<rss version="2.0" xmlns:atom="https://www.w3.org/2005/Atom">
  <channel>
    <title>Research in Progress (RIP)</title>
    <link>https://rip.trb.org/</link>
    <atom:link href="https://rip.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>Research in Progress (RIP)</title>
      <url>https://rip.trb.org/Images/PageHeader-wTitle-RIP.jpg</url>
      <link>https://rip.trb.org/</link>
    </image>
    <item>
      <title>High Load Hit Prevention</title>
      <link>https://rip.trb.org/View/2562331</link>
      <description><![CDATA[Many department owned bridges are impacted with over height vehicles every year, however, there are many bridges that
have been struck repeatedly. Bridges being repeatedly struck by over height vehicles leads to structural damage to
department infrastructure as well as impedance to the traveled roadway(s). This damage can cause immediate lane or road
closures while the damage is inspected and repaired, shortened service life of the structure, and potential bridge component or
structure replacement. In order to address this problem, the department seeks to identify methods to locate and inform over
height drivers prior to striking the structure in addition to detecting impacts. Not having these methods will allow over height
vehicles to continue to damage department structures and for damage to go unreported. Michigan Department of Transportation (MDOT) has interest in identifying
which method(s) of over height vehicle impact prevention is best suitable for structures susceptible to damage. The
department is expecting this research to yield the implementation of technology to reduce damage of bridges caused by over
height vehicles impacts. Additionally, Michigan has dozens of bridges located within navigable waterways and following the
collapse of the Francis Scott Key bridge in Baltimore, MDOT would like to evaluate the inherent risk of damage from vessel
allision at those bridges with substructure units within the waterway.]]></description>
      <pubDate>Fri, 17 Jul 2026 10:22:15 GMT</pubDate>
      <guid>https://rip.trb.org/View/2562331</guid>
    </item>
    <item>
      <title>Machine Vision Toolkit for Automated Fleet Composition Assessment and Reporting</title>
      <link>https://rip.trb.org/View/2691665</link>
      <description><![CDATA[State Departments of Transportation (DOTs) and Metropolitan Planning Organizations (MPOs) employ fleet composition data (e.g., passenger vehicles, single-unit trucks, and combination trucks) in a variety of planning, economic, roadway performance, and safety applications. Accurate fleet composition data is essential for pavement management, safety analysis, and fuel consumption modeling. However, traditional methods are labor-intensive, costly, and often lack the temporal or spatial resolution required to capture variations between freeways, arterials, and managed lanes vs. general-purpose lanes. Using machine vision tools to quickly, efficiently, and accurately capture on-road percentages of light-duty vehicle, light-duty truck, medium-duty truck, and a variety of heavy-duty truck classifications will enhance analytical and modeling accuracy and reduce state DOT data management costs. Building upon prior National Center for Sustainable Transportation (NCST) research that developed machine vision algorithms for vehicle identification, this project will package those research findings into a deployable, open-source Automated Fleet Classification Toolkit for practitioners and researchers. The research team will develop and release comprehensive Standard Operating Procedures (SOPs) and software tools allowing agencies to convert standard roadside or overpass video feeds into high-resolution fleet composition data. The toolkit will utilize advanced object detection (e.g., YOLO architectures) to automate the identification of vehicle classes (aligning with FHWA 13-category schemes where possible) and propulsion types based on visual vehicle features. The system is designed to distinguish traffic conditions on complex roadway geometries, allowing users to generate separate classification profiles for managed lanes vs. general-purpose lanes, and separating freeway mainlines from adjacent arterial service roads. The project focuses on technology transfer: providing the "how-to" manuals, open-source code, and data processing protocols so that State DOTs, consultants, university partners and research institutes can replicate the data collection and extraction without relying on proprietary "black box" services.]]></description>
      <pubDate>Sun, 12 Apr 2026 23:29:28 GMT</pubDate>
      <guid>https://rip.trb.org/View/2691665</guid>
    </item>
    <item>
      <title>LLM-Orchestrated Multi-Layer Digital Twin Network for Cyber-Resilient Traffic Management</title>
      <link>https://rip.trb.org/View/2663602</link>
      <description><![CDATA[Modern connected traffic systems are increasingly vulnerable to cyberattacks capable of propagating rapidly across networked infrastructure, inducing unsafe signal states, traffic congestion, and emergency response delays. Existing anomaly detection approaches including statistical thresholds, rule-based Automated Traffic Signal Performance Measures (ATSPM) and Signal Phase and Timing (SPaT) flags, and classical machine-learning methods such as Isolation Forest and one-class Support Vector Machines operate on limited data modalities and cannot capture cross-layer cyber-physical interactions or operator intent, leaving critical detection gaps in complex attack scenarios.
This project develops a distributed multi-layer digital twin (DT) network for urban traffic systems, enhanced by a large language model (LLM) for context-aware cyber anomaly detection. The framework mirrors physical traffic behavior, cyber infrastructure status, and operational decision processes across a corridor of 4–6 interconnected intersections, enabling early identification of unsafe and malicious events that threaten roadway safety. Each traffic unit is represented by coordinated Physical, Cyber, and Decision Layers: the Physical Layer models real-time mobility and safety conditions using ATSPM, SPaT/MAP data, and detector activity; the Cyber Layer mirrors controller firmware, communication telemetry, and roadside unit status; and the Decision Layer captures operator actions, timing plan updates, and agency-defined safety constraints. A customized transportation-aware LLM ingests both structured telemetry and unstructured logs to generate semantic feature embeddings that capture cross-layer and cross-node dependencies.
A hybrid neural anomaly detection engine integrates Temporal Convolutional Networks (TCNs) to learn evolving traffic and communication behaviors over time with Graph Neural Networks (GNNs) to capture spatial interactions and coordinated disruptions across interconnected intersections. This TCN–GNN architecture enables accurate recognition of both localized cyber intrusions and distributed corridor-level attacks. Detection performance is validated against controlled cyber-attack scenarios—including SPaT spoofing, firmware manipulation, and malicious timing-plan overrides—executed within the DT environment. Upon anomaly detection, the LLM generates actionable mitigation suggestions, such as isolating compromised controllers or reverting to safe fallback signal plans, which are evaluated within the digital twin to ensure that every recommendation supports operational safety, low latency, and service continuity.
The 12-month effort proceeds in two phases: development and calibration of the distributed multi-layer DTs with LLM integration for context modeling, followed by anomaly detection training, validation, and mitigation evaluation. Target performance metrics include detection accuracy of at least 90%, false-positive rates below 10%, decision-support latency improvements of at least 30%, and safety metric improvements of at least 20%. The project delivers a pilot-ready prototype, detailed deployment guidelines, and an open software repository to accelerate adoption by transportation agencies. 
]]></description>
      <pubDate>Tue, 03 Feb 2026 15:28:33 GMT</pubDate>
      <guid>https://rip.trb.org/View/2663602</guid>
    </item>
    <item>
      <title>Adaptive Cyber Threat Detection for Rail SCADA Systems: A Hybrid Machine Learning and Statistical Approach</title>
      <link>https://rip.trb.org/View/2655702</link>
      <description><![CDATA[Supervisory Control and Data Acquisition (SCADA) systems form the digital backbone of modern railway operations, enabling real-time monitoring of critical track geometry parameters including gage, cross-level, alignment, and warp that are essential for preventing derailments and ensuring passenger safety. While SCADA-driven sensing has advanced continuous condition monitoring, it has also introduced new cyber-physical vulnerabilities, particularly stealthy False Data Injection Attacks (FDIAs) capable of masking real defects or fabricating false positives without detection.

Existing rule-based and signature-based detection systems fail to identify subtle or novel attacks in high-dimensional, noisy rail geometry data, and most current models require labeled attack datasets that are rarely available. Although unsupervised methods such as autoencoders and Variational Autoencoders (VAEs) can detect deviations from learned normal behavior, they remain limited by non-stationary data characteristics and static detection thresholds.

This research proposes a Hybrid VAE with Median Absolute Deviation (MAD) scoring to enable robust, adaptive anomaly detection based on the statistical significance of reconstruction errors. The study investigates whether this approach enhances detection of both subtle and overt FDIAs compared to Isolation Forest and static-threshold VAE baselines, evaluates the effectiveness of MAD-based adaptive thresholding against fixed percentile methods, and examines trade-offs in interpretability, computational load, and reliability across attack intensities.

Using an operational track geometry dataset (18,290 samples, 87 features) from Colorado rail testing, the methodology simulates FDIAs through additive spikes, multiplicative distortion, and high-variance noise injection on safety-critical features. Model performance is evaluated using precision, recall, F1-score, and accuracy, with PCA and t-SNE visualization for validation. Findings will provide actionable deployment guidelines for enhancing cyber-physical resilience in railway SCADA systems.]]></description>
      <pubDate>Mon, 19 Jan 2026 16:16:59 GMT</pubDate>
      <guid>https://rip.trb.org/View/2655702</guid>
    </item>
    <item>
      <title>Colorado Traffic and Animal Detection (COTAD) Model - Training the YOLO algorithm to Reliably Detect Colorado Wildlife, Pedestrians, and Vehicles</title>
      <link>https://rip.trb.org/View/2643439</link>
      <description><![CDATA[Reliable, automated detection of wildlife in real-time from optical and thermal camera images, and using detection to alert drivers by using dynamic wildlife warning signs (DWW), can save lives and protect property, as well as protect wildlife. This research will develop and train COTAD to detect Colorado large wildlife and test the algorithm near highways. If successful and paired with DWW, this project could become a highly used and cost effective tool to continue Colorado Department of Transportation's (CDOT’s) progress and leadership in reducing wildlife-vehicle collisions.]]></description>
      <pubDate>Tue, 23 Dec 2025 13:50:23 GMT</pubDate>
      <guid>https://rip.trb.org/View/2643439</guid>
    </item>
    <item>
      <title>Evaluation of Vehicle Telematics and Infrastructure-based Connected Vehicle Data for Real-Time Safety and Mobility Application
</title>
      <link>https://rip.trb.org/View/2625309</link>
      <description><![CDATA[The emergence of connected vehicle (CV) data has provided unprecedented opportunities for developing real-time, proactive applications to enhance safety and mobility. This project utilizes and compares telematics and infrastructure-based CV data to determine optimal applications for each and explore integration strategies for safety and mobility solutions. Specifically, telematics CV data provide the location, speed, and other key information on approximately 5-10% of vehicles on the road. In contrast, infrastructure-based CV data from the connected corridor in the City of Madison contain information about traffic signals, vehicles, and road geometry. By comparing and integrating these data sources, this project proposes physics models and neural network algorithms to detect real-time safety issues such as crashes. The detection results can be used to issue immediate warnings to drivers, traffic managers, and automated vehicle systems. To disseminate these warnings, the research team proposes utilizing roadside variable message signs and in-app notifications via platforms like HAAS, Google Maps, and Waze. The proposed applications can be piloted through field tests in the University of Wisconsin-Madison’s Level 3 CAV testbed and possibly later at Mcity.]]></description>
      <pubDate>Thu, 13 Nov 2025 15:31:52 GMT</pubDate>
      <guid>https://rip.trb.org/View/2625309</guid>
    </item>
    <item>
      <title>Evaluating V2X Network Performance and Enhancing Safety and Security in Sensor Data Sharing for Connected and Automated Driving
</title>
      <link>https://rip.trb.org/View/2625308</link>
      <description><![CDATA[This project will investigate the sensor data sharing mechanism with C-V2X and networked vehicle-to-everything (V2X) communication technology in terms of safety, cybersecurity, and network performance with current bandwidth allocations. The research team will (1) develop a comprehensive evaluation framework of the V2X network performance (e.g., latency, throughput) under real – world complexities; (2) develop a data fusion model that fuses Sensor Data Sharing Messages (SDSMs) from multiple sources considering uncertainties in real-world V2X communication networks, errors in sensor-based object detection; and (3) develop a misbehavior detection model that can detect anomaly in SDSMs and evaluate the trustworthiness of the message within a short time.]]></description>
      <pubDate>Thu, 13 Nov 2025 15:28:02 GMT</pubDate>
      <guid>https://rip.trb.org/View/2625308</guid>
    </item>
    <item>
      <title>Eastern States Institutional Issues Study for Commercial Vehicle Operations</title>
      <link>https://rip.trb.org/View/2616148</link>
      <description><![CDATA[The objective is to fund work orders that support: (1) a regional commercial vehicle operations (CVO) forum program; (2) automated data entry and access for safety information; and (3) prototype integration of EZ Pass and CVO Automated Vehicle Identification (AVI) requirements.]]></description>
      <pubDate>Tue, 28 Oct 2025 19:35:54 GMT</pubDate>
      <guid>https://rip.trb.org/View/2616148</guid>
    </item>
    <item>
      <title>Milkweed Presence Detection from High-Resolution Mobile Roadway Photography and
LiDAR</title>
      <link>https://rip.trb.org/View/2601429</link>
      <description><![CDATA[This project builds on previous Idaho Transportation Department (ITD) studies to identify and map milkweed along highway rights-of-way using existing roadway photography and LiDAR data. The effort will produce a geographic information system (GIS)-based inventory that documents route, milepost, area, and density information for each identified milkweed cluster, supporting ITD’s initiatives to prioritize and preserve high-quality Monarch and pollinator habitat. Key objectives include detecting milkweed using LiDAR and photo imagery, mapping occurrences by route and milepost, estimating cluster size and plant density, and analyzing correlations between milkweed area and density.]]></description>
      <pubDate>Wed, 17 Sep 2025 16:30:46 GMT</pubDate>
      <guid>https://rip.trb.org/View/2601429</guid>
    </item>
    <item>
      <title>Automated Wildlife Detection for Wildlife Vehicle Collision Reduction</title>
      <link>https://rip.trb.org/View/2593929</link>
      <description><![CDATA[Large mammals cross highways to access core habitat areas, presenting significant safety hazards for Oregon drivers. Transportation infrastructure can in turn present significant disruption for wildlife connectivity. Automated wildlife detection systems tailored for Oregon conditions could enable Oregon Department of Transportation (ODOT) to efficiently assess the performance of constructed wildlife passage features, validate predicted wildlife crossing locations, monitor changes in the timing and location of crossing patterns as the climate and populations change, provide cost-effective detection systems for the many areas where crossing structures are not feasible, and most importantly reduce wildlife vehicle collisions.]]></description>
      <pubDate>Thu, 28 Aug 2025 11:42:32 GMT</pubDate>
      <guid>https://rip.trb.org/View/2593929</guid>
    </item>
    <item>
      <title>Phase 2: Quantitative Evaluation Process for Improved Rockslope Safety and Reduced Maintenance</title>
      <link>https://rip.trb.org/View/2593928</link>
      <description><![CDATA[Thousands of problematic rock slopes adversely impact Oregon Department of Transportation's (ODOT’s) infrastructure and will soon require substantial mitigation as Oregon’s highway rockcuts near the end of their respective design life. This deterioration is predicted to be further exacerbated with climate change. Many short-term mitigation techniques such as scaling or blasting are very expensive and dangerous as these techniques require personnel to physically scale the cliff, while longer term mitigation solutions (e.g., nailing) are often too cost prohibitive to deploy beyond a few select slopes. Rockfall mitigation efforts are further compounded by the subjective, ad-hoc nature of the mitigation process, which can lead to uncertainty regarding actual slope improvement and predicted improvement duration. It is also uncertain how long—and how effective—mitigation efforts actually are, particularly with respect to the longevity of scaling before the eroding slopes return to a similar or even more precarious state. Further, when rock slopes are not clearly improved with mitigation, clean-up costs for rockfall accumulate for our District maintenance crews. Taken together, the lack of quantitative information regarding how mitigation efforts actually improve the overall stability of rock slopes runs counter to ODOT’s aim for budget efficiency and transparency, with potential negative impacts on an already highly limited rockfall mitigation budget which could then in-turn negatively impact the safety of the traveling public. Quantitative, objective methods that better inform successful rockfall mitigation techniques, maintenance strategies, and overall asset management is needed. Recently, ODOT Research invested in the development of a series of lidar-based tools for quick assessment and monitoring of rockfall events (SPR809). Specifically, SPR809 refined the Rockfall Activity Index (RAI) hazard screening tool, analyzed seismic hazards, and developed a user-friendly Graphical User Interface (GUI) for implementation. RAI is a point cloud-derived, high-resolution, low-cost, morphology-based, objective approach for assessing rock slopes that can objectively and efficiently analyze large geographic regions. RAI may also be used to identify specific locations to target for mitigation by identifying rapidly eroding areas of the slope and precarious overhangs. This proposed research leverages these recently developed tools to further develop a new methodology for verifying rockfall mitigation effectiveness that can also be used to direct future mitigation and maintenance efforts.]]></description>
      <pubDate>Thu, 28 Aug 2025 11:23:39 GMT</pubDate>
      <guid>https://rip.trb.org/View/2593928</guid>
    </item>
    <item>
      <title>Traffic Safety Improvements at Low Water Crossings</title>
      <link>https://rip.trb.org/View/2593185</link>
      <description><![CDATA[The Texas Department of Transportation (TxDOT) Project 0-6992 "Traffic Safety Improvements at Low Water Crossings (LWCs)" proved how easy, low-cost countermeasures improve safety at LWCs by focusing on LWC delineation, using flood-detection sensors, and warning systems to alert travelers of flooded crossings. In addition to recommending the use of raised retroreflective pavement markers (RRPMs) to improve longitudinal markings at low-water crossings, the research team recommended experimenting with internally illuminated raised pavement markers (IRPMs) at problematic locations where drivers regularly drive through high water conditions. The research team will improve safety and operations at LWCs by incorporating a tool which integrates the National Oceanic and Atmospheric Administration (NOAA's) Multi-Radar/Multi-Sensor (MRMS) system, which combines radar, stream gauges, and environmental data to estimate rainfall rates with 1-kilometer resolution across the United States. By leveraging virtual sensor data, TxDOT will enhance the effectiveness of roadway flood warning systems, making early flood detection more robust and improving overall driver safety.]]></description>
      <pubDate>Tue, 26 Aug 2025 12:29:36 GMT</pubDate>
      <guid>https://rip.trb.org/View/2593185</guid>
    </item>
    <item>
      <title>Large-scale Testing for Detecting Changes in Track Modulus with Low-cost Sensors Installed on Rolling Stock



</title>
      <link>https://rip.trb.org/View/2572334</link>
      <description><![CDATA[U.S. railroads transport 1.6 billion tons of freight over more than 140,000 miles of track each year. Safe and efficient operation of such a vast infrastructure requires extensive monitoring, evaluation, and maintenance of its track systems. Track modulus is a critical parameter in the design, analysis, and maintenance of railroad track, as it is indication of  the material stiffness below the rail comprising the combined stiffnesses per unit length of rail, plates, ties, ballast, and subgrade. Locations of “soft” track can cause increases in rail deflections and stresses, which increases the rate of rail deterioration. Traditionally, track modulus is measured onsite using static deflection testing where a known load is applied to the track, the resulting deflection is measured, and this deflection is extrapolated to the track modulus. However, this is time consuming and labor-intensive, especially if measurements are to be taken at multiple points along the track. To address the challenges with track modulus measurement, this research attempts to refine existing methods and expand the scale of monitoring by leveraging data from railcars to identify problem locations on the track and relate measured responses to specific track deficiencies. The approach allows making continuous estimations over long sections of the track and also is less costly, as self-contained acceleration and data acquisition systems are inexpensive and easy to attach to the vehicle bodies. This research will use technologies such as the ground penetrating radar (GPR) and track geometry cars in combination with low-cost sensors placed on the existing plant of rolling stock for accelerated track monitoring. Track response to track conditions will be measured. Constitutive load-deflection relationships will be established between track condition, loading, and deflection to determine the track modulus. The modulus will be characterized by leveraging the mechanics that relates vehicle accelerations to the condition of track upon which the instrumented vehicle travels. The research team is experienced in the use of low-cost accelerometers for bridge monitoring and assessment, where it established the mechanistic relationships between loading and response under various conditions and showed the feasibility of determining quantifiable, specific conditions from the acceleration data. Using this expertise and experience, the team now seeks to develop mechanics-based relationships that correlate railcar body acceleration profiles to track behavior and, eventually, track condition. This project is the next step towards full implementation, where it will use previously characterized conditions on a model test to support the constitutive load-deflection relationships. The overall impact of this work will be the widespread monitoring of track infrastructure through the installation of low-cost accelerometers on existing rollingstock. The industrial partner, BNSF, will provide support and help with implementation.]]></description>
      <pubDate>Wed, 09 Jul 2025 16:08:22 GMT</pubDate>
      <guid>https://rip.trb.org/View/2572334</guid>
    </item>
    <item>
      <title>Development of a Prototype Turnkey Artificial Intelligence Aided Automated Trespassing Detection Solution Based on Stationary Cameras



</title>
      <link>https://rip.trb.org/View/2572329</link>
      <description><![CDATA[This Type II IDEA project will develop and test a prototype turnkey artificial intelligence aided trespassing detection system.  The system consists of integrated hardware (solar security trailer, networking equipment, etc.) and software that was proven in an earlier project funded by the Federal Transit Administration and Federal Railroad Administration. This system will be developed and tested in collaboration with the industry partner, SunRail, a commuter rail system in the greater Orlando, Florida area. The system hardware will be assembled and installed at selected locations. Data will be collected in those locations for 12 months, and the information will be analyzed to provide actionable safety data to SunRail, the industry collaborator. SunRail will install fencing along their right-of-way. This system could be used to gather trespassing data before and after the fencing installation to evaluate the effectiveness of the solution. At grade crossings, violation data could be used to justify upgrades like the installation of quad gates, gate skirts, or dynamic envelopes based on the types of violations observed. This data can improve trespassing mitigation decision making and support grant applications for further actions. Following this task, sample video data will be collected and analyzed to ensure system accuracy and data quality. The developed system will benefit railroad industry by enabling the collection of previously unavailable trespassing and grade crossing violation information.  It is rather unfeasible to have railroad staff manually annotate video feeds to acquire trespassing data.  This system, on the other hand,  will automatically watch and understand trespass behavior from video feeds at remote locations. Trespass and grade crossing violation information will be aggregated in a trespasser database, presenting users with a video clip of the trespassing event and corresponding metadata (time, weather, type: person, car, motorcycle etc.). Trends and common behaviors can be determined once enough of these events are aggregated.]]></description>
      <pubDate>Tue, 08 Jul 2025 16:55:33 GMT</pubDate>
      <guid>https://rip.trb.org/View/2572329</guid>
    </item>
    <item>
      <title>Multimodal 3D Perception System for Active Safety at Accident-Prone Locations</title>
      <link>https://rip.trb.org/View/2562265</link>
      <description><![CDATA[Accident-prone intersections in Michigan continue to account for a disproportionate share of severe crashes, highlighting the need for infrastructure-based active safety capabilities that can detect, anticipate, and mitigate imminent conflicts in real time. This project develops and demonstrates a full-stack roadside sensing and warning system that integrates LiDAR and wide-area cameras with edge computing, cloud-based data management, and V2X communications. By fusing complementary sensor modalities using Bird’s Eye View (BEV) fusion and conflict prediction, the system aims to improve detection accuracy, 3D localization, robustness under varied conditions, and early identification of potential collisions. The resulting system will generate timely, targeted safety warnings via roadside units (RSUs) and will be validated through staged data collection, model training, and controlled field testing at Mcity, establishing a scalable technical foundation for future deployment at high-risk intersections and similar safety-critical locations.

]]></description>
      <pubDate>Fri, 06 Jun 2025 14:48:04 GMT</pubDate>
      <guid>https://rip.trb.org/View/2562265</guid>
    </item>
  </channel>
</rss>