Traffic Inj Prev [JOURNAL]
Sun
200 papers
Study of adaptation parameters of merging zones for freeway interchange ramps considering CAV & AV mixed traffic environments.
Zhang S, Yang Y, Wang L … +3 more , He J, Gao J, Zhang H
Traffic Inj Prev · 2026 Jul · PMID 42391553 · Publisher ↗
OBJECTIVE: This study aims to investigate the adaptability of existing freeway interchange ramp merging zone geometries to mixed traffic flows comprising connected automated vehicles (CAVs), autonomous vehicles (AVs), an...
OBJECTIVE: This study aims to investigate the adaptability of existing freeway interchange ramp merging zone geometries to mixed traffic flows comprising connected automated vehicles (CAVs), autonomous vehicles (AVs), and human-driven vehicles (HDVs). It seeks to identify optimal road geometric parameters to enhance traffic safety and efficiency in such environments. METHODS: Using a merging zone on a Chengdu expressway interchange as a case study, a VISSIM microsimulation model was constructed for a three-lane mainline with a single-ramp parallel merging zone. The study incorporated the length of forbidden lane-changing markings as a key road parameter. An orthogonal experimental design was employed to structure simulation experiments. A weighted grey clustering model assessed the safety level of the merging zone, while the entropy-weighted TOPSIS method comprehensively evaluated the merits of road parameter combinations to determine the optimal geometric design. RESULTS: The results indicate that deploying CAVs can moderately improve safety in merging zones. The study also proposes the following recommended values for road parameters under mixed traffic flow conditions: a sight triangle angle of 7°; and for low CAV permeability, an acceleration lane length of 290 m, a merge taper length of 100 m, and forbidden lane-changing marking of 41 m; for high CAV permeability, these values become 275, 110, and 37 m, respectively. CONCLUSIONS: Under mixed traffic flow conditions involving HDVs, AVs, and CAVs, the existing geometric characteristics and traffic flow requirements of merging zones differ to some extent from conventional conditions. The parameter optimization guidelines proposed in this study can inform the design optimization and adaptive retrofitting of freeway merging zones in comparable mixed traffic environments.
Behavioral determinants of motorcyclist personal protective equipment (PPE) use in five European cities.
Papafoti L, Pérez Avendaño MA, Kontaxi A … +4 more , Carroll J, Huster D, Auerbach K, Lopez-Valdes FJ
Traffic Inj Prev · 2026 Jul · PMID 42391532 · Publisher ↗
OBJECTIVES: Motorcyclists face high risk of severe and long-term injuries, yet the use of personal protective equipment (PPE) beyond helmets remains inconsistent across European cities. This study examined determinants o...
OBJECTIVES: Motorcyclists face high risk of severe and long-term injuries, yet the use of personal protective equipment (PPE) beyond helmets remains inconsistent across European cities. This study examined determinants of motorcycle PPE use to understand barriers and enablers influencing protective behavior. METHODS: A cross-sectional survey was conducted in five European cities (Athens, Barcelona, Copenhagen, London, and Rome). Data were collected through an online questionnaire designed using the COM-B model of behavior (Capability, Opportunity, Motivation). The survey examined ownership and use of multiple motorcycle PPE types beyond helmets (protective jackets and leg protection). Additional measures captured riding context (e.g., urban/non-urban riding, trip purpose, and riding frequency), perceived barriers to PPE, and social norms. Statistical analyses examined associations between PPE use and behavioral determinants across the COM-B domains. Awareness and interest in innovative PPE were also analyzed. RESULTS: Among the 905 respondents, helmet ownership and use were high, whereas ownership and use of non-mandatory PPE such as jackets and leg protection were lower. A gap emerged between access to PPE and consistent use, particularly for jackets and leg protection. Multivariable analyses showed that physical opportunity factors played a central role in PPE use: riders reporting greater comfort when wearing protective equipment and greater convenience in carrying it had significantly higher odds of consistent use. Among riders who did not always wear PPE, the most frequently reported barriers were heat, discomfort, and inconvenience when not riding. Reflective motivation also influenced behavior. Although most riders recognized the protective value of jackets and leg protection, many reported that such equipment is less necessary for short journeys or urban riding, indicating context-dependent risk perceptions. Social opportunity factors were also present: perceptions that other riders commonly use non-mandatory PPE were relatively low. Respondents also showed interest in innovative PPE, particularly upper-body protection, although affordability, comfort, and usability were identified as key considerations for adoption. CONCLUSIONS: Among riders participating in the study, PPE use is shaped by interacting opportunity-, and motivation-related factors, with practical barriers such as heat, discomfort, and inconvenience playing a central role in limiting consistent use of non-mandatory equipment. The findings also highlight the importance of context-specific risk perceptions, as many riders consider PPE less necessary for short or urban trips, and of social influences linked to the perceived behavior of other riders. Improving PPE adoption requires behaviorally informed strategies addressing usability and comfort of protective equipment, strengthening risk communication in everyday riding contexts, and promoting social normalization of PPE use. Interest in innovative PPE concepts suggests opportunities for further uptake, prioritizing affordability, comfort, and ease of use alongside protection.
Impact of grade designs of long mountainous freeway tunnel with crest vertical curve on traffic safety at tunnel portals.
Hu Y, Yang Y, Wang Z … +3 more , Cheng J, Xu C, Zhang W
Traffic Inj Prev · 2026 Jul · PMID 42391531 · Publisher ↗
OBJECTIVE: Crest vertical curves are frequently employed in long mountainous freeway tunnels to facilitate drainage and ventilation, yet the mechanisms by which specific grades and grade changes within these curves influ...
OBJECTIVE: Crest vertical curves are frequently employed in long mountainous freeway tunnels to facilitate drainage and ventilation, yet the mechanisms by which specific grades and grade changes within these curves influence driving behavior and traffic safety at tunnel portals remain insufficiently understood. This study aims to explore the effects of crest vertical grade designs, specifically entrance grades, exit grades, and grade changes, on driving behavior at the portals of long mountainous freeway tunnels. METHODS: Based on a survey of alignment design indicators of over 100 tunnels in Guangdong Province, this study identified that most long tunnels adopt a crest vertical grade design and their typical alignment indicators. Based on the investigation, a 26-km-long freeway model comprising 10 tunnels with distinct vertical profile was constructed. Subsequently, a driving simulation experiment was conducted with 32 recruited participants. Driving behavior data were collected within 100-meter zones inside and outside each portal. Eight key behavioral indicators, including mean and standard deviation of speed, time headway, lane departure, and acceleration, were extracted. Repeated-measures one-way analysis of variance (ANOVA) was then used to identify significant differences across various design conditions. Subsequently, indicators exhibiting statistical significance were integrated into a Fuzzy Comprehensive Evaluation (FCE) model, combining with entropy weight methods to objectively quantify the overall safety performance of each design scheme. RESULTS: ANOVA results indicated that five specific indicators showed significant variations under different grade designs: mean and standard deviation of speed, mean and standard deviation of time headway, and mean acceleration. Lane departure and standard deviation of acceleration were not significantly affected by grade designs. The Fuzzy Comprehensive Evaluation revealed distinct optimal conditions for portal safety. At tunnel entrances, uphill grades yielded significantly higher safety scores compared to flat or downhill sections, with the 1.5% uphill grade achieving the highest comprehensive score. Conversely, flat (0%) entrances resulted in the lowest safety ratings. At tunnel exits, safety scores generally decreased as the exit grade became less steep; the 1.5% exit grade produced the optimal safety outcome. Regarding internal grade changes, a moderate change of 3% resulted in the highest safety scores, whereas larger changes led to increased speed dispersion and reduced time headways, lowering the overall safety evaluation. CONCLUSIONS: This study confirms that vertical alignment design critically influences driver performance and safety at portals of long tunnels. These findings provide quantitative guidance for updating freeway tunnel design specifications, recommending that designers prioritize uphill approaches and moderate grade transitions. Furthermore, the results offer a scientific basis for traffic authorities to implement targeted safety management measures, thereby mitigating crash risks in critical tunnel portal zones.
Associations between blood ethanol concentration, clinical impairment test results, and traffic accident involvement among apprehended drivers.
Høiseth G, Olsen T, Mørland J … +2 more , Hjelmeland K, Strand M
Traffic Inj Prev · 2026 Jul · PMID 42391523 · Publisher ↗
OBJECTIVE: Driving under the influence of drugs is a major cause of road traffic accidents, and alcohol markedly increases crash risk. A clinical test of impairment (CTI) is in some countries performed alongside blood sa...
OBJECTIVE: Driving under the influence of drugs is a major cause of road traffic accidents, and alcohol markedly increases crash risk. A clinical test of impairment (CTI) is in some countries performed alongside blood sampling in apprehended drivers, yielding a conclusion of either "impaired" or "not impaired". While the relationship between CTI results and blood drug concentrations has been examined, no previous studies have evaluated CTI outcomes in relation to traffic accidents. This study assesses whether ethanol-positive drivers judged as "impaired" have a higher risk of traffic accident involvement than those assessed as "not impaired", and whether accident involvement increases with impairment severity and ethanol concentration. METHODS: Drivers positive for ethanol alone were included if a valid CTI conclusion was available and blood sampling occurred 0.25-3 h after the incident. Cases were categorized as either traffic accident group or non-accident controls, the latter including police stops unrelated to accidents (e.g., routine checks, license issues, erratic driving). Age, sex, blood ethanol concentration and CTI results were compared between groups. Associations between impairment level and accident risk were analyzed using multivariable logistic regression. RESULTS: The accident included group 5,290 cases and the non-accident group included 17,596 cases. Accident-involved drivers were younger (median 30 vs. 34 years, < 0.001), had higher ethanol concentrations (median [IQR] 1.60 g/kg [1.06, 2.05] vs. 1.29 g/kg [0.63, 1.89], < 0.001) and were more often assessed as "impaired" (94.8% vs. 90.3%, < 0.001) compared to non-accident drivers. After adjusting for age, sex, ethanol concentration and driving time, CTI-assessed impairment remained significantly associated with accident involvement (OR 1.47; 95% CI 1.28, 1.69). Accident odds increased progressively with both impairment severity and ethanol concentration (p < 0.001). CONCLUSION: Ethanol-positive drivers assessed as "impaired" by the CTI had 47% higher odds of being involved in traffic accidents compared with drivers apprehended for other reasons, independent of ethanol concentration and demographic factors. Being assessed as "impaired" by the CTI therefore appear to be independently associated with accident risk among alcohol-positive drivers.
Crash severity determinants in Saudi Arabia's Eastern Province: Evidence from logistic regression and neural networks.
Minhas KS, Dalhat MA
Traffic Inj Prev · 2026 Jul · PMID 42391129 · Publisher ↗
Road traffic crashes remain a major public health concern in Saudi Arabia, which recorded a high fatality rate of 18.5 per 100,000 population in 2021. The objective of this study is to develop machine learning models whi...
Road traffic crashes remain a major public health concern in Saudi Arabia, which recorded a high fatality rate of 18.5 per 100,000 population in 2021. The objective of this study is to develop machine learning models which can identify crash severity determinants and predict the severity of crashes with high accuracy. This study analyzes 4,979 crashes from 2017-2022 in the Eastern Province to identify factors influencing crash severity. A multinomial logistic regression model was developed using crash severity as the dependent variable, achieving an overall prediction accuracy of 88.8%. A Multilayer Perceptron Neural Network was also developed with achieved 88.6-89.7% accuracy. Significant predictors of fatal crashes included the number of people involved, peak-hour periods, autumn season, clear weather, sandstorms, and driver speeding. For injury crashes, key predictors included road geometry, surface condition, collision type, and driver distraction. Findings provide locally grounded insights to guide targeted road safety interventions which can help in improving the road safety situation to support the national Vision 2030's goal of reducing fatalities below 10 per 100,000.
Seeing the automation malfunction before it happens: designing information cues that support driver situation awareness.
Garakani G, Guo H, Chen B … +2 more , Pulver E, Bao S
Traffic Inj Prev · 2026 Jul · PMID 42390282 · Publisher ↗
OBJECTIVE: Advanced Driver Assistance Systems (ADAS) may degrade in complex environments, producing malfunction states that require driver readiness to intervene. However, empirical guidance remains limited on how vehicl...
OBJECTIVE: Advanced Driver Assistance Systems (ADAS) may degrade in complex environments, producing malfunction states that require driver readiness to intervene. However, empirical guidance remains limited on how vehicle-driver communication should be structured to support situation awareness (SA) during such events. METHODS: This study evaluates three structured information cue functions, Compromised Function (CF), Tracking Status (TS), and Action Rationale (AR), designed to support perception (Occurrence), comprehension (Cause), and projection (Solution) during ADAS malfunctions. Twenty-seven ADAS-equipped drivers evaluated cue effectiveness across three representative malfunction scenarios using a mixed experimental design. Linear mixed-effects models assessed absolute effectiveness, relative differences among cue types, and demographic moderation. RESULTS: CF demonstrated the most stable and robust advantage at lower SA levels. At the Occurrence stage, CF was rated significantly higher than TS (Δ = 1.33) and AR (Δ = 0.93). A similar pattern emerged at the Cause stage, where CF exceeded AR (Δ = 0.73) and TS (Δ = 0.72). Demographic moderation effects, including age, gender, education, and ADAS experience, were consistently observed, indicating stable differences in cue effectiveness across user groups. CONCLUSIONS: These findings highlight the importance of clearly communicating system vulnerabilities to support early malfunction recognition and to guide more effective transparency design for drivers facing automation malfunctions.
YOLO-MIRNet: detection of distracted driving with enhanced generalization using a lightweight multi-scale interaction optimized model.
Zhang C, Ba K, Bai X … +1 more , Hu H
Traffic Inj Prev · 2026 Jul · PMID 42385235 · Publisher ↗
OBJECTIVE: Distracted driving is one of the key factors contributing to road crashes. Real-time and robust monitoring of drivers' states is of great significance for improving the level of road traffic safety. Existing m...
OBJECTIVE: Distracted driving is one of the key factors contributing to road crashes. Real-time and robust monitoring of drivers' states is of great significance for improving the level of road traffic safety. Existing methods based on convolutional neural networks still face challenges in terms of generalization ability and practicality under cross-scenario, cross-device, and complex environmental conditions. METHODS: To enhance the adaptability and robustness of the model in diverse real driving scenarios, this paper proposes a lightweight detection framework with enhanced generalization ability, namely YOLO-MIRNet. Based on the YOLOv8 architecture, the model integrates three core innovations: first, a Global and Detail Feature Aggregation (GDFA) module is designed, which collaboratively extracts multi-scale features through multi-branch dilated convolutions and channel shuffling; second, a Multi-scale Adaptive Gated Fusion (MSAGF) module is constructed, which realizes adaptive fusion of cross-scale features by utilizing pixel-wise attention and gating mechanism; third, an improved Lightweight Channel Attention (iMLCA) module is embedded to strengthen the discriminative ability of key features with low computational cost. Model training and evaluation are conducted on the well-partitioned public StateFarm dataset and the more challenging AUC dataset (partitioned by driver). RESULTS: Experimental results demonstrate that YOLO-MIRNet achieves a detection precision of 99.92% and a mean Average Precision (mAP) of 99.48% on the StateFarm dataset; it also attains a precision of 87.03% and an mAP of 93.99% on the AUC dataset, significantly outperforming the comparison models. With a parameter count of only 12.68M, the model exhibits excellent lightweight characteristics, and the model's single-frame inference time remains stable between 42 and 57 milliseconds. The generalization performance test results are prominent: in the cross-dataset evaluation ("StateFarm→AUC"), the precision reaches 67.03%; under the reverse setting ("AUC→StateFarm"), the accuracy is further improved to 72.61%. In addition, on the 100-driver multi-modal dataset containing day and night scenarios, the model's detection performance during both day and night significantly surpasses the baselines, demonstrating strong environmental adaptability. CONCLUSIONS: The proposed YOLO-MIRNet framework achieves an optimal balance among detection accuracy, model lightweight, and cross-domain generalization ability. It provides an efficient and reliable technical solution for real-time monitoring of distracted driving, which is conducive to promoting the practical application and deployment of intelligent traffic safety systems.
Naturalistic driving data can identify elevated driving risk in adults with Type 1 diabetes.
Joshi A, Rizzo M, Sharma A
Traffic Inj Prev · 2026 Jul · PMID 42384762 · Publisher ↗
OBJECTIVE: Individuals with Type 1 Diabetes Mellitus (T1DM) may experience acute glucose events, such as hypoglycemia and severe hyperglycemia, that can impair attention, reaction time, and decision-making during activit...
OBJECTIVE: Individuals with Type 1 Diabetes Mellitus (T1DM) may experience acute glucose events, such as hypoglycemia and severe hyperglycemia, that can impair attention, reaction time, and decision-making during activities of daily living (ADL), such as driving. While continuous glucose monitoring (CGM) detects such events, less is known about whether everyday driving behavior reflects days when these disturbances occur during active vehicle operation. We evaluate whether daily driving behavior differs on days when acute glucose-related impairment occurs during driving among individuals with T1DM. METHODS: This study followed 18 adults with T1DM (mean age = 30.78; 11 females) over 4 wk. Real-world driving data were continuously recorded, and sleep was assessed using wrist-worn actigraphy as a background daily contextual factor. Glucose levels were measured using CGM. The primary outcome was a high-risk driving day, defined as a day on which hypoglycemia or severe hyperglycemia occurred during active driving, as determined from CGM data. Gradient-boosted decision tree models (XGBoost) were evaluated using a nested leave-one-subject-out cross-validation framework to distinguish high-risk driving days based on behavioral features. Models were trained using driving-only, sleep-only, and combined driving-and-sleep feature sets. RESULTS: Models trained using driving features alone demonstrated the most consistent discriminative performance in identifying high-risk driving days (AUROC = 0.67; F1 = 0.65), outperforming the sleep-only and sleep + driving combined model. Influential driving features reflect exposure and temporal patterns, including afternoon peak trips, trips near home, and total driving time. Sleep features alone showed limited ability to distinguish high-risk driving days (AUROC = 0.48). CONCLUSIONS: Driving exposure features, particularly those capturing trip frequency, timing, and duration, were informative for distinguishing days when glucose-related impairment occurred during driving in individuals with T1DM. These findings demonstrate the feasibility of behavior-based approaches for flagging elevated driving risk when glucose-related impairment occurs. Such approaches may complement physiological monitoring in transportation safety and injury-prevention contexts.
Explicit and implicit HMI for tunnel blind spot: insights from a naturalistic driving experiment.
Zhu B, Yuan L, Wang R … +4 more , Yao W, Rong D, Xu C, Jin S
Traffic Inj Prev · 2026 Jun · PMID 42372100 · Publisher ↗
OBJECTIVES: Cooperative Vehicle-Infrastructure Systems (CVIS) can significantly enhance tunnel safety, yet most research relies on driving simulators that lack the psychological realism of actual confined environments. T...
OBJECTIVES: Cooperative Vehicle-Infrastructure Systems (CVIS) can significantly enhance tunnel safety, yet most research relies on driving simulators that lack the psychological realism of actual confined environments. This study addresses the gap by evaluating Human-Machine Interaction (HMI) effectiveness in real-world tunnel blind spots, where restricted visibility and confined roadway geometry create high-risk scenarios. METHODS: A naturalistic driving experiment was conducted in the Guayanling Tunnel, China. Participants drove a vehicle equipped with eye-tracking and onboard data units through a curved blind spot under four conditions: no warning, informative (implicit), instructive, and commanding (explicit) warnings. The study analyzed vehicle data (speed, braking, lateral acceleration) and eye-tracking metrics to assess safety and effectiveness, while also examining the influence of driver gender, experience, and driving style. RESULTS: Across the fixed sequential warning rounds, commanding warnings were associated with the largest changes in speed regulation and earlier braking, but they were also associated with increased lateral-control variability. Instructive warnings also advanced braking, whereas informative warnings showed smaller behavioral effects. Formal interaction analyses provided limited support for subgroup moderation: braking responses varied by driving experience, whereas gender- and driving-style-related differences were weak or exploratory. CONCLUSIONS: Explicit, commanding HMIs are superior for immediate hazard response in tunnel blind spots but impose a higher cognitive load. The effectiveness of warnings is not universal; it varies significantly based on individual driver characteristics. Therefore, relying solely on generic warning strategies may be insufficient for diverse driver populations.
Comparison of head-related injuries between standing electric scooter riders and pedestrians involved in motor vehicle collisions in Japan.
Sakamoto R
Traffic Inj Prev · 2026 Jun · PMID 42347807 · Publisher ↗
OBJECTIVE: While previous research suggests that e-scooter riders and pedestrians may exhibit similar injury patterns due to a shared standing posture, direct comparisons using unified datasets are lacking. This study ai...
OBJECTIVE: While previous research suggests that e-scooter riders and pedestrians may exhibit similar injury patterns due to a shared standing posture, direct comparisons using unified datasets are lacking. This study aims to identify the distinctive injury characteristics of e-scooter riders by comparing them directly with pedestrians in motor vehicle collisions, with a particular focus on head-related injury risk. METHODS: Using traffic accident statistics from the National Police Agency of Japan (January 2024-June 2025), I analyzed 165 injured e-scooter riders and 23,599 injured pedestrians ranging in age from 18 to 59 years. I statistically compared demographic, accident and injury characteristics (e.g., sex, injury source, primary injury region). To ensure a focused analysis, the primary injury region was collapsed into a binary outcome (Head/Face/Neck vs. Other). A logistic regression model was performed to estimate the adjusted odds ratio for head-related (Head/Face/Neck) injuries, adjusting for demographics and accident characteristics. RESULTS: E-scooter riders were significantly younger than pedestrians ( < 0.001). Although no significant difference was observed in injury severity, e-scooter riders had a significantly higher proportion of head-related (Head/Face/Neck) injuries (41.8% vs. 23.1%; < 0.001). Multivariable analysis revealed that unhelmeted e-scooter riders exhibited a significantly higher risk of head-related injuries compared with pedestrians (adjusted odds ratio 2.15, 95% confidence interval 1.53-3.00). CONCLUSIONS: This study suggests that a shared standing posture does not necessarily result in similar injury outcomes. The results indicate that e-scooter injury profiles, particularly for head-related injuries, differ from those of pedestrians. These findings suggest that it may be appropriate for traffic safety policies to consider e-scooter riders as a distinct category and prioritize the promotion of helmet use to mitigate head-related injury risks.
Effects of cabin thermal conditions and road type on driver workload and performance: A driving simulator study.
Lu B, Liu X, Zhang P … +3 more , Lin Y, Xie M, Chen C
Traffic Inj Prev · 2026 Jun · PMID 42347777 · Publisher ↗
OBJECTIVE: Driver workload is a safety-relevant factor that may contribute to impaired driving performance and an increased risk of traffic injury. Cabin temperature is a modifiable in-vehicle environmental condition, bu...
OBJECTIVE: Driver workload is a safety-relevant factor that may contribute to impaired driving performance and an increased risk of traffic injury. Cabin temperature is a modifiable in-vehicle environmental condition, but its influence on driver cognitive workload under different road scenarios remains insufficiently understood. This study examined the effects of cabin thermal conditions and road type on electroencephalography (EEG)-derived driver workload, subjective state, and task performance in a controlled driving simulator experiment. METHODS: Seventeen licensed drivers completed simulated driving tasks under three cabin thermal conditions corresponding to PMV = -1, 0, and 1 and two road scenarios: urban road and expressway. Electroencephalography was recorded during each driving task, and theta, alpha, and beta band powers were extracted. Two ratio-based EEG workload indicators, β/α and β/(θ + α), were calculated. Subjective workload, fatigue, thermal comfort, and driving task error rate were also calculated. Two-way repeated-measures ANOVA was used to examine the effects of thermal conditions and road type. An exploratory thermal human cognitive workload index (TH-CWI) was constructed by integrating EEG-derived workload, subjective workload, fatigue, task error rate, and thermal comfort. RESULTS: Warmer cabin conditions were associated with higher beta band power and higher EEG-derived workload-related ratios, particularly under PMV = 1. Among the two EEG ratio indicators, β/(θ + α) showed a clearer differentiation across thermal conditions than β/α. The thermally neutral condition was associated with lower subjective workload, lower fatigue ratings, and more stable task performance. Road type was associated with differences in EEG spectral activity; higher alpha and beta power was observed during expressway driving than during urban driving. Road type did not significantly affect the ratio-based EEG workload indicators. The exploratory TH-CWI showed the lowest integrated workload under PMV = 0 and the highest under PMV = 1, suggesting that cabin thermal conditions had a more consistent influence on workload-related outcomes than road type in the simulator setting. CONCLUSIONS: This simulator-based study provides preliminary evidence that PMV-defined cabin thermal conditions may influence driver workload-related EEG activity, subjective state, and task performance. Slightly warm cabin conditions were associated with elevated EEG-derived workload-related indicators and less favorable subjective and behavioral outcomes, whereas thermally neutral conditions appeared to be more favorable for reducing fatigue and maintaining stable performance. Road-type effects should be interpreted as differences in EEG spectral activation patterns rather than as direct evidence of higher workload during expressway driving. The exploratory TH-CWI may provide a preliminary multimodal summary of driver state and may have potential relevance for future driver-state monitoring, but further validation is needed.
Cannabis use among motorcyclists in Brazil: prevalence and associated factors from a cross-sectional study in two cities.
Bombana HS, Greve JMD, Sinagawa DM … +6 more , de Oliveira RA, Gonçalves Lopes L, Góes Endo L, de Carvalho HB, Yonamine M, Leyton V
Traffic Inj Prev · 2026 Jun · PMID 42347744 · Publisher ↗
OBJECTIVE: To estimate the prevalence of cannabis use among motorcyclists and identify factors associated with tetrahydrocannabinol (THC) positivity in two major Brazilian cities. METHODS: A cross-sectional roadside stud...
OBJECTIVE: To estimate the prevalence of cannabis use among motorcyclists and identify factors associated with tetrahydrocannabinol (THC) positivity in two major Brazilian cities. METHODS: A cross-sectional roadside study was conducted with motorcyclists in São Paulo (2023) and Campinas (2026), Brazil. Participants completed a structured questionnaire and provided oral fluid samples, which were analyzed using liquid chromatography-tandem mass spectrometry. Multivariable logistic regression was used to assess factors associated with THC positivity. RESULTS: Among 502 motorcyclists, 13.1% tested positive for at least one psychoactive substance, with THC detected in 9.6%. In adjusted analyses, age showed a non-linear association with THC positivity, with the probability of testing positive peaking at approximately 32 years and declining thereafter (age: aOR = 1.636, 95% CI: 1.080-2.480; age: aOR = 0.992, 95% CI: 0.986-0.998). A history of prior motorcycle crash was independently associated with higher odds of THC positivity (aOR = 2.623; 95% CI: 1.053-6.534), though this estimate should be interpreted with caution given the wide confidence interval. The association between app-based delivery work and THC positivity was not statistically significant after adjustment. CONCLUSIONS: Cannabis use was frequently identified among motorcyclists, a high-risk group for road traffic injuries. These findings underscore the importance of strengthening monitoring strategies and generating evidence to support policies addressing drug-impaired driving in Brazil.
Determinants of risky driving behaviors in professional bus drivers: A questionnaire investigation.
Gao Z, Meng Z, Xiang J … +2 more , Yin Y, Wang J
Traffic Inj Prev · 2026 Jun · PMID 42347704 · Publisher ↗
OBJECTIVE: Bus drivers operate under demanding and challenging conditions, characterized by elevated workloads, multitasking requirements, and navigation through complex mixed-traffic environments. This study aimed to co...
OBJECTIVE: Bus drivers operate under demanding and challenging conditions, characterized by elevated workloads, multitasking requirements, and navigation through complex mixed-traffic environments. This study aimed to comprehensively identify the determinants of risky driving behaviors among professional bus drivers. METHOD: A questionnaire-based investigation was conducted among 1,400 professional bus drivers in Changsha, China. Based on drivers' self-reported historical risky driving behaviors (including speeding, fatigued driving, distracted driving, and aggressive lane changes), K-means clustering analysis was used to categorize the drivers into three distinct risk groups: low-risk, medium-risk, and high-risk. Subsequently, logistic regression models (multinomial Logit for risk group classification and ordered Logit for specific risky behaviors) were employed to examine the influences of driver workload, safety awareness, bus enterprise management, traffic environments, and demographic characteristics on risk level classification and the occurrence of specific risky behaviors. RESULTS: Compared to the low-risk group, high-risk driver groups were associated with higher workloads, drivers aged 45 years and above, lower driving experience, and driving experience of less than five years. Notably, drivers with 11-20 years of experience showed higher risk than those with 6-10 years, suggesting a "burnout" effect. Meanwhile, the medium-risk group was correlated with higher workloads, diminished safety awareness, divorced drivers, and sleep duration of less than seven hours. Furthermore, the results revealed that higher workload and lower safety awareness were significantly associated with increased likelihoods of all four types of risk behaviors, while inadequate enterprise management and adverse traffic environments were only associated with certain risk behaviors. This study provides empirical evidence for the identification of risks among bus drivers and advocates for differentiated safety management strategies, emphasizing the necessity for collaborative interventions across multiple dimensions, including stress management, traffic environment optimization, and targeted interventions for high-risk demographic subgroups or specific bus routes.
Driving safety evaluation for hazardous materials vehicle drivers based on visual characteristics.
Xian H, Liu G, Zhang M … +3 more , Wang J, Kou J, Zhao J
Traffic Inj Prev · 2026 Jun · PMID 42340372 · Publisher ↗
OBJECTIVE: To develop a driving safety evaluation model for hazardous materials vehicle drivers based on visual characteristics. METHODS: Twenty-three professional hazardous materials vehicle drivers were recruited to pa...
OBJECTIVE: To develop a driving safety evaluation model for hazardous materials vehicle drivers based on visual characteristics. METHODS: Twenty-three professional hazardous materials vehicle drivers were recruited to participate in driving simulation experiments under normal, cognitively distracted, and operationally distracted conditions. Based on the Big Five personality questionnaire, K-means clustering was applied to classify the drivers into three types: balanced, calm, and impulsive. Significance analyses were conducted to identify differences in visual characteristics among driver types across driving states. An evaluation model was subsequently constructed by combining the Analytic Hierarchy Process (AHP)-Entropy Weight Method with fuzzy comprehensive evaluation, and the model was validated with case studies. RESULTS: Under distracted driving, significant differences were observed in single fixation duration, cumulative fixation duration, fixation frequency, single saccade duration, blink frequency, and pupil diameter compared to normal driving. Pupil diameter emerged as a universal sensitive indicator across all personality types, while the response patterns of visual characteristics to distraction type varied heterogeneously by personality. The model evaluations for the three types-balanced, calm, and impulsive-were highly consistent with the drivers' personality traits and visual behavior changes. CONCLUSION: This study integrates personality traits with visual characteristic parameters, filling a gap in the visual-based safety evaluation of hazardous materials vehicle drivers. The proposed AHP-Entropy Weight fuzzy comprehensive evaluation model provides a scientific and reliable method for the safety management of hazardous materials road transportation and for personalized driver safety early warning systems.
Unraveling vulnerable road user crash severity: a latent class and random parameter approach with COVID-19 temporal insights.
Fuad N, Xiao RI, Qian X … +1 more , Jaller M
Traffic Inj Prev · 2026 Jun · PMID 42308496 · Publisher ↗
OBJECTIVE: This study investigates the determinants of vulnerable road user (VRU) crash severity by accounting for unobserved heterogeneity both across and within crash subpopulations. Conventional single-model approache...
OBJECTIVE: This study investigates the determinants of vulnerable road user (VRU) crash severity by accounting for unobserved heterogeneity both across and within crash subpopulations. Conventional single-model approaches assume homogeneous effects across all crashes, potentially masking context-dependent severity mechanisms. METHODS: Crash data for 15,578 pedestrian and 11,433 bicyclist collisions occurring at or near intersections in 20 California cities (2016-2025) were extracted from the Statewide Integrated Traffic Records System (SWITRS). A two-stage analytical framework was employed. First, latent class analysis identified three distinct crash typologies for each VRU mode based on movement patterns, lighting, weather, and collision factors. Second, mixed logit (MXL) models were estimated for each latent class to capture within-cluster heterogeneity through random parameters. Pseudo-elasticity analysis quantified the practical magnitude of variable effects. Temporal stability was assessed by estimating separate models across pre-COVID, during-COVID, and post-COVID periods. RESULTS: Truck involvement, dark conditions without streetlights, and state highway location consistently elevated severe outcome odds across all clusters for both VRU types, while VRU age 65+ shifted injury distributions toward moderate rather than the most severe outcomes. Critically, several factors exhibited context-dependent effects. VRU fault increased severity when drivers traveled straight, but decreased severity in turning-driver crashes for bicyclists, indicating fundamentally different causal mechanisms. State highway effects ranged from the strongest fatal predictor in straight-driver pedestrian crashes to non-significant in other configurations. Different random parameters were identified across clusters, confirming that unobserved heterogeneity operates through distinct mechanisms in different crash contexts. CONCLUSIONS: Crash severity determinants are both universally important and context-dependent, with the same variable capable of opposing effects across crash configurations. These findings demonstrate that aggregate models pooling heterogeneous crash types obscure critical variation and support the adoption of context-sensitive approaches to crash modeling.
Enhancing road safety with a power-aware approach to traffic sign detection and recognition in driving assistance systems based on machine learning using optimization techniques.
Pratap Singh K
Traffic Inj Prev · 2026 Jun · PMID 42308429 · Publisher ↗
OBJECTIVES: This study aims to improve road safety by creating a more effective deep learning model for traffic sign detection and recognition in driving assistance systems. The model, called Optimized Convo Sequential R...
OBJECTIVES: This study aims to improve road safety by creating a more effective deep learning model for traffic sign detection and recognition in driving assistance systems. The model, called Optimized Convo Sequential Recurrent Neuro Ant Hierarchical Net (OCSRN-AHN), combines convolutional and sequential methods with optimization strategies. This enhances recognition accuracy and processing speed in different environmental conditions. METHODS: A Kaggle dataset of 58 traffic sign classes was used to train and evaluate the model. The OCSRN-AHN model was trained using Python-based deep learning tools. This involved combining Convolutional Neural Network (CNN) and Sequential Recurrent Neural Network (SRNN) with Ant Hierarchical Network (AHN) optimization to fine-tune parameters. We evaluated model performance using metrics such as accuracy, precision, recall, F1-score, mean average precision (mAP), and mAP@0.5 (mean average precision at an intersection over a union threshold of 0.5). RESULTS: The OCSRN-AHN model achieved 99% accuracy, 99% precision, 99% recall, and an F1-score of 98%. Additionally, mAP@0.5 reached 98%, and the overall mAP was 98%. It performed better than existing models, including PFANet, YOLO v3, and YOLO v7t, in both detection accuracy and robustness for small, medium, and large traffic signs. CONCLUSIONS: The OCSRN-AHN framework not only achieved superior accuracy but also exhibited a more robust detection performance than all existing methods. Besides the higher detection accuracy, the proposed OCSRN-AHN has a reduced model complexity (3.2 M parameters), lower computational complexity (4.8 FLOPs), shorter inference time (21.4 ms), and lower estimated power consumption (0.39 J per image). These findings suggest the appropriateness of its use in real-time scenarios.
Tailoring the message: Communication strategies to promote safer driving behaviors among cannabis users.
Dell'Acqua R, Hacker S, Ageze D … +3 more , Baird S, Marcotte TD, Hill LL
Traffic Inj Prev · 2026 Jun · PMID 42308416 · Publisher ↗
OBJECTIVE: Cannabis-impaired driving is an increasing public health concern. Effective communication strategies are essential for shaping risk perceptions, influencing normative beliefs, and encouraging safer behaviors....
OBJECTIVE: Cannabis-impaired driving is an increasing public health concern. Effective communication strategies are essential for shaping risk perceptions, influencing normative beliefs, and encouraging safer behaviors. UC San Diego Transportation Research and Education for Driving Safety Center evaluated how cannabis consumers perceive and respond to cannabis-impaired driving messages, message sources, and strategies to promote safer driving behaviors. METHODS: Eligible participants were adults who reported cannabis use within the past three months, perceived it as safe to drive on the same day of use, and resided in one of eight U.S. states selected for diversity in cannabis policy contexts. The study aimed to recruit 800 participants. Using a cross-sectional, mixed-methods design, participants reported cannabis use patterns and driving behaviors before reviewing a series of cannabis-impaired driving messages. Messages were developed through an iterative process informed by literature, expert review, and formative testing, and represented communication styles commonly used in safety campaigns. Participants rated each message on attention, appeal, relevance, believability, influence on behavioral intentions, and source credibility. RESULTS: 846 cannabis users participated. Messages emphasizing concrete effects and impairment (e.g., ) consistently received the highest ratings across attention, appeal, relevance, and believability. In contrast, more informational or evaluative messages (e.g., ) performed significantly less favorably. Factual messages yielded the highest proportion of participants reporting they were to increase wait time before driving (59%), while self-reflective messages were most effective in encouraging alternative transportation (55%). All message types produced similar effects on intentions to remain in the same location (56-58%), and substantially fewer participants reported intentions to reduce cannabis use overall (23-29%). Message responsiveness varied by driving risk profile. Ultra-high-risk drivers reported lower likelihood of engaging in safer behaviors compared with medium- and high-risk drivers, although differences were not uniformly statistically significant. Qualitative findings indicated that exaggerated, fear-based, or stigmatizing messages were viewed negatively across groups, whereas messaging that was clear, evidence-based, and nonjudgmental was perceived as credible and effective. Source trust also varied, with healthcare providers and science-based organizations rated most credible and celebrities and social media influencers rated least trustworthy. CONCLUSION: This study offers new evidence on how cannabis users respond to messaging about cannabis-impaired driving. Messages emphasizing concrete effects and impairment, particularly those that are factual, direct, and evidence-based, were most consistently associated with higher ratings on outcome measures. However, differences in behavioral intentions across message types were modest, with limited impact on intentions to reduce cannabis use. Findings also indicate that individuals reporting higher-risk cannabis use and driving behaviors, particularly ultra-high-risk drivers, may be less responsive to messaging overall and may require more tailored approaches and complementary strategies to mitigate impaired driving. Across groups, messages perceived as exaggerated, stigmatizing, or fear-based were viewed as less credible and potentially counterproductive. These results can inform the development of user-centered, harm-reduction safety campaigns as cannabis legalization expands, while underscoring the need for continued evaluation and refinement of messaging strategies.
Why drivers rush into "safe-looking" short tunnels: the misleading effect of spatially intervisible.
Bei R, Sun W, Wang Y … +3 more , Wan H, Lyu N, Du Z
Traffic Inj Prev · 2026 Jun · PMID 42308413 · Publisher ↗
OBJECTIVE: The access zone of short tunnels exhibits an abnormally high accident rate, yet the underlying environmental causes remain unclear. This study identifies the key environmental feature behind this safety parado...
OBJECTIVE: The access zone of short tunnels exhibits an abnormally high accident rate, yet the underlying environmental causes remain unclear. This study identifies the key environmental feature behind this safety paradox and explains how it induces abnormal speed control. METHODS: We collected eye movement, driving performance, and video data from 30 drivers in 18 tunnel approach zones using a combined field and subjective approach. Subjective perception experiments identified "spatial intervisibility", the ability to see through the tunnel from the approach, as the critical feature. We compared two nearly identical tunnels differing only in intervisibility and applied Task Analysis of Driving Scenarios (TADS) to track behavioral evolution. RESULTS: Subjective results showed 77% of drivers perceived the intervisible short tunnel as most distinct. Objectively, despite similar recognition of speed‑limit signs, drivers in intervisible tunnels delayed deceleration onset by 41%, began slowing at a 6% higher speed, and reached the portal 6% faster. Speed compliance plummeted from 70% to 17%. Eye‑tracking revealed the mechanism: the see‑through tunnel increased fixations on itself by 26%, reducing attention to surroundings. CONCLUSIONS: The misleadingly open vista causes visual guidance failure, inducing delayed deceleration and speed violations that elevate collision risk. These findings directly support targeted interventions such as dynamic speed guidance for such tunnels.
Effect of medium and tall hood leading edge vehicle front-end characteristics on pedestrian torso injuries.
Hu J, Lin YS, Narayanan VS … +3 more , Mueller B, Olsson H, Jermakian J
Traffic Inj Prev · 2026 Jun · PMID 42308404 · Publisher ↗
OBJECTIVE: Field data analysis has shown that SUVs and pickup trucks cause more torso injuries than sedans, and the rapid increasing proportion of SUVs among the U.S. vehicle fleet will likely increase the importance of...
OBJECTIVE: Field data analysis has shown that SUVs and pickup trucks cause more torso injuries than sedans, and the rapid increasing proportion of SUVs among the U.S. vehicle fleet will likely increase the importance of pedestrian torso protection. The objective of this study is to use finite element (FE) vehicle and human body models to investigate effects of vehicle front-end geometry and stiffness characteristics on pedestrian injuries, specifically focusing on SUVs and pickup trucks and pedestrian torso injuries. METHODS: Front-end geometries of 74 U.S. vehicles, including 41 with hood leading edge (HLE) > 1000 mm and 33 with 750 mm < HLE < 1000 mm, were collected and analyzed using principal component analysis (PCA). The resulting parametric vehicle front-end geometry model was then linked to an FE generic vehicle (GV) model, so that the GV model can be morphed into a wide range of vehicle front-end geometries representing the fleet. Impact simulations using GHBMC F05, M50, and M95 pedestrian models and three detailed vehicle FE models were conducted with the pedestrian perpendicular to the vehicle front-end located at the center of the vehicle. These simulation results were used to calibrate the stiffness values and contact definitions of the hood and hood leading edge components of the morphed GV models. After GV model calibration, several parametric studies were conducted, resulting in a total of 306 vehicle-to-pedestrian crash simulations using 34 morphed GV models with varied front-end geometric and stiffness characteristics and three pedestrian models under three impact velocities (30, 40, and 50 kph). Pedestrian torso injuries were measured by lateral torso deflections at 17 locations across the chest and abdomen regions. Multiple regression was used to test the significance of the variables. RESULTS: PCA results showed that the top three principal components (PCs) captured over 90% of the variation in vehicle front-end geometries, primarily reflecting HLE height/length, HLE roundness, and overall front-end shape. Simulation results suggested that HLE height and impact velocity were the two dominant factors influencing pedestrian torso injury predictions. Torso injury metrics were the highest when the HLE height was equal to or slightly lower than (<150 mm) the pedestrian's mid-sternum height. In addition, increased HLE roundness and a more compliant HLE were associated with reduced pedestrian torso injuries. CONCLUSIONS: This study generated a comprehensive set of vehicle-to-pedestrian impact simulation data, enabling a systematic evaluation of how vehicle front-end geometric and stiffness characteristics influence pedestrian torso injuries.
Optimization of restraint system parameters for reclined drivers in frontal collisions based on finite element modeling.
Zhang D, Liu X, Yan J … +2 more , Lei Y, Zhang Y
Traffic Inj Prev · 2026 Jun · PMID 42308051 · Publisher ↗
OBJECTIVE: To address the increased injury risk for drivers in reclined postures during frontal collisions, this study aims to identify an optimal set of key restraint system parameters (including Belt Limit Force, Prete...
OBJECTIVE: To address the increased injury risk for drivers in reclined postures during frontal collisions, this study aims to identify an optimal set of key restraint system parameters (including Belt Limit Force, Pretensioner Firing Time, and Airbag Mass Flow Coefficient) for a 30° reclined seating position. The goal is to minimize upper body injury criteria (the Head Injury Criterion , the Neck Injury Criterion , and the Chest Deflection ) through a systematic optimization framework, thereby providing a quantitative design reference for restraint systems in autonomous driving scenarios. METHODS: This study constructed a finite element model of the driver restraint system by integrating the Hybrid III 50 percentile dummy, seatbelt, and airbag models, based on the finite element model of a certain type of sedan, and verified its validity. To simulate the reclined seating posture, the seatback angle was adjusted to 30°. Five restraint system parameters were first investigated through an orthogonal experimental design, and three key parameters were subsequently selected for response surface modeling and NSGA-II optimization. A second-order response surface model was established based on the experimental data, and the NSGA-II algorithm was combined to find the optimal parameters of the restraint system. RESULTS: The results show that the validation results showed good agreement between simulation and test responses, with CORA scores above 0.75 for all key channels. The optimal parameter combination is a belt limit force of 2000 N, a pretensioner firing time of 16.41 ms, and an airbag mass flow coefficient of 0.68. Compared with the initial parameters, the optimized restraint system parameters reduce the driver's by 22.73% (from 379.20 to 308.98), the by 11.96% (from 0.4101 to 0.3663), and the by 32.37% (from 25.19 mm to 19.03 mm). The maximum deviation between the response surface model and the simulation results is 12.20%. CONCLUSION: The model constructed in this study is effective and reliable, and the optimized parameters can significantly reduce the risk of upper body injuries for reclined drivers in frontal collisions. It provides a quantitative reference for the design of restraint systems for reclined postures in autonomous driving scenarios, lays a foundation for future research on safety protection for nontraditional postures, and is of great significance for improving the riding safety of intelligent vehicles.
About
- Frequency
- Sun
- Papers found
- 200
- RSS feed
- Subscribe