<rss version="2.0" xmlns:atom="https://www.w3.org/2005/Atom">
  <channel>
    <title>Transport Research International Documentation (TRID)</title>
    <link>https://trid.trb.org/</link>
    <atom:link href="https://trid.trb.org/Record/RSS?s=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" rel="self" type="application/rss+xml" />
    <description></description>
    <language>en-us</language>
    <copyright>Copyright © 2026. National Academy of Sciences. All rights reserved.</copyright>
    <docs>http://blogs.law.harvard.edu/tech/rss</docs>
    <managingEditor>tris-trb@nas.edu (Bill McLeod)</managingEditor>
    <webMaster>tris-trb@nas.edu (Bill McLeod)</webMaster>
    <image>
      <title>Transport Research International Documentation (TRID)</title>
      <url>https://trid.trb.org/Images/PageHeader-wTitle.jpg</url>
      <link>https://trid.trb.org/</link>
    </image>
    <item>
      <title>Prioritizing Railway Level Crossing Safety Equipment Using Stabilized and Explainable Machine Learning: A Large-Scale Study of the French Railway Network</title>
      <link>https://trid.trb.org/View/2717033</link>
      <description><![CDATA[This study develops a systematic analysis of Level Crossing configurations combining road and rail factors. It provides a decision-support framework for prioritizing safety equipment deployment at railway level crossings using explainable machine learning and robust clustering. Accident risk at level crossings emerges from the interaction of heterogeneous infrastructural, operational, and environmental variables, whose combined effects cannot be captured through single-factor analysis. Rather than focusing solely on prediction, the methodology combines dimensionality reduction, clustering algorithms, supervised classification, and SHAP-based explainability to identify homogeneous crossing typologies and classify them according to structural risk profiles. Separate analyses of road and rail characteristics enable clearer identification of their interdependencies and improve interpretability. Results show that the highest-risk clusters are characterized by the joint presence of high volumes of road and rail traffic, dense surrounding road environments, confirming the multifactorial nature of accident risk. The cluster-based approach mitigates the stochasticity of rare events and reveals robust comparative patterns that support comparative risk profiling of level crossing infrastructure. The study demonstrates how interpretable machine learning can be structured as a decision-aid framework for exploratory risk priorization for infrastructure management.]]></description>
      <pubDate>Wed, 23 Sep 2026 15:55:45 GMT</pubDate>
      <guid>https://trid.trb.org/View/2717033</guid>
    </item>
    <item>
      <title>In-vehicle multimodal alerts near active highway-rail grade crossings and their effects on driver behavior</title>
      <link>https://trid.trb.org/View/2711672</link>
      <description><![CDATA[Risky driving behaviors and distractions are one of the leading causes of vehicle-train accidents at Highway-Rail Grade Crossings (HRGCs), leading to immense loss of life and property. To improve driver behavior near active HRGCs, the current study investigated the effectiveness of in-vehicle visual and auditory alerts. Fifty-one participants experienced a preliminary intelligent alert about an upcoming rail crossing, an intermediate alert to slow down, and a final alert to stop before an active HRGC in a driving simulator study, across two sites. Results showed that both the intermediate and final alerts produced significant reductions in approach speed compared to no-alert conditions. Eye-tracking data showed that, without the intermediate alert, drivers spent more time fixating on the right peripheral roadway; when the intermediate alert was present, right-side fixation durations decreased, indicating a more evenly distributed visual scan. Overall, these findings demonstrate that sequential multimodal alerts can meaningfully improve driver behavior at active HRGCs. Implementing tiered, multimodal alerts in vehicles could enhance drivers' compliance with crossing warnings and encourage more distributed visual scanning, thereby potentially reducing collision risk at railroad crossings.]]></description>
      <pubDate>Thu, 18 Jun 2026 16:34:46 GMT</pubDate>
      <guid>https://trid.trb.org/View/2711672</guid>
    </item>
    <item>
      <title>Group Size and Social Capital Impact on Free Riding Behavior in Padang’s Self-Guards Managed Railway Crossing</title>
      <link>https://trid.trb.org/View/2676039</link>
      <description><![CDATA[An unguarded railway level crossing in Padang, Indonesia, presents a critical safety risk. This study investigates how community group size, social capital, and duration of residence in the neighborhood interact to influence collective action for shared safety. The collective action takes the form of conscious participation in an initiative to finance railway crossing guards, provided by and for the community. Using two-stage probit regression, the analysis uses the duration of residence in the neighborhood as an instrumental variable for social capital to address potential endogeneity. Not necessarily, the longer residents live in the neighborhood, the greater the willingness to participate. It is possible that living longer in an area leads to a decline in social capital due to economic stagnation and social fragmentation. Especially in disadvantaged neighborhoods. These dynamics illustrate the complex interactions between the duration of residence, group size, and social capital. These interactions will lead to a wide variety of responses in maintaining grassroots safety efforts.]]></description>
      <pubDate>Wed, 17 Jun 2026 12:23:23 GMT</pubDate>
      <guid>https://trid.trb.org/View/2676039</guid>
    </item>
    <item>
      <title>Crash occurrence and severity at railway level crossings in Bangladesh</title>
      <link>https://trid.trb.org/View/2667074</link>
      <description><![CDATA[Although train accidents are less frequent than road accidents, their consequences are often detrimental, especially at Railway Level Crossings (RLCs), the intersection points between rails and roads. RLC safety in developing countries like Bangladesh has received limited research attention despite the unique characteristics of these crossing types. This study aims to identify factors of occurrence and severity at RLCs in Bangladesh Railway East Zone from 2014 to 2021, during which crashes injured 89 and killed 54. An analysis of 121 RLCs across 18 districts examined how crossing, roadway, environmental, and exposure factors and crash reasons affect crash likelihood and severity. An extensive survey was conducted to build a structured database by collecting RLC-specific information. After accounting random effects, no significant heterogeneity was detected in either the occurrence or severity models. Logistic modeling revealed that train speed, C-type and illegal RLCs, warning sign-to-barrier distance, poor gate lodge visibility, track-to-lodge distance, divisional community areas, heavy vehicle percentage, geometric orientation, and the Sholoshohor-Fatehbad segment increased crash likelihood. In contrast, A-type RLCs, crossing width/length ratios, four-half barriers, speed bumps, crossing angles (<45°), residential areas, and standard straight sections decreased occurrence. Additionally, the ordered logit model demonstrated that train speed, truck involvement, traffic-controlled RLCs, warning sign distance (<28 feet), off-peak timings, train-hitting-vehicle collisions, gatemen operational failures, law violations, and reckless traversing significantly impacted severity. Policy priorities include standardizing warning sign placement (<28 feet), improving lodge visibility with paint and clear sightlines, installing signs at T-intersections, and utilizing four-half bamboo gates for safer RLC design.]]></description>
      <pubDate>Wed, 18 Mar 2026 09:00:01 GMT</pubDate>
      <guid>https://trid.trb.org/View/2667074</guid>
    </item>
    <item>
      <title>Improving nighttime visibility and safety at passive railway level crossings: New designs incorporating photoluminescent markings and signs</title>
      <link>https://trid.trb.org/View/2655593</link>
      <description><![CDATA[Railway level crossings (RLXs) equipped with passive protective measures, particularly prevalent in rural and regional areas, pose substantial nighttime safety risks due to poor visibility associated with the absence of electrical infrastructure. This study investigates new passive level crossing designs incorporating photoluminescent (glow-in-the-dark) markings and signs to address visibility issues at nighttime. A virtual reality (VR) driving simulator experiment with 20 participants was conducted to evaluate 3 new designs: Photoluminescent markings over 150 m (PL150), 500 m (PL500), and 500 m combined with photoluminescent signs (PL500 + S) compared against a standard passive RLX. Results demonstrated that the PL500 + S design significantly reduced approach speeds and increased compliance, with speeds 7.6 to 9.9 km/h lower within 200 m of the crossing and stopping compliance 2.7 times higher compared to the standard passive design. The PL500 design showed modest improvements, whereas the PL150 design had negligible impact. Furthermore, participants perceived the photoluminescent designs as safer than the standard RLX in nighttime conditions. The results of this study highlight the combination of photoluminescent markings and signs as cost-effective interventions for regional RLXs lacking electrical infrastructure.]]></description>
      <pubDate>Wed, 18 Feb 2026 11:59:29 GMT</pubDate>
      <guid>https://trid.trb.org/View/2655593</guid>
    </item>
    <item>
      <title>Time-to-event crash severity prediction at highway-rail grade crossings with monotonic neural networks</title>
      <link>https://trid.trb.org/View/2647548</link>
      <description><![CDATA[Despite advances in highway-rail grade crossing (HRGC) safety, including widespread use of active control devices, crashes at these intersections still lead to severe outcomes. Conventional crash prediction models often fail to capture severity-level dependencies, rely on assumption-driven and computationally intensive methods, and overlook links between severity and time between events. This study introduces a predictive framework based on positive monotonic neural networks (MNNs) for modeling time-to-crash outcomes with severity at HRGCs. Considering the relative newness of the time-to-crash paradigm in HRGC safety, the Neural Fine–Gray model is adopted as a core MNN implementation to estimate severity-specific crash likelihoods. This approach eliminates the numerical integration required in traditional time-to-event models, substantially reducing computational burden and accelerating training for large datasets. The framework naturally handles imbalanced HRGC data by treating event-free records as right-censored, avoiding the resampling required in traditional machine-learning approaches. To examine severity-level dependencies—an aspect largely overlooked in the literature—four MNN architectures are developed and evaluated. Using a 29-year North Dakota HRGC dataset, results show trade-offs between predictive accuracy and computational efficiency. The cause-specific MNN performs best for medium- and long-term horizons, whereas the multi-head MNN converges faster and excels at short horizons. Moreover, benchmarking against traditional time-to-event models—cause-specific Cox and Fine–Gray—shows modest calibration gains and 2%–50% stronger discrimination, reflecting the alignment between MNNs and the nonlinear, high-dimensional HRGC data. The framework also enhances interpretability by revealing paradoxical effects, including the “adding flashing lights paradox” and the “adding stop signs paradox.”]]></description>
      <pubDate>Fri, 06 Feb 2026 13:52:06 GMT</pubDate>
      <guid>https://trid.trb.org/View/2647548</guid>
    </item>
    <item>
      <title>Conceptual Framework for Improving Safety at Level Crossings in the European Union</title>
      <link>https://trid.trb.org/View/2627504</link>
      <description><![CDATA[Safety at level crossings (LCs) represents one of the most serious issues in railway transport within the European Union, given that approximately 26% of all fatalities in railway accidents occur at these locations. This paper analyzes available data from the Common Safety Indicators (CSI) and reports by the European Union Agency for Railways (ERA) to assess the current state of safety at LCs. The focus is placed on the number of accidents, their consequences, and the types of protection systems implemented at crossings. Based on identified patterns and indicators, the paper proposes a conceptual framework for improving safety, which includes strategic guidelines for the modernization of protection systems, the elimination of passive crossings, and the strengthening of institutional coordination in the field of safety management.]]></description>
      <pubDate>Fri, 09 Jan 2026 14:44:06 GMT</pubDate>
      <guid>https://trid.trb.org/View/2627504</guid>
    </item>
    <item>
      <title>Non-Contact Energy Harvesting for Rural Grade Crossings – Year 1</title>
      <link>https://trid.trb.org/View/2633303</link>
      <description><![CDATA[The network of US railroads often spans remote parts of the country that are sparsely populated. In these areas, rail grade crossings are much less likely to have warning lights or crossing gates primarily due to the lack of electricity. Such unprotected or passive crossings have the majority of the grade crossing fatalities and accidents. In order to reduce rail accidents, enhanced warning systems are needed at as many passive crossings as possible. We propose to create a new energy harvesting approach based on the motion of the wheels to generate sufficient power for an LED-based grade crossing warning system. Recent advances to create small and powerful magnets allow for the design of a non-contact power generation approach that is activated with each passing wheel. The feasibility of this approach has been shown in the first year of this project and an initial prototype will be designed and tested during the second year.]]></description>
      <pubDate>Mon, 15 Dec 2025 11:20:07 GMT</pubDate>
      <guid>https://trid.trb.org/View/2633303</guid>
    </item>
    <item>
      <title>Safety analysis at railway level crossing</title>
      <link>https://trid.trb.org/View/2612946</link>
      <description><![CDATA[Indian Railways has a complex track network that passes from plain areas, hilly areas, forests, above from the rivers, and tunnels. In context of railway and roadway, level crossing is a place where track cross the road. Due to the different transportation pattern, speed, and size of the vehicle, railway trains and roadway vehicles cannot run simultaneously. Level crossing is a place where frequent train to vehicle accidents occur, so trains are given priority over road vehicles to pass through the level crossing. To minimize the accidental risk and allow road vehicles to pass from the level crossing railway authority deployed some safety methods i.e., closure, merger, manning, subway. This paper describes a new safety method i.e., automatic level crossing gate as a 5th class to improve the decision tree prediction model performed on the dataset, and tries to conduct a comparative analysis of 4-class classification model and 5-class classification model. Dataset is taken from the authorized source of Indian Railway/CAMTECH/RDSO (2021). Automatic level crossing as a safety method tries to solve multiple conflicts of the safety decision as present in the dataset and improve the prediction model to provide suitable methods in the context of safety, facility and cost.]]></description>
      <pubDate>Wed, 19 Nov 2025 17:09:40 GMT</pubDate>
      <guid>https://trid.trb.org/View/2612946</guid>
    </item>
    <item>
      <title>Non-Contact Energy Harvesting for Rural Grade Crossings – Year 3</title>
      <link>https://trid.trb.org/View/2574027</link>
      <description><![CDATA[The network of United States railroads often spans remote parts of the country that are sparsely populated. In these areas, rail grade crossings are much less likely to have warning lights or crossing gates primarily due to the lack of electricity. Such unprotected or passive crossings have the majority of the grade crossing fatalities and accidents. To reduce rail accidents, enhanced warning systems are needed at as many passive crossings as possible. The research team  proposes to create a new energy harvesting approach based on the motion of the wheels to generate sufficient power for a ight-emitting diode (LED)-based grade crossing warning system. Recent advances to create small and powerful magnets allows for the design of a non-contact power generation approach that is activated with each passing wheel. The feasibility of this approach has been demonstrated and an initial prototype will provide data for optimization of the energy harvesting capabilities. ]]></description>
      <pubDate>Mon, 14 Jul 2025 19:06:39 GMT</pubDate>
      <guid>https://trid.trb.org/View/2574027</guid>
    </item>
    <item>
      <title>Pedestrian and Bicyclist Safety at Highway-Rail Grade Crossings (Year 1 Report)</title>
      <link>https://trid.trb.org/View/2479883</link>
      <description><![CDATA[Published literature is sparse on non-motorist (pedestrians and bicyclists) crashes at Highway-rail Grade Crossings (HRGCs), despite their involvement in rail-related crashes. A critical aspect of crash prediction models for HRGs is crash exposure, which measures activities of interest at specific locations. While data on motor vehicle and train traffic at HRGCs are available from different sources, non-motorist traffic counts are not readily available. Current Federal Railroad Administration (FRA) models focus on vehicular crash exposure, overlooking non-motorized traffic; gathering non-motorized traffic data is crucial for improving crash prediction models. In this study, non-motorist traffic videos were recorded for 1,848 hours from various urban and suburban HRGCs in Nebraska followed by application of the AIbased You Only Look Once Version 8 (YOLOv8) algorithm for automated non-motorist volume detection. Additionally, data on grade crossing characteristics, including population density and land use, were collected to create a comprehensive non-motorist database. Statistical and AI models were developed to analyze non-motorist exposure in terms of daily traffic volumes, utilizing physical, dynamic, and temporal characteristics of HRGCs. The models indicated that sidewalks, improved visibility, and cloudy weather conditions were associated with increased non-motorist traffic volume. Conversely, higher motor vehicle traffic levels, adverse weather conditions (rain and snow), industrial zones, and greater number of traffic lanes were linked with lower non-motorist traffic. This foundational study aims to enhance crash prediction models at HRGCs by incorporating non-motorist traffic factors, potentially improving crash prediction models and non-motorist safety at HRGCs.]]></description>
      <pubDate>Mon, 23 Dec 2024 17:04:51 GMT</pubDate>
      <guid>https://trid.trb.org/View/2479883</guid>
    </item>
    <item>
      <title>AI-Enabled Vibration Sensing System for Early Detection of Trains at Active Highway-Rail Grade Crossings</title>
      <link>https://trid.trb.org/View/2479887</link>
      <description><![CDATA[Highway-rail grade crossings (HRGCs) play an essential role in ensuring the secure traversal of road users across railway tracks. However, despite their significance, they present safety challenges, particularly when trains go undetected, heightening the risk of potential collisions between the road user and train. This study aims to explore the viability of employing vibration sensors for detection and characterization of an approaching train’s speed at HRGCs. The methodology involves analyzing rail vibrations and developing a time series predictive machine learning (ML) model. To accomplish this, a Finite Element (FE) model of a ballasted track railway is created in SAP2000, consisting of essential track components such as rails, sleepers, rail pads, ballast, and subgrade. The acceleration time history of the rails, induced by a moving single car, is recorded, and employed as input data for training and testing a Long Short-Term Memory (LSTM) network. The LSTM model predicts crucial parameters of the approaching train, including its location and speed. Successful prediction of these parameters is anticipated to enable the determination of the precise moment when the train reaches the HRGC, affording road users sufficient reaction time to safely traverse the railway track.]]></description>
      <pubDate>Mon, 23 Dec 2024 17:04:51 GMT</pubDate>
      <guid>https://trid.trb.org/View/2479887</guid>
    </item>
    <item>
      <title>Track Intrusion Detection and Track Integrity Evaluation</title>
      <link>https://trid.trb.org/View/2479882</link>
      <description><![CDATA[Track intrusion, particularly trespassing within railroad rights-of-way, poses a major safety risk, leading to more fatalities than train-vehicle collisions. This research introduces a Hybrid Region-based Convolutional Network (Hybrid-RCNN) that integrates foreground segmentation and object detection to enhance surveillance at rail crossings. The model performs multiple tasks, including object detection, classification, and tracking, to efficiently monitor unauthorized activities. Testing shows the HybridRCNN's superior accuracy compared to models like Mask-RCNN, highlighting its potential to improve railway safety by identifying and mitigating hazards more effectively.]]></description>
      <pubDate>Mon, 23 Dec 2024 17:04:51 GMT</pubDate>
      <guid>https://trid.trb.org/View/2479882</guid>
    </item>
    <item>
      <title>Development of On-board-based Autonomous Train Control Systems</title>
      <link>https://trid.trb.org/View/2458862</link>
      <description><![CDATA[As a basic technology for realizing autonomous train operation, the authors propose two functions: a function for creating a running profile on board a train and a function for creating a running profile on board to recover from delays caused by hazards. Running profiles are generated based on information such as timetables. The function for direct control of ground equipment from a train is designed to allow on-board equipment to set train routes, control turnouts on that route, and control level crossing warnings as trains approach.]]></description>
      <pubDate>Mon, 02 Dec 2024 12:49:31 GMT</pubDate>
      <guid>https://trid.trb.org/View/2458862</guid>
    </item>
    <item>
      <title>Predictability of driver’s stop/go decisions at flashing-light-controlled grade crossings by coupling functional brain network and deep learning methods</title>
      <link>https://trid.trb.org/View/2426592</link>
      <description><![CDATA[Detecting and predicting the stop/go decisions of drivers at grade crossings is crucial for enhancing road safety. Electroencephalography (EEG) data, which provides direct and effective physiological indicators for recognizing driver states, combined with associated machine-learning techniques, can be used to monitor driver decisions. However, the ability of EEG to predict a driver’s stop/go decisions remains unclear. To investigate this, the authors collected both EEG and behavioral data from drivers at a flashing-light-controlled grade crossing, where stop/go decisions are critical, using a driving simulator. Herein, the authors propose an EEG-based prediction framework that combines functional brain network analysis with conventional neural networks (FBN-CNNs) to predict drivers’ stop/go decisions. The functional brain network was measured using phase-lag index matrices and minimum-spanning tree techniques. The authors subsequently compared the obtained results of the FBN-CNN with those from traditional machine learning methods, specifically random forest (RF) and Support Vector Machines (SVM). The results indicate that when facing a flashing red light, drivers who decide to stop exhibit stronger alpha band connectivity and weaker delta and theta activity than those who run the red-light. Furthermore, the FBN-CNN model outperformed the machine learning methods (RF and SVM) in both extracting EEG features and achieving high prediction accuracy. Interestingly, the EEGs of drivers during normal driving stages could help to predict their stop-or-go behavior at the onset of a flashing red light. In the typical dilemma zone, combining EEG data from the normal driving stage with those from the pre-decision stage improved the accuracy from 76% to 90%. These findings demonstrate the efficacy of EEG and deep learning methods in driver decision monitoring.]]></description>
      <pubDate>Fri, 04 Oct 2024 13:51:43 GMT</pubDate>
      <guid>https://trid.trb.org/View/2426592</guid>
    </item>
  </channel>
</rss>