash modification factor (CMF) will be applied if an intersection has conditions different from the base condition. However, the HSM model does not take account of curvature. It is well known that curved TWSC intersections are less safe than non-curved ones, particularly on rural two-lane roadways. This paper presents the development of crash prediction models incorporating intersection geometrics for TWSC intersections on rural two-lane highways in Louisiana. Then, it compares the results from the developed model with the calibrated HSM model. The negative binomial model was used with 5126 TWSC intersections verified one by one, including both three- and four-leg intersections from all parishes (counties). The estimation results indicate that AADT, curve radius, and intersection skewness angle significantly impact expected crash frequency for both three- and four-leg intersections. This research utilizes cumulative residual plots, mean absolute error, and root mean square error for comparative analysis of HSM models, HSM models with calibration, and Louisiana-specific models. The results show that Louisiana-specific SPFs outperformed the calibrated SPFs with superior reliability. Calibration factors of 0.58 for three-leg intersections and 0.46 for four-leg intersections are estimated, suggesting that the original HSM model overpredicts crashes in Louisiana.

Determining an appropriate segment length for highway safety evaluations in low- and middle-income countries (LMICs) poses a significant challenge. This study aims to address this issue by recommending a suitable segment length for such evaluations in India, using a 167 km intercity expressway as a case study. We employed negative binomial (NB) models on datasets segmented from 100 m to 1000 m with 100 m increments. Our findings strongly suggest that segment lengths from 300 m to 700-m suit various safety assessments. However, the study reveals that parameter estimates vary significantly with both segment length and sample size. This highlights the sensitivity of parameters to data aggregation and sample size across different segment lengths, making it difficult to identify a single optimal length. Therefore, we propose selecting the segment length and segmentation approach based on specific local conditions, highway context, data availability and quality. The methodology presented here can guide policymakers in LMICs to make informed choices regarding segment length for safety evaluations, including blackspot identification and treatment on their highways.

The main goal of this study is to investigate the impact of COVID-19 on road crashes in Thailand using time series and interrupted time series analysis. To achieve the goal, road crash data from the Department of Highway (DOH), which includes total crashes, single vehicle crashes, fatalities, fatal crashes, speeding crashes, and drunk driving crashes, was obtained to conduct Seasonal Autoregressive Integrated Moving Average (SARIMA) time series models and Interrupted Time Series (ITS) models. SARIMA models were applied to forecast the number of crashes in the absence of COVID-19 then compare them to the observed values to identify the difference. The impact of a policy change aimed at addressing the spread of COVID-19 was assessed using ITS models on a time series accident dataset. The goal was to ascertain if the intervention had a meaningful and causative impact on the outcome. The result showed that the first wave of COVID-19 caused a significant reduction in all road crash indicators instead of skyrocketing to a peak. After releasing the lockdown measures from the first wave of spreading, an increase was found in all of the crash indicators as well. However, the third wave of COVID-19, which lasted longest for nearly 7 months, also caused a decrease in the number of crashes, but not as much as the first wave of the outbreak. Moreover, the result from the interrupted time series also revealed that curfews and the closure of entertainment places are associated with a significant decrease in the number of speeding crashes and drunk driving crashes from 10 p.m. to 4 a.m., respectively. It can be observed that the COVID-19 countermeasures, such as curfews and bans on the sales of alcoholic beverages, led to a drop in the number of speeding and drunk driving crashes.

Herein, we propose a “Comprehensive Road Safety Management for School Routes” to realize “Vision Zero for School Routes”, launched by International Association of Traffic and Safety Sciences (IATSS). Specifically, we discuss Niigata City's “Vision Zero” project rolled out at Hiyoriyama Elementary School, which aims to eliminate accidents resulting in deaths of or serious injuries to children. Hiyoriyama Elementary School was to be opened by merging four elementary schools in Chuo-ku, Niigata City, and traffic safety measures were discussed with the participation of local residents to ensure the safety of the new route to the school, which would bring in many children from a wide range of localities. In a workshop attended by road administrators, police, school officials, and local residents, issues were identified, a traffic survey was conducted to understand the current situation, and a traffic simulation was conducted to verify concerns that might arise subsequent to the implementation of traffic control measures. The traffic in front of the elementary school was restricted during school hours; furthermore, to ensure compliance with these restrictions, rising bollards were installed for the first time along a school route in Japan. We also clarified the long-term effects of the rising bollards installed on a public road in Japan for the first time in an IATSS project. We hope that this report contributes to the expansion of Vision Zero, which secures the safety of schoolchildren.

It is difficult to accurately predict when and where natural disasters such as major earthquakes will occur, and it is difficult to establish and permanently maintain a large-scale disaster response system for the entire country in order to realize a sustainable, safe, and secure society. Therefore, the key point is how to quickly and efficiently collect on-site information and establish evacuation, rescue, and support systems in the affected areas when a disaster occurs.

The authors launched a research project in April 2012, following the IATSS earthquake disaster investigation project in 2011, immediately after the Great East Japan Earthquake, to investigate the use of probe information obtained from vehicle-mounted sensors to assess the on-site situation in the initial response to such disasters. In the project, a prototype of a system that uses a probe function mounted on a motorcycle to assess road surface damage and detect the presence or absence of people in need of rescue while driving was built and its performance was checked, and the effectiveness of the proposed system was found. After the completion of the research project, the author has spent about 10 years trying to improve the accuracy of the system's detection of minute road surface wear and tear, which can be used for road management, as well as to verify the social acceptability of the system through long-term experiments in social implementation.

In this paper, the author summarizes the research trends to date, reviews the progress and results, and reports the results of discussions on the role of vehicles equipped with the probe function for the implementation and diffusion of sustainable disaster mitigation and disaster prevention systems in the future.