Search Results (60)

Crash severity prediction is critical for emergency response, infrastructure spending, and risk communication. Although machine learning has been widely applied to this problem, three gaps prevent practical deployment: uncalibrated probability scores, SHAP-based explanations whose faithfulness has not been verified, and models never tested outside their training jurisdiction. The proposed framework, SAE-XCrash (Safety-Aware and Explainable Crash Severity Prediction), addresses all three using two public datasets—US-Accidents (7.0 million records, 2016–2023) and UK STATS19 (approximately 1,010,000 records, 2016–2022)—with strict temporal splits throughout. Notably, the US-Accidents severity label measures traffic disruption duration, not injury outcome; results should be interpreted accordingly. Previously unknown label-schema drift led to a revised binary target with Severity 4 as the only positive class. Five classifiers are compared. Post hoc isotonic calibration reduces Expected Calibration Error by 97.3% at negligible discrimination cost. A four-step quantitative SHAP audit confirms statistically significant deletion faithfulness; however, explanation stability fails at realistic perturbation levels (54.3% low-stability fraction at sigma = 0.05), driven by spatial data sparsity in sparse geohash cells—a negative result that carries direct operational implications for deployment. A three-tier cross-dataset transfer experiment (zero-shot, recalibration, full retrain) shows that temporal features transfer robustly across jurisdictions, while spatial memorization is the primary generalization barrier. All code, split indices, and model artifacts are publicly available. Full article

Crashes involving freight and commercial vehicles impose substantial human and economic costs, yet most severity studies pool vehicle types or focus exclusively on heavy trucks, masking class-specific risk mechanisms. This study estimates separate Random Parameters Binary Logit models with heterogeneity in means and variances for three vehicle categories—heavy-duty multi-axle trucks (n = 6512), two-axle trucks (n = 2656), and light-duty pickup trucks (n = 23,477)—using 32,645 crash records from Thailand’s national highway network (May 2022–December 2024). Pairwise transferability tests rejected parameter transferability, with four of six comparisons exceeding the 97 percent confidence level (three of these above 99 percent; χ² = 85.38 to 240.01), confirming that disaggregate estimation is statistically warranted. Three core findings emerge: First, although barrier medians, cut-in-front maneuvers, and sideswipe crashes affect severity in consistent directions across all vehicle types, their magnitudes differ sharply: the protective effect of barrier medians is nearly six times larger for two-axle trucks (ME = −0.160) compared to heavy-duty trucks (ME = −0.028). Second, several determinants are class-specific: dark unlit conditions elevate severity only for two-axle trucks (ME = 0.128), flush medians only for heavy-duty trucks (ME = 0.040), and raised medians only for light-duty pickups (ME = 0.042). Third, no random parameter is common to all three models. Pooled models, therefore, impose misleading homogeneity assumptions; vehicle-type-specific estimation is essential for targeted safety policy. Full article

Until now, there have been no published attempts to utilize ensemble learning approaches to pre-crash velocity estimation. In this research article, we focus on the method of vehicle crash velocity prediction based on the random forest regression approach. In particular, the study aims to develop and validate a random forest-based non-linear model for estimating pre-crash velocity using EES-related parameters for compact vehicles in a crash scenario against an immovable, stationary barrier. The estimation technique is trained and evaluated using the compact vehicle class from the NHTSA database, which consists of 399 records of frontal impacts against a rigid barrier. The relative error obtained for the presented calculation method is $7.57\mathrm{\%}$, with absolute error being equal to $1.12 \mathrm{m}/\mathrm{s}$. We subsequently compare our results with some other techniques which were tested on this dataset. Despite the simplicity of random forest regression, we obtain surprisingly good results, as the method outperforms linear regressor and artificial neural network predictors, which have relative errors of $8.17\mathrm{\%}$ and $9.63\mathrm{\%}$, respectively. The independence of Event Data Recorders along with the ease of obtaining the necessary data makes the proposed approach a highly desirable tool in forensic analysis, especially in cases involving older vehicles. Full article

Movable guardrail systems are increasingly used in work zones, reversible lanes, and temporary traffic operations; however, evidence on their crashworthiness, material performance, and operational reliability remains dispersed across multiple design typologies and regulatory frameworks. This PRISMA-compliant systematic review synthesizes 78 studies involving full-scale crash tests, validated finite-element simulations, field performance evaluations, and compliance evaluations under MASH, EN 1317, NCHRP 350, and AS/NZS 3845.1. The findings indicate that modular rigid barriers reliably achieve TL-3/TL-4 performance when joint alignment and foundation conditions are properly controlled; semi-rigid steel systems provide a practical balance between containment capacity and redeployability, but remain sensitive to post spacing and connector detailing; and flexible polymer systems are best suited for short-duration, low-speed applications. Material-focused research highlights the advantages of UHPC section refinement, high-strength steels, and hybrid FRP–metal configurations in enhancing energy absorption without exceeding occupant-risk thresholds. Across studies, connection integrity consistently emerges as the dominant factor governing redirection stability and working-width performance. Field evaluations confirm satisfactory operational performance in constrained environments, while life-cycle assessments identify refurbishment intervals and mass-related logistics as major cost contributors. This review provides an integrated, evidence-based synthesis and a structured engineering foundation for advancing next-generation movable barrier designs, testing protocols, and deployment strategies. Full article

Median crossover crashes are among the most severe roadway safety events due to their high-energy nature and strong association with fatal and incapacitating injuries, posing a substantial public health burden. This study develops a multimodal intelligent analytics framework to evaluate the cable median barrier performance in Tennessee by integrating structured crash data, roadway and traffic characteristics, post-impact vehicle responses, and unstructured police narratives. Across 6094 crashes on 576 cable barrier segments, 1196 involved barrier impacts and 914 included complete post-impact response information. Deep learning-based text mining using a BERT transformer model was applied to narrative descriptions from fatal, serious injury, and minor injury crashes to extract contextual indicators of loss of control, impact dynamics, and injury mechanisms. Safety effectiveness evaluation using Empirical Bayes methods showed substantial reductions after installation, including a 96% decrease in fatal crashes and an 88% reduction in serious-injury crashes. Vehicle–barrier interactions—classified as containment, redirection, rollover, or penetration—were modeled using a multinomial logit framework with marginal effects to assess the influence of geometric, operational, and vehicle-related factors. Reduced barrier offset, narrow shoulders, high traffic volumes, outer-lane departures, and heavy-vehicle involvement significantly increased the likelihood of rollover and penetration events, which are strongly linked to higher injury severity. Through fusing multimodal data and combining explainable statistical models with deep learning text analysis, this study provided a scalable, trustworthy approach to characterizing injury risk, aligning transportation safety analytics with emerging intelligent healthcare and big-data methodologies aimed at preventing severe and fatal trauma. Full article

This study examines crash involvement, safety training exposure, and e-learning readiness among commercial heavy goods vehicle (HGV) drivers in Ethiopia. Data were collected through a cross-sectional survey of 202 male drivers operating along the Addis Ababa–Djibouti trade corridor, a high-risk freight route that carries approximately 95% of Ethiopia’s international trade and serves as the country’s primary gateway to global markets. The survey assessed crash history, safety training experiences, perceived safety challenges, and barriers to and motivators for e-learning adoption. Results indicate persistently high crash involvement despite widespread participation in conventional classroom-based training, suggesting a gap between training provision and real-world safety outcomes. Older and mid-career drivers exhibited higher crash involvement, highlighting a gap between training provision and behavioral or operational safety outcomes, while younger and more educated drivers showed greater readiness for technology-enhanced training. Although most drivers valued safety training, many perceived existing programs as repetitive, insufficiently interactive, and poorly aligned with operational demands. Key facilitators for e-learning adoption included flexible schedules, ease of use, and motivational support, whereas limited digital skills and low perceived usefulness remained barriers for some groups. The findings highlight the need for age-responsive, flexible, and interactive e-learning approaches to complement traditional training and address persistent safety risks, such as fatigue and unsafe driving behaviors. These approaches also support scalable, technology-enhanced interventions tailored to Ethiopia’s high-risk freight corridors, while guiding future research directions. Full article

Noise barriers are commonly used to reduce the adverse effects of traffic noise in both urban and suburban settings. While conventional systems constructed from concrete and steel provide reliable acoustic and structural performance, they raise sustainability concerns due to high embodied energy and carbon emissions. Glued-laminated (glulam) timber has emerged as a sustainable alternative, offering a reduced carbon footprint, aesthetic appeal, and effective acoustic performance. However, the crashworthiness of timber-based noise wall systems remains under investigated, particularly with respect to the safety criteria established in the 2016 edition of the American Association of State Highway and Transportation Officials (AASHTO) Manual for Assessing Safety Hardware (MASH). This study presents the full-scale crash testing and evaluation of glulam rubrail and noise wall systems under MASH Test Level 3 (TL-3) impact conditions. Building on a previously tested system compliant with National Cooperative Highway Research Program (NCHRP) Report 350, modifications were made to increase rubrail dimensions to meet higher lateral design loads. Three full-scale vehicle crash tests were conducted using 1100C and 2270P vehicles at 100 km/h and 25 degrees, covering both front- and back-mounted wall configurations. All tested systems demonstrated acceptable structural performance, effective vehicle redirection, and compliance with MASH 2016 occupant risk criteria. There was no penetration or potential for debris intrusion into the occupant compartment, and all measured occupant risk values remained well below allowable thresholds. Minimal damage to structural components was observed. The results confirm that the modified glulam noise wall system meets current impact safety standards and is suitable for use along high-speed roadways. This work supports the integration of sustainable materials into roadside safety infrastructure without compromising crash performance. Full article

Upright-posture motorcycle crashes against steel safety barriers (SSBs) often result in severe upper-body injuries due to the sharp upper edge of the rail. While solutions for sliding crashes on curves, called a ‘motorcyclist-friendly barrier’, are already implemented in practice, protective measures for upright-posture impacts remain underdeveloped. This study systematically reviews patents and patent applications addressing upper guardrail protection for motorcyclists. We identified and analysed a small number of existing innovations aimed at mitigating the consequences of upright crashes. The selected solutions were evaluated according to their technical design, ease of installation, potential for recycling, environmental compatibility, and expected costs. Our comparative analysis reveals that while some patents or patent applications offer promising features, such as flexible caps, bent plates, or modular attachments, none comprehensively address all safety, environmental, and economic requirements. The findings provide a basis for further development of motorcyclist-friendly SSB designs and suggest specific criteria that should be included in future guidelines and standard updates. Full article

The accurate modeling of dynamic soil–structure interaction processes under impact loading is critical for advancing the design of soil-embedded barrier systems. Full-scale crash testing remains the benchmark for evaluating barrier performance; however, such tests are costly, logistically demanding, and subject to variability that limits repeatability. Recent advancements in computational methods, particularly the development of large-deformation numerical schemes, such as the multi-material arbitrary Lagrangian–Eulerian (MM-ALE) and smoothed particle hydrodynamics (SPH) approaches, offer viable alternatives for simulating soil behavior under impact loading. These methods have enabled a more realistic representation of granular soil dynamics, particularly that of the Manual for Assessing Safety Hardware (MASH) strong soil, a well-graded gravelly soil commonly used in crash testing of soil-embedded barriers and safety features. This soil exhibits complex mechanical responses governed by inter-particle friction, dilatancy, confining pressure, and moisture content. Nonetheless, the predictive fidelity of these simulations is governed by the selection and implementation of soil constitutive models, which must capture the nonlinear, dilatant, and pressure-sensitive behavior of granular materials under high strain rate loading. This review critically examines the theoretical foundations and practical applications of a range of soil constitutive models embedded in the LS-DYNA hydrocode, including elastic, elastoplastic, elasto-viscoplastic, and multi-yield surface formulations. Emphasis is placed on the unique behaviors of MASH strong soil, such as confining-pressure dependence, limited elastic range, and strong dilatancy, which must be accurately represented to model the soil’s transition between solid-like and fluid-like states during impact loading. This paper addresses existing gaps in the literature by offering a structured basis for selecting and evaluating constitutive models in simulations of high-energy vehicular impact events involving soil–structure systems. This framework supports researchers working to improve the numerical analysis of impact-induced responses in soil-embedded structural systems. Full article

As a key roadside safety feature, longitudinal guardrail steel barriers are purposefully designed to contain and redirect errant vehicles to prevent roadway departure, dissipate impact energy through plastic deformation, and reduce the severity of vehicle crashes. Nevertheless, these systems should be carefully designed and assessed, as localized rupturing, especially near splice or impact locations, can lead to catastrophic failures, compromising vehicle containment, violating crash safety standards, and ultimately jeopardizing the safety of occupants and other road users. Before conducting full-scale crash testing, finite element analysis (FEA) tools are widely employed to evaluate the design efficiency, optimize system configurations, and preemptively identify potential failure modes prior to expensive physical crash testing. To accurately assess system behavior, calibrated material models and precise failure criteria must be utilized in these simulations. Despite the existence of numerous failure criteria and material models, the material characteristics of AASHTO M-180 guardrail steel have not been fully investigated. This paper significantly advances the FE modeling of ductile fracture in guardrail steel, addressing a critical need within the roadside safety community. This study formulates stress-state-dependent failure criteria and proposes advanced material modeling techniques. Extensive experimental testing was conducted on steel specimens having various triaxiality and Lode parameter values to reproduce a wide spectrum of complex, three-dimensional stress-state loading conditions. The test results were then used to identify material properties and construct a failure surface. Subsequent FEA, which incorporated the Generalized Incremental Stress-State-Dependent Damage Model (GISSMO) in conjunction with two LS-DYNA material models, illustrates the capability of the developed surface and material input parameters to predict material behavior under various stress states accurately. A parametric study was completed to further validate the proposed models, highlighting their robustness and reliability. Full article

Concrete median barriers are prone to damage from low-velocity impacts. However, there is a limited understanding of how damage from initial impacts affects barriers’ long-term performance and whether they maintain safe continued service or must be replaced. Therefore, this paper evaluates the performance of the concrete barriers under sequential low-velocity impact using finite-element analysis. Crash test simulations were performed by impacting the concrete barrier twice with an 80,000 lb (36-ton) tractor-trailer at a target impact velocity and angle. The first impact’s velocities varied between 30 mph (48 kmph) and 54 mph (87 kmph) at 10°, 15°, and 20° crash angles, and the damaged barrier was subsequently subjected to the second impact conforming to the American Association of State Highway and Transportation Officials’ (AASHTO) Manual for Assessing Safety Hardware (MASH) for Test Level 5 criteria (i.e., representative velocity of 52.7 mph (85 kmph) at 15°). Therefore, a total of 78 impact simulations were conducted, and statistical analysis was performed to investigate the relationship between the peak impact forces of the first and second impacts and the crash angle and velocity across distinct phases of the crash simulation and over the entire crash history. The results show that while the peak impact force of the first impact was linearly related to both velocity and angle, the maximum impact force at the second impact did not follow the same trend. However, when considering the localized peak forces in each phase of the crash, the peak forces from the later stages of the second impact (i.e., rebound and final interaction phases) were highly correlated with the initial impact’s velocity and angle, substantially reducing the barrier’s capability to resist vehicular impact loads. In particular, for initial velocities above 46 mph (74 kmph) at angles of 15° and 20°, barriers formed shear cracks traversing across their cross-section, which resulted in excessive fragmentation during the second impact and consequent failure to meet the MASH criteria in terms of structural adequacy. Full article

Reinforced Concrete (RC) barriers are used for different purposes in the highway inventory. An important purpose is the use of concrete barriers to act as railing that protects bridge piers against vehicular collision force (VCF). Therefore, these barriers are designed to absorb the collision energy and/or redirect the vehicle away from the parts being protected. Accurate estimation of the capacity of RC barriers during crash events is an important consideration in their design and placement. The American Association of State Highway and Transportation Officials (AASHTO) considers yield line analysis (YLA) with the V-shape failure pattern to predict the barrier capacity. AASHTO’s analysis method involves some assumptions that are intended to simplify the analysis process. Some of these assumptions have been shown to underestimate the actual barrier capacity and might disqualify many existing RC barriers from acting as intervening structures due to structural inadequacy. Many researchers have proposed alternative failure patterns and methodologies in an attempt to better predict the capacity of RC barriers. This research shows that AASHTO’s YLA, with the current V-shape failure pattern, can be improved and still predict the barrier capacity when some of the simplifying assumptions are eliminated. Also, the research presents an alternative innovative truss analogy model to predict the capacity of RC barriers. The results of the improved YLA and the proposed truss model are validated by finite element analysis (FEA) using Abaqus. The results of this research will help structural engineers in the highway industry to initially design new barriers for the intended capacity as well as estimate the capacity of existing ones. Full article

Cor-Ten steels, also known as weathering steels, are construction materials of growing importance in the field of architecture and crash barriers, not only due to their good mechanical and corrosion resistance properties but also for the appealing color of their oxides. However, a complete description of the fracture locus of Cor-Ten steels in both low and high triaxiality ranges is still lacking. The present study aims at integrating and extending the data available in the literature for this peculiar material by evaluating four different planar specimens with a mixed numerical–experimental methodology. A non-notched specimen was tested in terms of tension to calibrate the true stress–strain curve of the material after necking by means of an iterative process involving the FEM. Once the model had been calibrated, a tensile test of each specimen was simulated, and the corresponding results were validated using the experimental test data. From the FEM results, the quantities of interests, namely, the stress triaxiality, the equivalent plastic strain, and the normalized Lode angle, were extrapolated. Subsequently, the fracture locus of the Cor-Ten steel was determined through the interpolation of the experimental data collected in the present study as well as data available in the literature for low triaxiality ranges. The results confirmed the parabolic trend characterizing the fracture locus at low triaxiality suggested in the literature, and an exponential decreasing trend was found at higher triaxiality values after reaching a local maximum. The results thus confirm that the fracture locus of Cor-Ten steels, as generally found for metallic materials, cannot be completely described by a monotonic function. Moreover, it was found that the highly ductile behavior of the material induces a significant topology change in the specimens before failure, thus making it more complex to forecast the location of crack nucleation and, as a consequence, the stress state. Full article

Background: Mobile health (mHealth) applications play a crucial role in enhancing healthcare accessibility, patient engagement, and chronic disease management. However, technical, usability, accessibility, and privacy-related barriers continue to hinder their widespread adoption. The Sehaty app, a government-managed mHealth platform in Saudi Arabia, is widely used for scheduling medical appointments, accessing health records, and communicating with healthcare providers. Understanding the challenges associated with its utilization is essential for optimizing its functionality and improving user experience. Aim: This study aims to identify and evaluate the key barriers affecting the adoption and usability of the Sehaty mHealth application among patients with chronic conditions in Saudi Arabia. Specifically, it examines challenges related to technical performance, usability, accessibility, privacy, and security and their impact on user satisfaction and engagement. Methods: A cross-sectional study was conducted using a structured questionnaire distributed to 344 participants selected through purposive sampling to ensure the inclusion of active Sehaty users with chronic conditions. The questionnaire assessed 10 primary usability barriers, including technical issues, navigation difficulties, privacy concerns, and accessibility limitations. Descriptive statistics and correlation analyses were performed to evaluate the prevalence and interrelationships of these barriers. Results: The findings indicate that technical barriers, including frequent application crashes, slow responsiveness, and system instability, significantly hinder user satisfaction. Usability challenges, such as difficulties in navigation and task completion, further impede engagement. Moreover, privacy and security concerns emerged as significant deterrents, with users expressing apprehensions about data safety and transparency. Accessibility barriers, particularly for older adults and individuals with disabilities, were associated with insufficient support and training, making the app less user-friendly for these populations. The study highlights the interconnected nature of usability challenges, suggesting that improvements in technical stability and interface design could lead to enhanced user confidence, engagement, and overall satisfaction. Conclusions: Addressing these barriers requires targeted technical enhancements, user-centered design improvements, and strengthened data security measures to promote trust and engagement. Additionally, implementing comprehensive user support systems and accessibility features is essential to ensuring equitable access to mHealth services. While the study’s generalizability is limited by its focus on a single government-managed platform, its findings offer valuable insights applicable to broader mHealth initiatives. Future research should incorporate longitudinal studies to assess the long-term impact of usability improvements on mHealth adoption and healthcare outcomes. Full article

Electric vehicles (EVs) are central to global efforts in reducing carbon emissions and transitioning toward sustainable transportation. This literature review emphasizes the critical role of safety in advancing EV adoption in Indonesia by drawing insights from global advancements and addressing local challenges. Key findings highlight that while EVs promise significant environmental benefits, safety concerns, such as battery thermal runaway risks and structural reliability in diverse road and climatic conditions, remain significant barriers. Issues such as battery safety, including thermal runaway risks, and the reliability of structural designs in Indonesia’s diverse road and climatic conditions are pivotal. Globally, advancements in battery management systems (BMS), crash-resistant vehicle designs, and autonomous driving technologies provide effective pathways to mitigate these safety risks. Locally, the development of safety standards tailored to tropical climates and robust infrastructure is essential. Leveraging Indonesia’s natural resources, such as nickel, offers opportunities to produce safer and cost-effective batteries. Additionally, policy frameworks like Presidential Regulation No. 55/2019 must prioritize safety measures, including rigorous testing, recycling protocols, and public education. This study concludes by advocating for an integrated approach that combines technological innovation, enhanced safety features, and supportive policies to accelerate EV adoption in Indonesia. Future research should focus on improving safety technologies, lifecycle assessments, and renewable energy integration to ensure the long-term success of EV adoption in the country. Full article