Traffic Injury Prevention最新文献

{"title":"Vehicle blind zones and pedestrian safety at intersections: identifying high-risk scenarios using mathematical simulations.","authors":"Sushant R Jagtap, Jessica S Jermakian","doi":"10.1080/15389588.2025.2520915","DOIUrl":"https://doi.org/10.1080/15389588.2025.2520915","url":null,"abstract":"<p><strong>Objective: </strong>This study presents a novel simulation-based methodology to systematically assess how vehicle blind zones, intersection geometry, and the speed and direction of both vehicles and pedestrians interact to influence pedestrian visibility at intersections. Focusing on a diverse set of vehicle designs and intersection scenarios, the framework identifies high-risk situations where blind zones may critically impair a driver's ability to detect and respond to crossing pedestrians.</p><p><strong>Methods: </strong>Blind zone data from 20 vehicles representing a range of classes were used. A custom simulation code modeled interactions between crossing pedestrians and vehicle blind zones in four intersection geometries: baseline, narrow lane, wide lane, and offset crosswalk. For each geometry, vehicle speed (4.2 m/s or 15 km/h, 5.6 m/s or 20 km/h, 6.9 m/s or 25 km/h), vehicle trajectory (straight, left, right), and turning trajectory (sharp, average, wide) were simulated. Crossing pedestrians were modeled for two travel directions (from the driver-side and passenger-side A-pillar) at three speeds (0.8, 1.2, 1.8 m/s). Output metrics included time the pedestrian spends in the blind zone, percentage of crossing time in the blind zone, and available driver reaction time once the pedestrian is visible to the driver.</p><p><strong>Results: </strong>Vehicle blind zone areas ranged from 242 to 478 m². Pedestrians were predominantly obscured during left-turn maneuvers, especially when approaching from the driver-side A-pillar, resulting in an average time in the blind zone of 1.7 s compared with 0.8 s for passenger-side approaches. Lower vehicle speeds combined with fast pedestrian speeds further increased this duration, while higher speeds reduced the available driver reaction time. Intersection geometry and turning radius also affected these metrics. Regression analyses indicated that although blind zone area did not predict total obscuration time, it was significantly associated with delayed pedestrian entry/exit and longer available driver reaction time.</p><p><strong>Conclusions: </strong>This study used a novel simulation methodology to demonstrate that vehicle blind zones can critically affect pedestrian visibility, particularly in left-turn scenarios involving driver-side pedestrian approaches. While blind zone area is an important metric, finer design features (e.g., A-pillar dimensions) and dynamic factors such as vehicle speed, turning behavior, and intersection layout play a crucial role in determining pedestrian visibility.</p>","PeriodicalId":54422,"journal":{"name":"Traffic Injury Prevention","volume":" ","pages":"1-8"},"PeriodicalIF":1.6,"publicationDate":"2025-07-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144664077","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Repositioning forward-leaning booster-seated child occupants in frontal-oblique low-acceleration impacts.","authors":"Valentina Graci, Madeline Griffith, Thomas Seacrist","doi":"10.1080/15389588.2025.2526611","DOIUrl":"https://doi.org/10.1080/15389588.2025.2526611","url":null,"abstract":"<p><strong>Objective: </strong>Recently, pre-pretensioner equipped seatbelts (PPT) were found to be effective in repositioning forward-leaning adults. The aim of this study was to evaluate the influence of the PPT on booster-seated forward-leaning children during sled-simulated frontal oblique low-acceleration impacts.</p><p><strong>Methods: </strong>Eight child volunteers (mean: 6.6 ± 0.5 years, four males, four females; weight 24.2 ± 2.1 kg; seated height 64.4 ± 3.9 cm) were seated on a low-acceleration impact sled in a high-back belt-positioning booster seat (BPB) on a rear vehicle bench seat restrained with a three-point seatbelt with a PPT. Volunteers were exposed to a far-side impact 30° from frontal (peak acceleration 1 g, duration 265 ms). Four randomized testing conditions were examined and repeated twice: seatbelt conditions with PPT (85 N at 250 ms prior sled onset) and without PPT; and a standard and a forward-leaning posture. A 3D motion capture system captured participants' kinematics. Repeated measure three-way ANOVAs tested the effect of seatbelt condition, posture, and repetition on head and trunk peak forward displacement. All displacements were measured relative to the initial position.</p><p><strong>Results: </strong>The PPT reduced peak forward head displacements in both standard (PPT 103 ± 21 mm vs no-PPT 165 ± 19 mm) and forward-leaning (PPT 42 ± 37 mm vs no-PPT 138 ± 25 mm) postures (<i>p</i> < .001). In the forward-leaning posture with the PPT, the maximum trunk displacement was the initial position: the peak trunk displacement during the sled pulse was less forward than the initial trunk position. Overall, the peak forward trunk displacement was greater in the standard posture (50 ± 37 mm) than in the forward-leaning posture (29 ± 45 mm) (<i>p</i> < .001) and was greater without the PPT (79 ± 13 mm) than with the PPT (1 ± 18 mm). Reduced values with the PPT were also found in the lateral head and trunk displacements (<i>p</i> < .005) regardless of initial posture.</p><p><strong>Conclusions: </strong>The PPT reduced head and trunk displacements in forward-leaning postures and prevented the trunk from moving further forward from the initial forward-leaning position. A PPT, with a relatively low force of 85 N, has the potential to reduce out-of-position postures in booster-seated children in nonstandard seating positions.</p>","PeriodicalId":54422,"journal":{"name":"Traffic Injury Prevention","volume":" ","pages":"1-9"},"PeriodicalIF":1.6,"publicationDate":"2025-07-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144664075","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Trends in road traffic injury mortality and age-period-cohort analysis among children and adolescents in China and globally.","authors":"Wenzhe Du, Hui Fan, Hezhou Li, Changfu Hao","doi":"10.1080/15389588.2025.2528147","DOIUrl":"https://doi.org/10.1080/15389588.2025.2528147","url":null,"abstract":"<p><strong>Objectives: </strong>Road traffic injuries are a leading cause of death among children and adolescents under 19. Despite a decline over the past 30 years, mortality rates remain high. This study analyzes road injury mortality rates trends among children and adolescents aged 0-19 in China and globally from 1990 to 2021.</p><p><strong>Methods: </strong>Using the Global Burden of Disease (GBD) 2021 database, we gathered age-standardized and crude mortality rates for road injuries among children and adolescents in China and worldwide from 1990 to 2021. Joinpoint regression models were used to describe trends, while age-period-cohort (APC) models and Estimated Annual Percentage Change (EAPC) analyses evaluated mortality trends and changes from 1990 to 2019.</p><p><strong>Results: </strong>In 2021, the age-standardized mortality rate (ASMR) for road traffic injuries in China was 4.38 per 100,000 population, lower than the global ASMR of 5.96 per 100,000. Joinpoint regression showed a decrease in road injury mortality for the under-5 age group in China from 1990 to 2021. For the 5-9, 10-14, and 15-19 age groups, mortality rates initially decreased, rose, and declined. Globally, road injury mortality rates decreased for all age groups under 15 and for females aged 15-19; for males aged 15-19, rates decreased, rose slightly, and then decreased again. The APC model indicated that the risk of road injury death varied by age, period, and cohort, with age-specific increases and decreases and a general downward trend across periods in China and globally. EAPC analysis from 2010 to 2019 showed a significant reduction in China (EAPC = -6.92%) compared to a modest decline globally (EAPC = -3.39%).</p><p><strong>Conclusion: </strong>From 1990 to 2021, China's road traffic injury mortality rate demonstrated a continuous decline. Since 2014, China's mortality rate from road traffic injuries has remained below the global average. However, the mortality rate among males remains higher than that of females. Notably, the mortality rate among Chinese children under the age of 5 has shown an upward trend.</p>","PeriodicalId":54422,"journal":{"name":"Traffic Injury Prevention","volume":" ","pages":"1-10"},"PeriodicalIF":1.6,"publicationDate":"2025-07-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144664076","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"The moderating effect of passenger presence on the relationship between driver age and crash culpability: The co-driving influence of passengers.","authors":"Gilsu Pae, Michelle L Reyes, Elizabeth O'Neal, Joseph Cavanaugh, Cara Hamann","doi":"10.1080/15389588.2025.2522242","DOIUrl":"https://doi.org/10.1080/15389588.2025.2522242","url":null,"abstract":"<p><strong>Objective: </strong>Passenger presence influences driver behavior and safety, with its effects shaped by the situational awareness shared between drivers and passengers. This awareness depends on factors such as driving experience and cognitive and physical abilities, which vary across age groups. However, little research has investigated how passenger presence moderates the relationship between driver age and crash culpability. This study examines the relationship between driver age and crash culpability, focusing on the moderating effect of passenger presence.</p><p><strong>Methods: </strong>This study analyzed crash data from the Iowa Department of Transportation from 2015 to 2020, focusing on drivers involved in failure-to-yield crashes (<i>N</i> = 36,577). The analytic sample was limited to multiple-vehicle collisions where the crash report indicated only one driver contributed to the crash. A multivariable logistic regression model was built to examine predictors of culpability among drivers involved in failure-to-yield crashes, with a focus on passenger presence and its interaction with driver age group.</p><p><strong>Results: </strong>Passenger presence was significantly associated with a reduced probability of drivers being culpable in failure-to-yield crashes, with front-seat passengers having a significant effect for middle- and older-aged driver groups. The association was not significant for drivers aged 14-44 but attained significance for those aged 45 and older. Middle-aged drivers (45-64), who may begin to experience age-related declines in driving ability, benefited the most from front-seat passenger presence. The oldest drivers (75+) had a substantially higher probability of being culpable without passengers compared to the youngest drivers (14-24), whereas their probability was similar to that of the youngest drivers when accompanied by passengers.</p><p><strong>Conclusions: </strong>The meaningful differences in the probability of drivers being culpable between presence and absence of front-seat passengers highlight the potential of passengers as co-drivers for safe driving, offering insights for developing effective road safety interventions.</p>","PeriodicalId":54422,"journal":{"name":"Traffic Injury Prevention","volume":" ","pages":"1-6"},"PeriodicalIF":1.6,"publicationDate":"2025-07-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144651248","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"LSTM-based classification of e-scooter trajectory features for single vs tandem riding detection.","authors":"Jiahui Zhao, Jiaming Wu, Sida Jiang, Zhibin Li, Pan Liu","doi":"10.1080/15389588.2025.2524466","DOIUrl":"https://doi.org/10.1080/15389588.2025.2524466","url":null,"abstract":"<p><strong>Objectives: </strong>Dockless electric scooters (e-scooters) have emerged as a popular mode of short-distance transportation in urban environments, offering convenience and flexibility in rental and usage. However, users often engage in unsafe behaviors while riding, posing risks to themselves and others. In this study, we aim to identify unsafe riding behaviors through analysis of riding trajectories, to reduce safety incidents and promote safer e-scooter usage.</p><p><strong>Methods: </strong>This study explores the classification of single and tandem e-scooter riding behaviors using a data-driven approach. Leveraging trajectory data from Gothenburg, Sweden, collected over 11 days in November 2023; we utilize Long Short-Term Memory (LSTM) neural networks to analyze dynamic temporal features.</p><p><strong>Results: </strong>The LSTM model demonstrated significant performance advantages over both RNN and Random Forest models, achieving an accuracy of 92.65%, precision of 91.69%, recall of 93.85%, F1 score of 95.56%, and an AUC of 0.9169.</p><p><strong>Conclusions: </strong>Additionally, optimizing the input sequence length to 240 s of continuous trajectory features balanced computational efficiency with prediction accuracy and stability. Dynamic trajectory features such as acceleration, turning angle, speed, and start SOC play pivotal roles in differentiating riding patterns. The proposed method can assist city authorities and e-scooter operators in real-time risk detection, operational monitoring, and targeted safety interventions, contributing to safer shared micromobility systems.</p>","PeriodicalId":54422,"journal":{"name":"Traffic Injury Prevention","volume":" ","pages":"1-10"},"PeriodicalIF":1.6,"publicationDate":"2025-07-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144651247","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Evaluation of THOR-50M-RS under reclined seating conditions.","authors":"Kathleen D Klinich, Nichole R Orton, Carl Miller, Tyler Vallier, Miriam A Manary, Anne Bonifas, Matthew P Reed, J P Donlon, Jason Forman","doi":"10.1080/15389588.2025.2522935","DOIUrl":"https://doi.org/10.1080/15389588.2025.2522935","url":null,"abstract":"<p><strong>Objective: </strong>Testing was conducted to evaluate the performance of the THOR-50M for Reclined Seating (THOR-50M-RS) with modifications to allow for a more realistic posture and impact response under reclined seating conditions. Results were compared to tests conducted with PMHS under the same conditions.</p><p><strong>Methods: </strong>Nine frontal sled tests were conducted using the THOR-50M-RS in a controlled response seat with the seatback angle set to 45 degrees. Three tests were conducted under each of three conditions: (1) nominal 32 km/h delta V, 4.5 kN seatbelt load limiter (LL), knee bolster condition 1 (180 mm spacing); (2) nominal 50 km/h delta V, 2.7 kN LL, knee bolster condition 1; and (3) nominal 50 km/h delta V, 2.7 kN LL, knee bolster condition 2 (100 mm spacing). The ATD was positioned using a seating procedure based on data from volunteers in reclined seats. THOR-50M-RS instrumentation prioritized channels that could be directly compared to PMHS tests, such as head, T1, and pelvis accelerations. Ten cameras were used to track matching kinematic targets for comparison. Emphasis was placed on qualitative comparison of the head, spine, hip, and knee kinematics, but differences were also quantified using Correlation and Analysis (CORA) techniques. Injury assessment reference values (IARVs) were also calculated.</p><p><strong>Results: </strong>The ATD showed good repeatability within test conditions. Minor durability issues included pelvis flesh damage that was repaired and did not reoccur, as well as upper thigh flesh fraying at the rear seams. Comparison to kinematics from PMHS tests run in the same conditions indicated that the response of the THOR-50M-RS generally aligned well with the PMHS signals. However, the initial narrow width of PMHS corridors often led to low CORA scores, with a range of 0.03 to 0.98 across all test conditions and signals.</p><p><strong>Conclusions: </strong>The THOR-50M-RS was able to meet target postures in reclined seats and showed good repeatability within a given test condition. These tests demonstrate the challenges of designing effective occupant protection systems for occupants in reclined postures, with chest and lumbar spine injury measures exceeding recommended limits (derived for non-reclined modes) in the higher-speed (50 km/h) testing condition.</p>","PeriodicalId":54422,"journal":{"name":"Traffic Injury Prevention","volume":" ","pages":"1-8"},"PeriodicalIF":1.6,"publicationDate":"2025-07-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144651245","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Geospatial analysis of motor vehicle crashes (MVCs) and trauma center accessibility in New York State.","authors":"Nwe Oo Mon, Elisa Szydziak, Gabriela Santos-Revilla, Renee Stolberg, Taleah Angus, Sara Cardozo-Stolberg","doi":"10.1080/15389588.2025.2524846","DOIUrl":"https://doi.org/10.1080/15389588.2025.2524846","url":null,"abstract":"<p><strong>Objective: </strong>To identify and visualize counties at high risk for motor vehicle crashes (MVCs) resulting in injuries, serious injuries, and fatalities in New York State from 2019 to 2023 using geospatial analysis. This study utilizes choropleth and heat maps to highlight crash distribution, overlays trauma center locations to evaluate accessibility and identify coverage gaps, and incorporates principal component analysis and negative binomial regression to explore regional patterns and predictors of severe and fatal crashes.</p><p><strong>Methods: </strong>Crash data were obtained from the Crash Location and Engineering Analysis Repository (CLEAR) database (2019-2023) maintained by the New York State Department of Transportation (NYSDOT). Key metrics include crashes per 10,000 population, and crashes per million vehicle miles traveled (VMT). Geospatial analyses were conducted using choropleth and heat maps to highlight high-risk regions and evaluate trauma center coverage. Principal Component Analysis (PCA) was conducted to explore patterns among county-level crash and access indicators. A negative binomial regression model was used to identify predictors of serious/fatal crash rates.</p><p><strong>Results: </strong>Serious and fatal crash numbers were concentrated in densely populated urban/suburban counties such as Kings, Suffolk, Queens, Erie, and Nassau, while rural counties like Hamilton, Greene exhibited elevated serious and fatal crash rates per 10,000 population. NYC counties had the highest serious and fatal crash rates per million VMT. Heatmaps revealed counties/areas with insufficient trauma center coverage suggesting potential delays in definitive care. Rural counties had more speed and animal-related crashes. PCA identified distinct rural and urban crash exposure patterns. Negative binomial regression results showed higher serious/fatal crash rates were significantly associated with greater rural population percentage and Health Service Regions outside NYC.</p><p><strong>Conclusion: </strong>Geospatial analysis highlights significant disparities in trauma center accessibility and MVC outcomes across New York State. Rural counties with higher crash severity and fatality rates face challenges due to limited trauma care resources. Enhancing trauma center accessibility in underserved regions and implementing targeted interventions such as road safety improvements, wildlife-related crash prevention measures, and emergency transport enhancements are critical to reducing MVC-related fatalities. These findings provide policymakers with actionable insights to allocate resources efficiently and prioritize trauma care expansion in high-risk areas.</p>","PeriodicalId":54422,"journal":{"name":"Traffic Injury Prevention","volume":" ","pages":"1-10"},"PeriodicalIF":1.6,"publicationDate":"2025-07-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144651246","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Injury patterns in motor vehicle collision-adult pedestrian deaths.","authors":"Moheem M Halari, Tanya Charyk Stewart, Kevin J McClafferty, Allison C Pellar, Michael J Pickup, Michael J Shkrum","doi":"10.1080/15389588.2025.2516717","DOIUrl":"https://doi.org/10.1080/15389588.2025.2516717","url":null,"abstract":"&lt;p&gt;&lt;strong&gt;Objectives: &lt;/strong&gt;To describe fatal pedestrian injury patterns in adults 25-64 years old and correlate them with motor vehicle collision (MVC) dynamics and pedestrian kinematics using medicolegal death investigations data of MVCs occurring in the current Canadian MV fleet.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Methods: &lt;/strong&gt;MVC-pedestrian injuries were collated in an Injury Data Collection Form (IDCF) and coded using the Abbreviated Injury Scale (AIS) 2015 revision. The AIS of the most frequent severe injury was noted for individual body regions. The Maximum AIS (MAIS) was used to define the most severe injury to the body overall and by body regions (MAISBR). This study focused on serious to maximal injuries (AIS 3-6), that had an increasing likelihood of causing death. The IDCF was used to extract collision and injury data from the Office of the Chief Coroner for Ontario database of postmortem examinations done at the Provincial Forensic Pathology Unit in Toronto, Canada and other provincial facilities between 2013 and 2019. Injury data were correlated with data about the MVs, and MV dynamics and pedestrian kinematics. The study was approved by the Western University Health Science Research Ethics Board.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Results: &lt;/strong&gt;There were 318 adults: 200 (62.9%) males and 118 (37.1%) females. Adult pedestrians comprised 47.5% (318/670) of all autopsied pedestrians. Vehicle type was known in 292 cases, and cars (&lt;i&gt;n&lt;/i&gt; = 99/292, 33.9%) were the most frequent type of vehicle in single vehicle impacts; however, collectively vehicles with high hood edges (i.e., greater distance between the ground and hood edge) such as light trucks, heavy trucks and buses were in the majority. Pedestrian kinematics were known in 288/299 single impact-related deaths. Forward projection (&lt;i&gt;n&lt;/i&gt; = 113/288, 39.2%) was the most frequent type and resulted from impacts with high hood edge vehicles. Compared to car impacts, pedestrians struck by high hood edge vehicles were more likely to be runover. Based on MAISBR ≥3 injuries, the head was the most severely injured (median MAISBR = 4), followed by neck (median MAISBR = 3), thorax (median MAISBR = 4), abdomen/retroperitoneum (median MAISBR = 4) and pelvis (median MAISBR = 3). About 70% of the pedestrians were in circumstances which increased their risk of being struck. More than half (176/318, 55.3%) had a positive toxicology result. About ¼ (27.4%) had a positive blood ethanol result. Nearly all pedestrians with positive alcohol results did not have the right of way when struck.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Conclusion: &lt;/strong&gt;The current study was a comprehensive analysis of fatal injury patterns and specific injuries in adult pedestrians struck by motor vehicles. By collation and analysis of comprehensive data derived from postmortem examinations, associations between injury patterns in the adult age group were correlated with a range of factors related to motor vehicle types, reflective of the current Canadian fleet, collision","PeriodicalId":54422,"journal":{"name":"Traffic Injury Prevention","volume":" ","pages":"1-11"},"PeriodicalIF":1.6,"publicationDate":"2025-07-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144644150","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Analysis of factors contributing to severity of AV crashes at intersections: Insights from the Autonomous Vehicle Operation Incident Dataset Across The Globe (AVOID).","authors":"Rui Guo, Yanyan Chen, Yunchao Zhang, Fuhua Yi","doi":"10.1080/15389588.2025.2520006","DOIUrl":"https://doi.org/10.1080/15389588.2025.2520006","url":null,"abstract":"<p><strong>Objectives: </strong>The market for autonomous vehicles (AVs) is developing rapidly, while safety concerns persist as a critical challenge hindering their widespread development and commercialization. Intersections, characterized by highly dynamic and unpredictable traffic conditions, represent particularly high-risk scenarios for AVs. This study systematically investigates the key risk factors influencing collision severity at intersections to enhance AV safety performance.</p><p><strong>Methods: </strong>This study employs the Autonomous Vehicle Operation Crash Dataset Across The Globe (AVOID) to investigate the key factors influencing the injury severity of AV related collisions. A hybrid analytical framework is proposed, integrating an XGBoost-SHAP model for feature importance analysis and a multinomial logit (MNL) model for statistical inference. Following the factor importance ranking, nine critical determinants were selected to examine their individual effects on injury severity levels. Furthermore, the XGBoost-SHAP approach was utilized to explore interaction effects among significant factors, revealing synergistic relationships between key features.</p><p><strong>Results: </strong>The results indicate that the majority of crashes at intersections occurred when AVs were stationary or moving at low speeds (17.71% while stopped, 42.76% at speeds below 10 mph). Approximately 58.14% of the crashes involved autonomous driving mode, with an injury rate 14.5% higher compared to manual mode. Factors such as pre-crash movement, crash scene, contact area, pre-crash speed, and autonomous mode significantly influenced injury severity. Crashes occurring during straight-line travel or lane changes in autonomous mode tended to result in more severe injury compared to manual driving. Additionally, crashes in steering direction scenes at speeds between 10 and 20 mph were associated with higher injury severity, and speeds exceeding 20 mph in traffic through scenes led to even more severe injuries.</p><p><strong>Conclusions: </strong>This study reveals the main factors influencing the severity of collisions in autonomous vehicles and the combination of the following factors that may increase the severity of injuries: autonomous driving mode, lane changing, turning or crossing scenarios, and high-speed driving. The results advance the understanding of autonomous vehicle safety and offer potential implications for enhancing self-driving systems.</p>","PeriodicalId":54422,"journal":{"name":"Traffic Injury Prevention","volume":" ","pages":"1-8"},"PeriodicalIF":1.6,"publicationDate":"2025-07-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144644138","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Analyzing extreme muti-vehicle rear-end collision risks in adverse weather through Generalized Pareto Regression Trees.","authors":"Li Li, Jing-Yu Yang, Meng Zhang, Qi Niu, Qing-Chang Lu, Feng Zhu","doi":"10.1080/15389588.2025.2517386","DOIUrl":"https://doi.org/10.1080/15389588.2025.2517386","url":null,"abstract":"<p><strong>Objectives: </strong>The study seeks to explore rear-end collision risks in multi-vehicle car-following scenarios under adverse weather conditions by proposing an integrated framework.</p><p><strong>Methods: </strong>The integrated framework is applied to a case study of three-vehicle car-following scenario in Norway without loss of generality. For identifying car-following groups with extreme collision risks, the collision risk of each group in the raw dataset is evaluated using an extended probabilistic driving risk field. Quantitative collision risks are analyzed to fit the Generalized Pareto distribution, and high-risk scenarios screened <i>via</i> mean residual life plots and threshold stability plots. To determine risk-contributing factors, Generalized Pareto Regression Trees (GPRT) are constructed to pinpoint significant influences on rear-end collision risks. By integrating the classification and regression trees with extreme value theory, the GPRT discards data assumptions and covariate continuity requirements of most extreme value analysis (e.g., extreme quantile regression). Moreover, the GPRT not only identifies the hierarchical structure of variables affecting rear-end collision risks but also determines risk-impact thresholds for covariates, offering superior interpretability and engineering applicability.</p><p><strong>Results: </strong>The results show that revealed risks conform well to the Generalized Pareto distribution, allowing for the formulating Generalized Pareto regression trees. Compared to the Generalized Additive Model (GAM) and Negative Binomial Regression (NBR) methods, the GPRT approach demonstrates superior performance in balancing risk fitting accuracy and model complexity. Vehicle speeds, weights, and headways emerge as critical factors for collision risks under clear, rainy, and snowy conditions. As weather conditions deteriorate from clear to rainy or snowy, the influence of vehicle speed and weight diminishes, while the influence of headway and road surface conditions becomes more pronounced. Collision risks are high on sunny days, regardless of whether the middle vehicles of three-vehicle groups are light or heavy vehicles.</p><p><strong>Conclusions: </strong>The integrated evaluation framework developed in this study provides a tool for car-following safety assessment under extreme weather conditions.</p>","PeriodicalId":54422,"journal":{"name":"Traffic Injury Prevention","volume":" ","pages":"1-9"},"PeriodicalIF":1.6,"publicationDate":"2025-07-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144644139","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}