<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/common/TRIS Suite/feeds/RSS.aspx?s=PHNlYXJjaD48cGFyYW1zPjxwYXJhbSBuYW1lPSJzdWJqZWN0aWQiIHZhbHVlPSIxODA0IiAvPjxwYXJhbSBuYW1lPSJkYXRlaW4iIHZhbHVlPSI3MzAiIC8%2BPHBhcmFtIG5hbWU9InN1YmplY3Rsb2dpYyIgdmFsdWU9Im9yIiAvPjxwYXJhbSBuYW1lPSJ0ZXJtc2xvZ2ljIiB2YWx1ZT0ib3IiIC8%2BPHBhcmFtIG5hbWU9ImxvY2F0aW9uIiB2YWx1ZT0iMTYiIC8%2BPC9wYXJhbXM%2BPGZpbHRlcnMgLz48cmFuZ2VzIC8%2BPHNvcnRzPjxzb3J0IGZpZWxkPSJwdWJsaXNoZWQiIG9yZGVyPSJkZXNjIiAvPjwvc29ydHM%2BPHBlcnNpc3RzPjxwZXJzaXN0IG5hbWU9InJhbmdldHlwZSIgdmFsdWU9InB1Ymxpc2hlZGRhdGUiIC8%2BPC9wZXJzaXN0cz48L3NlYXJjaD4%3D" 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>Drone Network Design for Comprehensive Emergency Response in Rural Utah</title>
      <link>https://rip.trb.org/View/2786771</link>
      <description><![CDATA[Develop an optimized drone network design for comprehensive emergency response in rural Utah to deliver much needed supplies during times of emergency.]]></description>
      <pubDate>Tue, 06 Oct 2026 17:13:13 GMT</pubDate>
      <guid>https://rip.trb.org/View/2786771</guid>
    </item>
    <item>
      <title>Roadmap for Virtual Traffic Light Control at Intersections Using Connectivity, AI and Digital Twins</title>
      <link>https://rip.trb.org/View/2786712</link>
      <description><![CDATA[Vehicle automation, connectivity, artificial intelligence (AI), and advanced sensing enable new approaches to intersection management. This project investigates a virtual traffic light (VTL) framework that over time can replace conventional signalized intersections, which carry high capital, maintenance, and operating costs. In the proposed system, connected and automated vehicles (CAVs), along with connected pedestrians and bicyclists, communicate continuously with a virtual control agent that maintains real-time estimates of their positions, speeds, and intended trajectories. Using these data, the agent computes individualized right-of-way assignments—equivalent to virtual green, yellow, and red indications—and transmits them directly via vehicle-to-everything (V2X) communication to each user. By coordinating movements through communication rather than fixed infrastructure, the system can improve traffic efficiency, reduce delays, enhance safety, and lower costs. An additional advantage of VTL systems is their flexibility: control logic and policies can be updated through software, enabling rapid, low-cost adaptation compared with conventional traffic signals. Because full connectivity will be achieved gradually, the research first establishes upper-bound performance under ideal, fully connected conditions without failures, providing quantitative benchmarks. It then develops a roadmap for transitioning from today’s traffic control to increasingly connected environments, accommodating mixed traffic with varying levels of connectivity. The results will guide phased deployment strategies aligned with USDOT priorities in safety, mobility, sustainability, and cost-effective infrastructure modernization.

]]></description>
      <pubDate>Mon, 05 Oct 2026 13:32:38 GMT</pubDate>
      <guid>https://rip.trb.org/View/2786712</guid>
    </item>
    <item>
      <title>Measuring Household Mobility and Travel Patterns with Partial Vehicle Automation</title>
      <link>https://rip.trb.org/View/2786708</link>
      <description><![CDATA[Vehicle automation has the potential to fundamentally reshape travel behavior by reducing the cognitive and physical burden of driving. A large body of research has explored the potential impacts of fully autonomous vehicles on travel demand. However, fully autonomous systems have not yet achieved widespread market adoption, limiting researchers’ ability to study these behavioral changes using real-world data and limiting the current potential of this technology to benefit society. Meanwhile, partial vehicle automation is being adopted and used by consumers. This project uses partial vehicle automation systems (SAE Level 2) as a real-world setting for studying how automation technologies may influence household mobility, travel demand, and transportation system performance. Unlike fully autonomous vehicles, partial automation systems are already integrated into millions of vehicles across a wide range of vehicle types and price points, allowing the study of automation-related behavioral changes under real market conditions. The project combines a nationwide cross-sectional survey merged with Vehicle Identification Number-decoded vehicle specifications and a longitudinal panel of California households. Using econometric analysis, the study will identify the types of travel associated with partial automation adoption and examine which households are most likely to benefit from these technologies. Findings will support transportation agencies and planners seeking to better understand how emerging vehicle technologies may shape regional travel patterns, infrastructure demand, and economic activity.]]></description>
      <pubDate>Mon, 05 Oct 2026 13:28:16 GMT</pubDate>
      <guid>https://rip.trb.org/View/2786708</guid>
    </item>
    <item>
      <title>Advancing Autonomous Vehicle Deployment in U.S. Cities: A Decision-Support Tool for Implementation and System Performance</title>
      <link>https://rip.trb.org/View/2786705</link>
      <description><![CDATA[Cities across the U.S. are central to advancing the next phase of autonomous vehicle (AV) deployment as these technologies move from pilot programs to broader implementation. As industry investment and activity accelerate, public agencies are responsible for enabling safe, efficient, and scalable operations within transportation systems. While some cities have developed operational experience, many are still establishing the policies, processes, and institutional capacity needed to support deployment at scale. This creates a timely opportunity to provide practical, evidence-based guidance to help agencies manage operations, coordinate with industry, and integrate autonomous systems effectively. This project will develop an AV Deployment Decision-Support Tool for Cities to support the efficient expansion of AV services across U.S. markets. The research will engage cities grouped into four categories: mature, growing, new, and potential markets. The study follows a structured three-phase approach. First, a baseline review will document deployment practices, regulatory approaches, and operational processes across cities. Second, targeted interviews with public agencies and national organizations will be conducted to identify key issues and refine discussion topics, followed by four online focus groups to examine common challenges, operational needs, and scalable strategies across different stages of deployment. Third, a one-day in-person workshop will validate findings, compare approaches, and identify shared priorities. The project will deliver a practical, structured decision-support tool that helps cities navigate deployment processes and reduce uncertainty as AV technologies scale. By supporting streamlined and efficient deployment practices, this work contributes to improved system performance and strengthens U.S. leadership in transportation innovation.]]></description>
      <pubDate>Mon, 05 Oct 2026 05:39:02 GMT</pubDate>
      <guid>https://rip.trb.org/View/2786705</guid>
    </item>
    <item>
      <title>Strategic Mapping of UAV Test Sites and Corridors to Support Emerging Development and Workforce Training</title>
      <link>https://rip.trb.org/View/2784730</link>
      <description><![CDATA[Develop a comprehensive network of test sites and corridors for electric aviation to assist policymakers, developers, and workforce training institutions in making informed, long-term investments.]]></description>
      <pubDate>Wed, 30 Sep 2026 18:29:43 GMT</pubDate>
      <guid>https://rip.trb.org/View/2784730</guid>
    </item>
    <item>
      <title>Development of a Zero-Emission Vehicles Incident-Management System for Arizona’s Roadways</title>
      <link>https://rip.trb.org/View/2780396</link>
      <description><![CDATA[As zero-emission vehicles (ZEVs) equipped with battery-electric and hydrogen fuel cells become more common in Arizona, managing roadway incidents involving these vehicles presents new challenges. Crashes involving ZEVs cause problems for first and second responders because there is a gap in procedures for addressing the unique safety hazards of ZEVs, including electrical systems that can produce up to 1,000 volts of electricity and on-board fuel storage systems that contain compressed hydrogen at up to 700 bar operating pressure. ZEV incidents can place both the public and emergency responders in jeopardy, and the damage to transportation infrastructure caused by ZEV failures can create problems for department of transportation staff tasked with restoring roadways to full functionality. Many first and second responders reported limited opportunities for learning how to respond to ZEV incidents in a coordinated fashion, and police, emergency medical services (EMS), and tow truck drivers are all less likely than firefighters to have ZEV-specific safety training. Developing protocols that coordinate responses to ZEV emergency incidents among various Arizona Department of Transportation (ADOT) units and other state and local emergency management agencies could help keep emergency responders and the traveling public safe on Arizona’s roads, as well as reducing congestion by helping ADOT improve response times to emergencies.
]]></description>
      <pubDate>Fri, 18 Sep 2026 16:27:13 GMT</pubDate>
      <guid>https://rip.trb.org/View/2780396</guid>
    </item>
    <item>
      <title>Public Transit Automation and the Future of Service Delivery: Scenario Planning for Operations and Workforce Readiness</title>
      <link>https://rip.trb.org/View/2780207</link>
      <description><![CDATA[Public transit agencies are beginning to confront a new generation of automation technologies that will almost certainly alter how service is planned, operated, staffed, and delivered. These technologies include fully automated transit vehicles as well as nearer-term applications in scheduling, dispatching, maintenance, customer information, and operations management. In this project, the research team will examine how transit automation could reshape the relationship between agency operating costs, service quality, and workforce needs. Labor is one of the largest components of transit operating costs; if automation eventually reduces the marginal cost of service delivery, agencies may be able to consider service models that are difficult to sustain today, including much higher frequencies, longer service spans, more extensive feeder service, and more responsive off-peak operations. At the same time, automation could significantly alter job tasks, training requirements, safety responsibilities, staffing models, and long-term employment opportunities.

The research will combine a national practice review, approximately 20–25 semi-structured interviews, at least two focus groups, and scenario planning. Participants will include transit agency staff, labor representatives, riders, advocates, researchers, consultants, technology developers and vendors, and national transit organizations such as the American Public Transportation Association (APTA). The work will identify opportunities and risks associated with automation, including potential changes in service frequency, service span, responsiveness, job tasks, training, staffing, safety oversight, procurement, and public communication. It will culminate in a scenario matrix and Automation Workforce Readiness Checklist that help agencies assess automation before pilot testing or deployment.]]></description>
      <pubDate>Fri, 18 Sep 2026 10:34:51 GMT</pubDate>
      <guid>https://rip.trb.org/View/2780207</guid>
    </item>
    <item>
      <title>Household Vehicle Holdings and Composition: A Multi-Dimensional Analysis of Body and Fuel Type Choices</title>
      <link>https://rip.trb.org/View/2780206</link>
      <description><![CDATA[Household vehicle ownership is the foundation of personal mobility for the vast majority of American families. The number of vehicles a household owns, and the specific types of vehicles in its fleet, determine how household members access employment, healthcare, education, and services, and directly generate the travel demand that shapes roadway utilization, network performance, and infrastructure needs. Accurately representing how households make vehicle fleet decisions is therefore essential for travel demand forecasting, long-range transportation planning, and the design of policies that support reliable and efficient system operations.

While a substantial body of research has examined vehicle holdings, several important gaps remain. Most existing studies treat vehicle ownership as a single aggregate outcome, failing to jointly model the count of vehicles owned alongside the specific body-type and fuel-type composition of the household fleet. Even studies that examine fleet composition, often using the multiple discrete continuous extreme value (MDCEV) modeling approach (Bhat and Sen, 2006; Augustin et al., 2015; Chakraborty et al., 2023), identify only whether at least one vehicle of each type is present, rather than modeling the count of vehicles in each category. This is a meaningful limitation as many households own more than one vehicle of the same body type, and the specific count in each category has direct implications for household mobility capabilities and system-level demand generation. Further, existing studies do not allow for a direct effect of total fleet size on preferences for specific vehicle types. This is an important consideration because a household adding a second or third vehicle faces a fundamentally different choice context than one purchasing its first, and the body type preferred as a secondary addition to an existing fleet may differ substantially from a household's primary vehicle choice. Failing to account for these factors limits the accuracy of fleet-level demand forecasts and the ability of planners to anticipate how household vehicle availability translates into system demand.

This project addresses these gaps by developing a multivariate ordered-response probit (MORP) framework that jointly models household fleet size alongside the count of vehicles in each of twelve body-type and fuel-type combinations, including six body types (SUV, sedan, pickup truck, compact, minivan, and sports car) and two fuel types (internal combustion engine and electric). A key methodological contribution is the normalization of the likelihood function to ensure that the sum of vehicle counts across all body-type and fuel-type categories equals the total fleet count, eliminating infeasible outcome combinations while enabling direct estimation of the effect of total fleet size on preferences for specific vehicle types. The analysis uses data from the 2024 Transportation Heartbeat of America (THA) survey, a large national sample of households, supplemented with built environment data from the EPA Smart Location Database. The modeling framework also incorporates three latent attitudinal constructs capturing the degree to which individuals prioritize sustainability in transportation decisions, the extent to which safety considerations shape personal travel choices, and the degree to which individuals view vehicles as social signals rather than functional assets. The inclusion of these attitudinal variables reveals the indirect pathways through which household demographics shape vehicle fleet composition, providing specific avenues for policy changes to influences vehicle holdings.

The results will enable metropolitan planning organizations and state DOTs to improve vehicle fleet representation in activity-based travel demand models, replacing aggregate ownership assumptions with behaviorally grounded estimates of]]></description>
      <pubDate>Fri, 18 Sep 2026 10:31:19 GMT</pubDate>
      <guid>https://rip.trb.org/View/2780206</guid>
    </item>
    <item>
      <title>Modeling Preferences for Autonomous Vehicle Access Types and Usage Intentions</title>
      <link>https://rip.trb.org/View/2780202</link>
      <description><![CDATA[As autonomous vehicle (AV) technologies continue to advance, it is increasingly important for transportation agencies to understand how individuals will access and use these technologies, as different access models have substantially different implications for travel demand, system performance, and infrastructure investment. This project examines preferences across two primary AV access modes, private ownership and ridehailing-based services, and analyzes how these preferences relate to anticipated AV usage behaviors. The study considers three key usage dimensions: willingness to use AVs for transporting children, anticipated in-vehicle time use, and willingness to relinquish driving control. In addition, the analysis incorporates three individual-level constructs that influence AV-related decisions, including perceptions of AV safety, prior experience with advanced vehicle technologies, and affinity toward shared mobility. Using data from the 2024 California Vehicle Survey, the project develops an integrated behavioral modeling framework that jointly represents AV access preferences and usage intentions. The results will provide actionable insights for improving travel demand forecasting, supporting infrastructure and fleet planning, and informing policy strategies related to AV deployment, system operations, and long-term transportation system performance.]]></description>
      <pubDate>Fri, 18 Sep 2026 10:17:18 GMT</pubDate>
      <guid>https://rip.trb.org/View/2780202</guid>
    </item>
    <item>
      <title>Planning and Policy for Safer Roads with Autonomous Vehicles: Decision Making Behavior in Dilemma-inducing Situations - Phase II</title>
      <link>https://rip.trb.org/View/2778266</link>
      <description><![CDATA[Despite the advances made in the technical availability of autonomous vehicles (AVs) for road tests and even limited commercial mobility services, a critical challenge in the integration of AVs into the real-world transportation systems and our lives is establishing safety and ethical regulations. In particular, principles are required for an AV to make safe and ethical decisions in high-impact low-frequency situations that induce dilemmas, such as whether to hit pedestrians to spare passengers onboard? The absence of such policies can concern the potential users about safety aspects of AVs, which is identified to negatively affect the public acceptance of the AV technology, as highlighted by the findings of the PI’s previous research.

This project is focused on designing these policies derived from the research on the potential autonomous system users’ expectations and perception of a safe AV through exploring how the users themselves would make decisions that entail a dilemma-inducing situation. This project will be the second Phase of the ongoing project from the previous year. In particular, this proposed work will be built on and complement the first Phase through accomplishing three tasks explained below.

Task 1 - Launching Stated Preferences (S.P.) Survey to Collect Sample Datasets on Two Groups of Individuals. The S.P. survey is designed in Phase I using experimental design methods which is demonstrated to be a rigorous and efficient tool for collecting datasets on individual persons’ opinions and preferences. In Phase II, the results of the pilot dataset collected in Phase 1 will be utilized for updating the survey. Then, the survey will be fully launched on two population groups with distinctive heterogenous attitudes. The PI has the experience of designing and dissemination S.P. surveys in a different context.

Task 2 - Estimating Modeling Frameworks Capturing Heterogeneity of Behavior on the Datasets Collected in Task 1. This task focuses on estimating methods developed in Phase I which are built on behavioral economics, mathematical psychology, and cognitive science. These methods are capable of capturing morality dimension of decisions as well as the heterogenous behavior of individual persons. One of such methods is choice theory-based model of latent class choice, which is capable of capturing “reason-based” morality and at the same time endogenously classifies persons into groups to be homogenous with respect to attitudes within the groups and heterogenous across the groups. The PI has the experience of developing and estimating such models in different contexts. The models are then empirically estimated on the datasets collected in Task 1.

Task 3 - Integrating Research into Teaching. The research findings from Tasks 1 and 2 will be presented to the students of the course on Transportation Engineering. The purpose is familiarizing undergraduate students with advanced research topics through presenting the findings of this study. This will then be followed by a discussion session by groups of students who will be sharing their ideas and discussion results to all their peers.]]></description>
      <pubDate>Fri, 18 Sep 2026 10:04:13 GMT</pubDate>
      <guid>https://rip.trb.org/View/2778266</guid>
    </item>
    <item>
      <title>Evaluating Autonomous Vehicles’ Pedestrian Safety Benefits in Road Networks</title>
      <link>https://rip.trb.org/View/2778253</link>
      <description><![CDATA[Pennsylvania experienced more than 3000 pedestrian-related injuries in 2022, according to PennDOT, and the Governor’s Highway Safety Association estimates that Pennsylvania ranked 10th among all 50 states in 2022 for the most pedestrian fatalities. This project will explore and evaluate the potential for connected autonomous vehicles (CAVs) to reduce these pedestrian-involved incidents. While many past works have explored using CAV capabilities to detect and avoid pedestrians, the research team will specifically focus on ensuring that these benefits are felt where and when they are most needed.

Recent studies show that pedestrian fatalities in the United States are especially common at night and on roads that do not have infrastructure designed to support pedestrians. By increasing the concentration of CAVs, which may be safer for pedestrians, on such dangerous roadways, the team may be able to increase overall pedestrian safety. Many pedestrian-related incidents, for example, involve human driver distraction, and CAVs can make use of non-visual (e.g., radar) environmental data to detect pedestrians, especially at night. However, CAVs may also be more dangerous for pedestrians if their detection systems fail. Moreover, intentionally diverting CAVs to areas that are dangerous for pedestrians may have implications for safety and traffic flow through the road network as a whole. For example, proactively routing CAVs to these areas may create more traffic congestion and potentially increase accidents involving vehicles, leading human-driven vehicles to avoid this area. Indeed, CAVs will likely be deployed initially in mixed-autonomy settings with human-driven vehicles, which adds an additional layer of complexity in analyzing how pedestrians will affect safety or traffic flow.

The team plans to build on their 2023-4 Safety21 project investigating CAVs’ safety effects in road networks. In particular, the team will make use of their mixed autonomy simulator from the previous project, which includes models of vehicle-involved accidents but does not explicitly seek to improve pedestrian safety. The work will comprise four parts: (i) integrating pedestrian fatality and injury statistics into the mixed autonomy simulator, (ii) developing analytical models for traffic flow and pedestrian injuries/fatalities given CAV routing, (iii) designing algorithms for CAVs to minimize pedestrian fatalities and injuries, and (iv) using the simulator and the analytical models to evaluate whether CAVs can effectively reduce pedestrian injuries and fatalities, as well as any tradeoffs with increasing traffic congestion.]]></description>
      <pubDate>Thu, 17 Sep 2026 17:12:04 GMT</pubDate>
      <guid>https://rip.trb.org/View/2778253</guid>
    </item>
    <item>
      <title>Deep Reinforcement Learning Based Driving Strategies Bootstrapped by Suboptimal Policies</title>
      <link>https://rip.trb.org/View/2778233</link>
      <description><![CDATA[Designing autonomous driving systems requires a modular approach. This includes vehicle sensing, perception, localization, scene representation, path planning, decision-making, and vehicle control. The study focuses on decision-making in complex driving scenarios. Traditional methods often use model-based, heuristic, or rule-based controllers.  Heuristic and rule-based controllers rely on human expertise to handle different driving situations. They include safety measures and rules to avoid deadlocks. However, these methods face limitations as designing rules to cover every possible driving scenario is highly challenging and resource intensive.

Reinforcement learning (RL), on the other hand, offers a self-supervised approach, learning control policies through interaction with the environment. With advancements in neural networks, deep reinforcement learning (DRL) has shown the ability to efficiently map state-action transitions in high-dimensional environments. 

However, a persistent challenge in RL is achieving sufficient exploration of the environment. Research has shown that RL agents require extensive exploration to accurately map state-action values and discover high-reward states. Even with mechanisms like stochastic DRL algorithms that aim to embed exploration throughout the training process, agents often fail to discover optimal policies in complex scenarios. As a result, RL agents struggle with delayed rewards, where limited and inefficient exploration can result in sub-optimal performance.

Driving environments encompass a range of challenging scenarios, such as roundabouts, overtaking, merging, intersections, and highway exits. These scenarios involve delayed rewards, which discourage exploration and complicate the RL learning process. Examples of delayed rewards include the cumulative benefit of using a free lane to achieve optimal cruising speed, waiting for the right moment to enter a roundabout safely, or merging into traffic while avoiding collisions—even if it requires temporarily stopping in the lane. Failure to explore and identify these states during training can lead to sub-optimal policies or even unsafe behaviors, such as remaining stuck behind a slow vehicle, rushing into a roundabout prematurely, or misjudging merges in traffic. 

In this study, the research team proposes to address these limitations by combining the stability of rule-based sub-optimal controllers with the adaptive learning capabilities of DRL. The team aims to utilize sub-optimal policies to bootstrap efficient training during the RL learning process and to overcome exploration barriers. Specifically, the team leverages a rule-based controller to prevent deadlocks and guides RL agents toward promising states, improving their ability to learn optimal driving strategies. This approach not only improves collision avoidance but also enhances training efficiency by guiding agents through challenging scenarios where rewards are delayed. By integrating suboptimal strategies, the method strikes a balance between simplicity and adaptability, enabling agents to learn more effectively.

The experiments are conducted in a complex driving scenario where the driving policy must optimize long-term rewards. The key contributions of this study are summarized as follows:
(1) The sub-optimal policies: The team designs sub-optimal controllers capable of making driving decisions under reasonable motivations.
(2) Guidance for RL agents: The sub-optimal controllers guide RL agent toward exploring states associated with delayed rewards, improving exploration and learning efficiency.
(3) Validation in complex driving scenarios: The team validates the effectiveness of the method in complex driving scenarios, demonstrating its ability to optimize long-term rewards and handle challenging conditions effectively.]]></description>
      <pubDate>Thu, 17 Sep 2026 16:49:20 GMT</pubDate>
      <guid>https://rip.trb.org/View/2778233</guid>
    </item>
    <item>
      <title>Safe Driving Across Domains</title>
      <link>https://rip.trb.org/View/2778232</link>
      <description><![CDATA[This project tackles the critical research question: How can pre-trained autonomous driving models be seamlessly and dynamically adapted across diverse operational design domains (ODDs) to achieve safe and robust deployment? ODDs define the specific conditions under which an autonomous driving system can operate safely—spanning factors like road types, traffic density, weather patterns, and geographic features. For instance, a model trained to navigate Pittsburgh’s urban intersections must also adapt to Pittsburgh’s rural routes with sparse signage and rolling hills, or even underdeveloped, unpaved, mixed-traffic roads. What is the right way for a robot to, in real time, smoothly switch between different planning trajectories to ensure safety and consistency amid changing driving conditions?

By focusing on adapting pre-trained models rather than retraining them from scratch, this project offers a scalable approach to ensure safe deployment. This is particularly critical in data-limited settings—collecting training data for every possible scenario can cost millions and take years. For example, equipping autonomous systems to operate in rural areas avoids costly new data collection while preserving high safety standards. A single deployment failure in an untested environment could cost lives and erode public trust, emphasizing the importance of rigorous adaptation. With an estimated 1.35 million road traffic deaths annually worldwide in drastically different driving conditions, this project aims to ensure that autonomous driving technologies can be deployed safely. In other words, this research algorithmically and proactively catalyzes translation of new technologies to data-limited areas, instead of passively relying on commercial technology spillover – the latter would take years or even decades. 

The research team leverages diverse datasets to ensure robust testing. These include driving data from TIER IV in Japan, and over 10 million miles of publicly available driving data from Waymo, collected across 25 U.S. cities and capturing diverse urban environments. In Pittsburgh, the team has access to truck-mounted camera data that covers mixed urban and rural routes, while stationary cameras funded through a PITA partnership with RIDC capture real-time trajectory data in high-traffic industrial areas. Simulation-based fine-tuning complements these datasets, addressing data gaps in under-documented scenarios, such as rural roads with limited signage. This rich dataset ecosystem provides the foundation for testing and validating model adaptations in diverse operational design domains.

The project employs open source, cutting-edge algorithms, including inverse reinforcement learning (RL), guided diffusion, blueprints, and fine-tuning. Inverse RL provides an overall framework to complement rule-based modules to directly and flexibly extract algorithm parameters (sometimes implicitly) from expert (human or simulated) driving. Guided diffusion allows for precise and smooth generation and adjustments of such parameters in constrained domains, such as adapting autonomous driving algorithms from Pittsburgh's dense urban streets to its hilly rural roads. Blueprinting ensures scalable and structured design of template models that are transferable across diverse contexts, while fine-tuning efficiently adapts pre-trained models to novel environments, particularly in low-data settings. Open source models already implement these algorithms. The team adapts these powerful emerging methods to address challenges in rural areas and infrastructure-limited regions.]]></description>
      <pubDate>Thu, 17 Sep 2026 16:37:29 GMT</pubDate>
      <guid>https://rip.trb.org/View/2778232</guid>
    </item>
    <item>
      <title>Cost-Effective Radar-Based Depth Perception and Scene Interpretation</title>
      <link>https://rip.trb.org/View/2778231</link>
      <description><![CDATA[One of the current challenges for autonomous vehicles (AV) is providing provable safety in urban driving and highway driving. LIDAR provides accurate depth and shape information, but remains quite expensive. In this research the research team explores the viability of fusing 4D radar images with camera images as an alternative to using LIDAR. The research aims to establish the data fusion of radar and camera images in autonomous driving as a replacement for data fusion using traditional LIDAR, radar, and camera images. The current approach demonstrates that a trained radar detector can achieve approximately 3.5 meters of Unidirectional Chamfer Distance (UCD) against ground truth LiDAR data on a dataset recently released by Delft University of Technology (DUT) in the Netherlands, which is 28%  better than the state of the art (SOTA).

The approach integrates the images obtained from deep neural network (DNN) 4D radar detectors with conventional camera RGB images. It introduces a novel pixel positional encoding algorithm inspired by Bartlett’s spectrum estimation technique. This algorithm transforms radar depth maps and RGB images into a unified pixel image subspace facilitating the learning of their potential correspondence. The method effectively leverages high-resolution camera images to train radar depth map generative models, addressing the limitations of conventional radar detectors in complex vehicular environments while sharpening the radar output. The research team develops spectrum estimation algorithms tailored for radar depth maps and RGB images, a comprehensive training framework for data-driven generative models, and a camera-radar deployment scheme for AV operation.

The research plan is as follows: (1) Develop a deeper analysis of the training framework to develop more advanced deep learning-based radar detectors.
(2) Scale current results on about 100 seconds of representative AV driving from a DUT dataset to a significantly larger number of frames, including a wider range of training and testing scenarios.
(3) Quantify additional performance benefit from real-time (instead of current offline) fusion schemes and deployment frameworks for 4D radars and cameras.
(4) Use the existing data to measure the underlying latency in the proposed data fusion of 4D radar and camera images. This will help quantify the capabilities and limitations of the scheme.
(5) Conduct a detailed error analysis to understand the distribution of error between the point cloud the team generates with the data fusion approach against the ground truth. This will help establish any existing bias in the error the team obtains in the UCD.
(6) DUT’s 4D radar training data set was generated in the Netherlands. The team will investigate the augmentation of these data by data collected in collaboration with the deployment partner Blue Fusion, who  will allow us the use of one or more of their high-performance radar. This effort will address one of the important pitfalls of machine learning systems, distribution shift or domain adaptation.

The results will be instrumental in providing a cost-effective and weather-resilient solution to the perception part of the provable safety challenge in L4 and L5 Autonomous Vehicles.]]></description>
      <pubDate>Thu, 17 Sep 2026 16:33:53 GMT</pubDate>
      <guid>https://rip.trb.org/View/2778231</guid>
    </item>
    <item>
      <title>Smart Trailer Technologies for Data-Driven Safety, Efficiency, and Policy Innovation in Heavy-Duty Fleet Operations</title>
      <link>https://rip.trb.org/View/2778229</link>
      <description><![CDATA[In 2022, 4,764 fatalities in the United States involved heavy-duty trucks, with 66% of these deaths affecting passenger vehicle occupants. Many resulted from underride collisions, where smaller vehicles slide under the high-clearance sides or rear of trailers, often causing catastrophic outcomes. Despite the severity of this issue, the transportation industry lacks a clear, standardized definition of “smart trailers” and guidelines for evaluating and mitigating such risks. This absence hinders the development of cohesive safety policies, the adoption of effective safety measures, and the creation of implementation roadmaps. Unlike Europe, where side underride guards are mandated and proven effective, the U.S. has yet to establish similar frameworks, contributing to uncertainty for policymakers, fleet managers, and technology developers.

This project addresses these challenges by clarifying the definition of smart trailers and establishing a data-driven framework to evaluate their safety, efficiency, and economic benefits. Smart trailers integrate advanced sensors, telematics, and artificial intelligence (AI) to monitor operations, predict risks, and optimize maintenance schedules. A key focus is determining the direct and indirect costs and benefits of equipping trailers with side and rear underride guards and evaluating proactive warning systems or additional technical and policy solutions to reduce trailer-related collisions. Tackling the limited real incident data, the study will integrate data-driven analysis with physics-informed machine learning to guide policymakers in developing practical, safety-enhancing regulations.

To achieve these objectives, the project will synthesize data and physics-based models from multiple sources, including accident reports, vehicle physics models, historical crash data, regulatory documents, and insurance industry records. This information will support the development of simulation-driven models to assess both the safety and economic implications of underride guards or alternative solutions. By leveraging these models, the research will provide actionable insights into the optimal design and implementation of safety measures and how existing smart trailer technologies can support evidence-based regulatory decisions.

Collaboration with key partners will ensure the research's practical relevance and impact. Clarience Technologies will evaluate current smart trailer technologies, identify limitations, and help define a standardized framework. Safety Emissions Solutions will contribute inspection and regulatory data to enhance cost-benefit analyses, while the Regional Industrial Development Corporation (RIDC) will assess how smart trailer technologies can address regional transportation challenges and resource allocation challenges. Pitt Ohio’s participation in validating the system through real-world testing and operational data collection is tentative and subject to confirmation.

The outcomes of this project include a standardized definition of smart trailers, recommendations for safety policy, and a framework for adopting advanced safety technologies. By addressing critical gaps in definitions, data, and evaluation methods, this research will guide the development of effective regulations and implementation roadmaps. Through collaboration with industry leaders, this project aims to enhance public safety, operational efficiency, and sustainable fleet management, paving the way for a safer and more technologically integrated future in heavy-duty transportation.]]></description>
      <pubDate>Thu, 17 Sep 2026 16:22:36 GMT</pubDate>
      <guid>https://rip.trb.org/View/2778229</guid>
    </item>
  </channel>
</rss>