<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>GenAI-Enabled Automated Traffic Simulation Management</title>
      <link>https://rip.trb.org/View/2742141</link>
      <description><![CDATA[Microscopic traffic simulation software allows users to model traffic flow, assess traffic management strategies, and optimize transportation systems for efficiency. However, building a simulation model for a real-world road network is a complex, manual, time-consuming, and error-prone task. GenAI-Enabled Automated Traffic Simulation Management uses Generative Artificial Intelligence (AI) (GenAI) to automate the preparation of input data, the running of simulations, and the extraction of output results, enabling traffic engineers to focus on the purpose of the simulation rather than the tedious manual work of building the model.

The project maps user text-based scenario descriptions to actual simulation scenarios, with step-by-step validation from the user before execution, by wrapping the INTEGRATION simulation software with a callable API via the Model Context Protocol (MCP) and a web-based interface to a large language model (LLM).

The project will deliver a baseline simulation model built with the INTEGRATION microscopic simulation software for a selected freeway corridor or road network; an INTEGRATION API that wraps the simulation software using the Model Context Protocol so it can be called by any large language model; and a web-based graphical user interface that guides users in prompting an LLM to build simulation models and answer questions using Retrieval-Augmented Generation from the INTEGRATION manual. The GenAI system will be validated by comparing GenAI-generated model files against the baseline model, through human-in-the-loop validation, and through real-world deployment with the City of Alexandria.]]></description>
      <pubDate>Sat, 01 Aug 2026 09:48:00 GMT</pubDate>
      <guid>https://rip.trb.org/View/2742141</guid>
    </item>
    <item>
      <title>Assessing Benefits of Mixed Traffic Platooning with Multi-Agent Reinforcement Learning and Cooperative Adaptive Cruise Control with Unconnected Vehicles </title>
      <link>https://rip.trb.org/View/2739300</link>
      <description><![CDATA[This project assesses the performance benefits of implementing multi-agent reinforcement learning (MARL)-based cooperative platooning and cooperative adaptive cruise control with unconnected vehicles (CACCu) in mixed traffic environments, comparing scenarios with and without these technologies under various connected automated vehicle (CAV) market penetrations. The main goal is to investigate when a policy to deploy these advanced technologies makes sense.

Cooperative platooning in mixed traffic, where CAVs must interact safely and efficiently with human-driven vehicles, remains a key barrier to realizing the full mobility, safety, and energy benefits of connected automation, a challenge amplified by uncertainty in human driving behavior. When a CAV’s immediate preceding vehicle is not connected, it may benefit from a lane change to follow a connected vehicle and form cooperative adaptive cruise control; the team’s MARL approach, built on a CNN QMIX architecture supporting centralized training with decentralized execution, learns coordination policies that adapt to surrounding vehicles rather than relying on fixed rules.

]]></description>
      <pubDate>Thu, 30 Jul 2026 16:05:40 GMT</pubDate>
      <guid>https://rip.trb.org/View/2739300</guid>
    </item>
    <item>
      <title>Potential Impact of Autonomous Vehicles on Reducing Congestion - Phase 2</title>
      <link>https://rip.trb.org/View/2733194</link>
      <description><![CDATA[Traffic congestion is a major problem in large metropolitan areas in the United States. In 2022, on average, a commuter lost about $1,259 in monetary terms annually due to congestion nationwide, which amounts to 8.7 billion lost hours in total.  The lack of coordination among individual users, who make routing decisions independently based on current traffic information without anticipating that others may follow similar decision-making patterns, contributes significantly to the high cost of congestion.    

The behavior of drivers optimizing their individual routes leads to a state known as the User Equilibrium, leading to travel times that can be significantly higher than travel times from the System Optimal, particularly in congested urban networks where the effects of individual decisions cascade throughout the system.  With the future emergence of autonomous vehicles, it is possible that organizations may now own more of the fleet of vehicles and control their routing, providing the organization more options for balancing route selections and thus making it possible to find routing solutions closer to the system optimal.  Driverless ride-hailing companies such as Waymo have already begun their service in five major cities across the United States and Tesla has started to test their Robotaxi service in Austin, Texas.

In Phase 1, the research team developed the research foundation for this problem. This work includes the literature review and the development of an online dispatch-and-relocation framework for a centrally controlled autonomous vehicle fleet. The Phase 1 framework matches requests to vehicles while accounting for pickup deadlines, near-term vehicle availability, and proactive repositioning toward forecasted demand. Phase 1 also establishes a comparison structure against a traditional human-driver ride-hailing system and an initial simulation capability that traces routes and estimates vehicle miles traveled, deadhead miles, passenger waiting time, revenue, and related performance measures.

Phase 2 will build directly on this foundation and is the primary focus of the next stage of the project. In Phase 2, the team will scale the optimization and simulation framework so it can solve problems at the size of major metropolitan areas. This includes extending the model to larger networks and richer demand patterns, improving computational tractability for larger instances, and strengthening the simulation module so it can evaluate passenger-vehicle matches and route decisions under more realistic operating conditions. To make the model scalable, the team will aggregate the service region into zones and solve the resulting problems repeatedly over short rolling horizons. The team will also need to calibrate the demand forecasting and routing inputs for large urban networks and test the algorithms on progressively larger instances to ensure that the solution quality and computation time remain practical. The purpose of Phase 2 is to determine how much centralized control of autonomous fleets can reduce system-wide travel, deadhead mileage, waiting times, and congestion when evaluated on realistic metropolitan-scale settings.

]]></description>
      <pubDate>Wed, 22 Jul 2026 17:37:19 GMT</pubDate>
      <guid>https://rip.trb.org/View/2733194</guid>
    </item>
    <item>
      <title>Routing Autonomous Trucks on Dedicate Lanes- Phase 2</title>
      <link>https://rip.trb.org/View/2733038</link>
      <description><![CDATA[Trucks are known to have a significant impact on congestion during traffic peak hours due to their size and slower dynamics. Human operated trucks for freight transport are faced with two constraints: those imposed by the service demand and those imposed by the human driver. For long haul operations, for example, truck drivers must meet the constraints of hours of service. For short haul they must meet family and personal constraints which often do not allow them to operate during odd hours. With automation the human constraints are removed which opens the way to view truck routing and scheduling under different and more flexible constraints. The major problem faced by automated trucks operating with the rest of traffic, however, is safety as due to the different sizes involved the sensing problem is more challenging and potential accidents can be catastrophic. Moving trucks from times of high congestion to times of no congestion will bring considerable benefits to trucking companies as well as to all other users of the road network, as fewer trucks will be operating during peak traffic hours. In addition, trucking companies that are short of truck drivers will be able to operate without disruptions and without human imposed constraints, saving on labor costs.

During the first phase of the project, the research team developed microscopic traffic simulation model which the team validated using real data from I-710. The network considered was part of I-710 and the team assumed as a first step single origin-destination (OD) flows. The team considered the scenario where trucks sharing the same road network as passenger cars become automated and operate on dedicated truck lanes at times that the traffic demand is very low, so that lanes can be switched dynamically to dedicated automated truck lanes without affecting traffic. By doing so we can keep the automated trucks separated from manually driven vehicles, thereby addressing the issue of safety.

The ongoing phase 1 study shows that by removing a number of trucks which are about 0.4% of all vehicles during a high peak traffic and have them automated and operating on dynamically dedicated lanes during off peak traffic the travel time for trucks is reduced by 4.5% while the travel time of passenger vehicles during the high peak traffic decreases by about 3%. These preliminary findings suggest that temporal rescheduling of freight demand, combined with dynamic lane management, could improve both freight and overall network performance. In phase 1 the team simply used the traffic simulator to test their ad hoc approach of moving trucks from high peak to low peak traffic without any form of optimization.

In phase 2 the team plans to extend the approach as follows: (1) The team will expand the road network to include some of the most popular truck routes covering short medium and long-haul scenarios. The issue of parking and refueling in the absence of driver will also be addressed. (2) The team will extend the results of phase 1 to multiple interacting OD pairs, allowing the framework to capture more realistic freight demand patterns and network-level coordination effects. (3) The team considers the case of truck platoons which will include fully automated truck platoons but also the more realistic case where the first truck in the platoon has a human driver. In other words, the lead truck will be driven by a human driving and following trucks will be electronically connected and fully automated. Truck platooning is an attractive concept as it has shown to have the potential of reducing aerodynamic drag and contribute to significant fuel savings. (4) The team plans to optimize their decisions of temporal rescheduling of freight demand, combined with dynamic lane management to achieve the best possible outcome. The team views the problem as assigning loads in 2 dimensions temporal and spatial in a way that reduces travel time and lowers fuel cost for both trucks and passenger vehicles.]]></description>
      <pubDate>Wed, 22 Jul 2026 17:28:54 GMT</pubDate>
      <guid>https://rip.trb.org/View/2733038</guid>
    </item>
    <item>
      <title>Community Response Enhanced Education on Disasters (CREED): Virtual Reality Training to Enhance Transportation</title>
      <link>https://rip.trb.org/View/2732358</link>
      <description><![CDATA[Artificial Intelligence/Virtual Reality (AI/VR) simulation activities enhance critical emergency management functions—planning, forecasting, threat detection, security, and information sharing—ultimately improving the preservation of life and property. These tools create a safe, immersive environment where participants build problem-solving skills, decision-making ability, and confidence in disaster response. This project will deliver education, professional development, and continuous improvement in emergency preparedness for both aspiring and current professionals. The research team will invite high school and community college students to participate alongside university students and practitioners. Participants will engage in hands-on training and live demonstrations using advanced technologies such as virtual reality (VR) and artificial intelligence (AI)-driven simulations. The program fosters multidisciplinary collaboration among disciplines: Emergency Management Technology, Meteorology, Computer Science/Engineering, Health Science, and Journalism, and Media Studies. Through classroom instruction, workshops, and interactive training demos, students will work directly with emergency management professionals, strengthening real-world skills, supporting school-to-work transitions, and enhancing career readiness.]]></description>
      <pubDate>Tue, 21 Jul 2026 16:29:54 GMT</pubDate>
      <guid>https://rip.trb.org/View/2732358</guid>
    </item>
    <item>
      <title>Developing Guidelines for Right-Turn Lane Pockets for Intersections in Nevada</title>
      <link>https://rip.trb.org/View/2713605</link>
      <description><![CDATA[Nevada is overrepresented in intersection-related crashes and has been designated as an Intersection-Focused state by the Federal Highway Administration (FHWA).  Further research exploring how dedicated right-turn lane pockets could mitigate fatal and serious injury crashes is needed.  Currently, the criteria used to determine when an exclusive right-turn lane is required may be inconsistently applied given today’s traffic volumes, vehicle mix, pedestrian activity, and safety expectations. In some cases, right-turn lanes are omitted even when they could improve operations or safety, while in others, they are added despite limited benefits and significant cost or right-of-way impacts.
The main objective of this research is to produce an updated, data-driven, and context-sensitive framework for determining when and where exclusive right-turn lane pockets should be required or clearly not required at access points in Nevada.  In particular, a tool will be provided that will improve the Nevada Department of Transportation’s (NDOT’s) ability to consistently evaluate access points to the state highway system and will maximize the safety and operational benefits of exclusive right-turn lanes. Additionally, the research results will provide NDOT with a process model that supports consistent, transparent, and technically-sound decision making for new developments and access modifications.
The University of Nevada, Reno team plans to reach the research objective by: (1) Synthesizing the current state of knowledge and practice for right-turn lane warrants nationally and among peer agencies. (2) Collecting high-resolution field data and developing calibrated VISSIM microsimulation models.  (3) Conducting systematic scenario-based simulations using the calibrated VISSIM models and crash data analyses.  (4) Developing guidelines, using the empirical findings from the aforementioned tasks, for determining when an exclusive right-turn lane pocket is warranted. (5) Developing a user-friendly, implementation-ready decision-support tool for use by NDOT reviewers, local agency staff, developers, and consulting engineers. (6) Compiling all research findings, guidelines, and tools into a comprehensive final report and conducting a training workshop to facilitate implementation.
The final deliverables will be designed from the outset for direct integration into NDOT’s policies, procedures, and operations. 
Additionally, the implementation plan will be structured as a phased approach that transitions from research completion through pilot application, full deployment, and sustained use.]]></description>
      <pubDate>Thu, 11 Jun 2026 14:33:30 GMT</pubDate>
      <guid>https://rip.trb.org/View/2713605</guid>
    </item>
    <item>
      <title>Development and Evaluation of Approach Guardrail Transition with Increased Span Length between Concrete bridge Rail and First Transition Post - Phase II </title>
      <link>https://rip.trb.org/View/2689395</link>
      <description><![CDATA[Phase I of this project, funded by the Nebraska Department of Transportation (NDOT), addressed this need at the concept and simulation level. Midwest Roadside Safety Facility (MwRSF) researchers developed and refined several long-span approach guardrail transition (AGT) concepts for the 34-inch tall NDOT thrie-beam system and used LS-DYNA simulations to evaluate their performance with increased span between the concrete buttress and the first transition post under MASH TL-3 impact conditions. The work included evaluation of the upstream W-beam to thrie-beam transition, the downstream thrie-beam to rigid buttress connection, and identification of critical impact points for both the pickup truck and small car tests. These analyses demonstrated that the selected long-span concept is a promising candidate, but they do not satisfy Federal Highway Administration (FHWA) requirements. Federal acceptance of new roadside safety hardware under the American Association of State Highway and Transportation Officials (AASHTO) Manual for Assessing Safety Hardware (MASH) requires full-scale crash testing. An FHWA eligibility letter cannot be obtained on the basis of simulations alone. Without full-scale crash testing, the long-span AGT system cannot be fully validated, adopted statewide, or included in NDOT standard plans. Phase II is therefore needed to conduct the required full-scale MASH TL-3 crash tests and provide an FHWA-compliant evaluation of the new long-span AGT system.]]></description>
      <pubDate>Tue, 02 Jun 2026 12:25:41 GMT</pubDate>
      <guid>https://rip.trb.org/View/2689395</guid>
    </item>
    <item>
      <title>Development of design guidelines for protection against erosion at bridge piers of rectangular cross section and estimating effects of pressurized flow on erosion potential</title>
      <link>https://rip.trb.org/View/2706034</link>
      <description><![CDATA[Bridge piers are vulnerable to severe erosion (scour) during high-flow and flooding conditions, which can compromise structural stability and, in extreme cases, lead to bridge failure. Existing riprap design methodologies used to protect bridge piers have limitations, particularly for rectangular piers and for conditions in which bridge decks become submerged and flow transitions from open channel to pressurized regimes. Inadequate riprap sizing under such conditions increases risk of structural distress, traffic interruption, and potential safety hazards.
This project develops improved design guidelines for riprap protection at rectangular bridge piers under both open channel and pressurized flow conditions. Using validated three-dimensional numerical simulations, the research will quantify how pier geometry, aspect ratio, angle of attack, and flow regime influence critical shear stress and the Froude number associated with stone failure. The project will propose a multi-parameter riprap sizing formula applicable to a broader range of geometrical and hydraulic conditions, including overtopping scenarios. Recommendations will be provided for adapting existing HEC-18 methodologies to account for pressurized flow conditions at bridge sites.

]]></description>
      <pubDate>Sat, 23 May 2026 17:39:45 GMT</pubDate>
      <guid>https://rip.trb.org/View/2706034</guid>
    </item>
    <item>
      <title>Enhancing Safety in Mixed-Autonomy Traffic via Prediction-Based Connected Autonomous Vehicle Control</title>
      <link>https://rip.trb.org/View/2703928</link>
      <description><![CDATA[This project proposes a new framework for prediction-based connected autonomous vehicle (CAV) control to enhance safety in mixed-autonomy traffic where CAVs and human-driven vehicles (HVs) coexist. Specifically, by predicting future traffic conditions behind a target CAV, the vehicle can be proactively controlled to improve the safety and efficiency of the overall traffic stream. This approach is motivated by the fact that a controlled CAV directly influences the behavior, safety, and performance of following HVs through car-following interactions. Accordingly, the proposed method jointly considers a CAV and its following HVs in the design of a safety-aware driving strategy. Although HVs do not communicate with CAVs, traffic states related to HVs can be estimated using partial traffic measurements collected by CAVs. Leveraging these predictions, the proposed control strategy will be formulated within a model predictive control (MPC) framework to improve safety and traffic efficiency for HVs following a CAV. Extensive simulation studies will be conducted under a range of traffic scenarios and HV driving styles to demonstrate the effectiveness of the proposed approach. In addition, multiple CAV penetration rates will be evaluated to examine scalability and deployment potential. 
This project is highly aligned with the Mid-America Transportation Center's (MATC’s)  mission to advance transportation safety through technology development, technology transfer, and deployment. It addresses a timely safety challenge: near-term traffic will be mixed-autonomy, where early-generation autonomous vehicles (e.g., ACC-equipped vehicles and emerging CAVs) operate alongside the majority of HVs. In this environment, safety risks arise not only from individual vehicle performance, but also from interactions between automated and human drivers—an issue that is often overlooked in existing CAV control design.
While this project will leverage an existing dataset collected in Minnesota and high-fidelity simulation data generated in Simulation of Urban MObility (SUMO) for numerical investigation and validation, the team anticipates extending the proposed methodology in future work using connected-vehicle and SPaT data to be collected by Dr. Li Zhao’s team at UNL, in collaboration with the Nebraska DOT. 
]]></description>
      <pubDate>Thu, 21 May 2026 22:41:02 GMT</pubDate>
      <guid>https://rip.trb.org/View/2703928</guid>
    </item>
    <item>
      <title>Deep Learning–Based Digital Image Correlation for Fatigue Crack  Characterization in Steel Structures
</title>
      <link>https://rip.trb.org/View/2703927</link>
      <description><![CDATA[This proposal presents a strategic approach to improving transportation safety through the advancement of deep learning–based Digital Image Correlation (DIC) for fatigue crack characterization in steel structural components. With aging transportation infrastructure and increasing cumulative traffic loading, fatigue-related deterioration in steel bridges and related systems presents ongoing safety risks. Accurate measurement of crack-induced displacement fields is critical for reliable structural assessment and informed maintenance decisions. The primary objectives of this proposal are to advance artificial intelligence (AI)-driven DIC methods beyond the limitations of conventional correlation-based approaches by enabling sub-pixel displacement learning through synthetic data generation, incorporating physics-informed modeling of crack-induced displacement discontinuities, and supporting high-resolution analysis of large image regions without loss of spatial detail. The methodology involves grayscale synthetic speckle data generation for sub-pixel displacement learning, mechanics-based displacement field modeling using finite element simulations, and development of an attention-enhanced deep learning architecture for full-field displacement prediction. Experimental validation against commercial DIC systems will establish a transferable methodology supporting safer fatigue crack evaluation practices.
]]></description>
      <pubDate>Tue, 19 May 2026 13:48:19 GMT</pubDate>
      <guid>https://rip.trb.org/View/2703927</guid>
    </item>
    <item>
      <title>Risk Evaluation of Transportation Systems to Tornadoes to Facilitate SAFE Digital Twin Development</title>
      <link>https://rip.trb.org/View/2703879</link>
      <description><![CDATA[This project will develop tornado hazard curves for tornado-prone regions. To achieve this, the tornado genesis model will be applied to generate a large number of tornado starting points. Then, the tornado track model will be applied to simulate track parameters (e.g., path width, length, heading direction and intensity). Next, for each tornado track, a wind field model will be established to obtain the information on the entire wind field, including wind velocity and pressure. For this, CFD simulations will be first applied to model a small number of tornadoes with representative flow structure and intensities; a surrogate model will then be developed using the multi-fidelity machine learning modeling technique, to replace the time-consuming CFD simulations. Finally, the generated data from the synthetic tornado tracks for a great number of years will be processed statistically to develop tornado hazard curves and tornado hazard maps for the states in Mainland America. The developed tornado hazard curves will help Department of Transportation properly assess the damage to vehicles on the road or in parking lots, informing stakeholders of preparation for future tornadoes. The developed curves can be integrated into catastrophe modeling to better estimate the risk of vehicles under tornadoes and thus better price the automobile insurance premium. In addition, these curves can improve the building codes related to a tornado-resistant design. ]]></description>
      <pubDate>Mon, 18 May 2026 17:13:44 GMT</pubDate>
      <guid>https://rip.trb.org/View/2703879</guid>
    </item>
    <item>
      <title>Agentic LLM-powered Synthetic Transportation Agent Response System (AL-STARS)</title>
      <link>https://rip.trb.org/View/2703689</link>
      <description><![CDATA[Household travel surveys help to gather data on travel trends and are a key part of transportation planning. Gathering this data has become challenging, however, as response rates have been decreasing. This project will develop a web-based travel survey tool (AL-STARS) that generates realistic, synthetic travel data. AL-STARS will use an advanced AI model to simulate travelers in Illinois, allowing planners and modelers to test transportation ideas and survey questions virtually before real-world use. The project will help IDOT make better decisions in infrastructure and policy by providing a more accurate, diverse and cost-effective way to understand how different communities — especially undersampled population groups — actually travel.]]></description>
      <pubDate>Fri, 15 May 2026 09:28:30 GMT</pubDate>
      <guid>https://rip.trb.org/View/2703689</guid>
    </item>
    <item>
      <title>TraCR Foundational Project: TraCR Collective Transportation Cybersecurity Testbeds</title>
      <link>https://rip.trb.org/View/2697454</link>
      <description><![CDATA[The National Center for Transportation Cybersecurity and Resiliency's (TraCR's)
foundational project aims to develop technological tools, prototypes, testing platforms, and facilities to ensure the cybersecurity and cyber-resilience of multimodal transportation systems and related infrastructure. The project is led by Clemson University (Clemson) under the strategic direction of Dr. Ronnie Chowdhury (Lead PI), with coordination support from Dr. Sabbir Salek (Co-PI), and involves all eight other TraCR partner institutions organized into four subgroups. A structured project governance framework, including biweekly subgroup meetings, monthly full-team coordination meetings, quarterly progress reporting and advisory board engagement, ensures alignment with project milestones, integration across teams, and effective monitoring of technical progress and deliverables. 

Clemson collaborates with Benedict College (Benedict), South Carolina State University (SCSU), and the University of Texas at Dallas (UTD) to advance a comprehensive, automated threat modeling capability for multimodal transportation systems. Building on the Transportation Cybersecurity and Resiliency Threat Modeling Framework (TraCR-TMF), the team conducts testbed-in-the-loop evaluations within Clemson’s real-world cybersecurity testbed, implementing digital-twin-based cybersecurity analysis of in-vehicle networks, and engaging state transportation agencies to assess operational transferability. Additionally, the team will work to integrate graph-based reasoning models into threat modeling, deploy supervised ModernBERT classifiers, and align with the MITRE Embedded Systems Threat Matrix to strengthen structured system-to-vulnerability mapping and improve threat coverage across transportation cyber-physical systems.

The other partner institutions will develop additional real-world and virtual testing platforms to support cybersecurity experimentation for multimodal transportation. Florida International University (FIU) and the University of Alabama at Tuscaloosa (UA) are jointly advancing the Open-Source Connected and Automated Mobility Co-Simulation (OpenCAMS) environment and related simulation platforms, integrating SUMO, CARLA, and network simulation tools, to evaluate privacy-aware multimodal large language models and post-quantum-secure C-V2X communications. Their efforts further include the development and validation of spoofing attack models targeting Basic Safety Message transmissions and multi-frequency GPS receivers, as well as investigations into backdoor-resilient perception systems and the security of vision-language models for intelligent transportation applications.

Purdue University (Purdue) and the University of California, Santa Cruz (UCSC) are advancing adversarial testing methodologies through integrated physical-virtual experimentation frameworks that combine miniature autonomous vehicle testbeds, CARLA/METS-R simulation coupling, and scenario-based vulnerability discovery. These activities include simulation-to-real validation of perception and traffic signal spoofing attacks, evaluation of V2X safety message vulnerabilities, cybersecurity analysis of shared micromobility Bluetooth pairing protocols, implementation of lightweight post-quantum cryptographic protections for vulnerable road user beacons, and closed-loop security assessments of traffic signal controller infrastructures, along with investigations of secure multimodal AI agents and memory-augmented reasoning architectures for autonomous robotic transportation systems.

In addition, Morgan State University (MSU) is enhancing its connected vehicle cybersecurity experimentation capabilities by developing replay-attack models targeting C-V2X onboard units and evaluating mitigation strategies in its real-world testbed environment, in collaboration with Clemson. These efforts quantify communication-level impacts on safety-critical applications and support the development of deployable countermeasures to strengthen resilience against wireless attack vectors affecting connected transportation infrastructure.
]]></description>
      <pubDate>Thu, 30 Apr 2026 12:19:22 GMT</pubDate>
      <guid>https://rip.trb.org/View/2697454</guid>
    </item>
    <item>
      <title>SentinelLab: A Plug-in Online Defender Testbed for Connected and Autonomous Vehicle (CAV) </title>
      <link>https://rip.trb.org/View/2696945</link>
      <description><![CDATA[This project develops SentinelLab, a closed-loop defender testbed designed to transform how Connected and Automated Vehicle (CAV) cybersecurity is validated. Currently, most research stops at detecting anomalies; this project bridges the gap to active defense by integrating the METS-R traffic simulator with the CARLA photo-realistic sensor simulator. The testbed utilizes a “Recognize, Decide, Act” framework. The project employs a multimodal Large Language Model (LLM) to recognize specific attack families (e.g., message replay, breaking provocation) from noisy streaming signals. It then uses an online Defender Workbench to decide on mitigation strategies via plug-in policies and automatically executes these actions in the simulation. This system enables researchers and public agencies to prepare defenses against realistic threats and quantify their impacts on safety and mobility.]]></description>
      <pubDate>Wed, 29 Apr 2026 11:32:19 GMT</pubDate>
      <guid>https://rip.trb.org/View/2696945</guid>
    </item>
    <item>
      <title>Compositional Modeling and Attack Analysis of End-to-end Autonomous Vehicles</title>
      <link>https://rip.trb.org/View/2696904</link>
      <description><![CDATA[The architecture of autonomous driving systems is shifting from modular pipelines to end-to-end systems powered by advanced Artificial Intelligence (AI) and Vision-Language Models (VLMs). This project develops a comprehensive framework for the compositional modeling, security analysis, and resilience testing of these next-generation systems. The research team will create formal models to abstract the behavior of AI components and build an AI-powered vulnerability analysis engine to identify semantic attacks. The project culminates in a high-fidelity, open-source software testbed that integrates these models to simulate attacks and evaluate the resilience of autonomous vehicles and drones.]]></description>
      <pubDate>Tue, 28 Apr 2026 15:45:25 GMT</pubDate>
      <guid>https://rip.trb.org/View/2696904</guid>
    </item>
  </channel>
</rss>