<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=PHNlYXJjaD48cGFyYW1zPjxwYXJhbSBuYW1lPSJzdWJqZWN0bG9naWMiIHZhbHVlPSJvciIgLz48cGFyYW0gbmFtZT0idGVybXNsb2dpYyIgdmFsdWU9Im9yIiAvPjxwYXJhbSBuYW1lPSJsb2NhdGlvbiIgdmFsdWU9IjE2IiAvPjwvcGFyYW1zPjxmaWx0ZXJzIC8+PHJhbmdlcyAvPjxzb3J0cz48c29ydCBmaWVsZD0icHVibGlzaGVkIiBvcmRlcj0iZGVzYyIgLz48L3NvcnRzPjxwZXJzaXN0cz48cGVyc2lzdCBuYW1lPSJyYW5nZXR5cGUiIHZhbHVlPSJwdWJsaXNoZWRkYXRlIiAvPjwvcGVyc2lzdHM+PC9zZWFyY2g+" 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>Full Closure Versus Lane Closures for Freeway Maintenance
</title>
      <link>https://rip.trb.org/View/2763062</link>
      <description><![CDATA[Freeway maintenance projects are essential but challenging operations that often entail significant safety risks, high costs, and traffic disruptions. In North Carolina, most freeway maintenance is performed using the lane closure option, where one or more lanes are closed and traffic is routed through remaining lanes or shoulders. A second option is a full closure, which shuts down the entire roadway (in one direction or both) and detours traffic via alternate routes. Each option has trade-offs: full closures can accelerate construction and enhance worker safety by removing traffic from the work zone, but they inconvenience motorists with detours; partial closures allow some traffic through, reducing detour impacts, but prolong the work and expose workers and drivers to work zone hazards. Currently, guidance on when to choose full closure versus lane closures is limited, leaving engineers to make case-by-case judgments without a consistent framework. Given the stakes, a data-driven approach is needed to optimize closure decisions.
This proposal outlines a two-year, research project to investigate the safety, construction cost, and operational impacts of full closures versus lane closures for freeway maintenance. The primary focus is traffic safety: understanding how full closures (with detoured traffic) compare to partial closures (with live traffic in a work zone) in terms of crash risk for motorists and workers. Secondary objectives include quantifying construction cost and time differences and assessing operational impacts such as traffic delay and rerouting effects, using practical analysis methods (avoiding the need for full regional traffic modeling). The ultimate goal is to develop guidance for NCDOT to use when determining whether to use full closure or lane closures. This guidance will directly help NCDOT staff (both at the Division level and central offices) plan maintenance projects that minimize total harm (safety and mobility impacts) while maximizing efficiency and cost-effectiveness.
To achieve these objectives, the research team will: (1) review national and international literature and practices on work zone closure strategies, with emphasis on safety outcomes; (2) collect and analyze relevant data from past projects (in North Carolina and other states) to compare crash rates, work durations, and costs under different closure approaches; and (3) perform targeted operational analyses (e.g. using deterministic models or simplified traffic analysis tools) to estimate delays and diversion impacts without full-scale regional simulation.
Significance: By focusing first on safety, this research will fill a critical gap. For example, while it is intuitively safer for workers to have no traffic in the work zone (full closure), questions remain about the overall safety impact once detour routes are considered. Preliminary evidence from other states is mixed but informative: some full closure implementations have been associated with improved safety for both workers and travelers (and even lower overall crash rates on alternate routes), while others note the importance of careful detour planning to mitigate increased exposure of local roads to additional traffic and associated safety risks. By systematically studying these impacts and also evaluating cost and mobility outcomes, this project will produce actionable guidance. The products will include a final report and a concise guideline document that NCDOT can circulate among engineers and project managers. An implementation plan is built into the project to ensure the guidelines are effectively communicated (through workshops or webinars) and pilot-tested on a real project scenario. The outcome will empower NCDOT to make well-informed decisions that improve work zone safety, minimize traveler inconvenience, save money, and expedite project delivery.
]]></description>
      <pubDate>Wed, 19 Aug 2026 17:20:57 GMT</pubDate>
      <guid>https://rip.trb.org/View/2763062</guid>
    </item>
    <item>
      <title>Development of a Crash Modification Factor for Pedestrian Crashes at Roundabout</title>
      <link>https://rip.trb.org/View/2762930</link>
      <description><![CDATA[Roundabouts have been widely adopted as a safety countermeasure at intersections due to their ability to elimate conflict points, reduce vehicle speeds, and improve operational performance. They are classified by the Federal Highway Administration (FHWA) as a Proven Safety Countermeasure based on strong evidence of reduced serious-injury and fatal vehicle crashes when replacing stop-controlled or signalized intersections. However, while the vehicle safety benefits of roundabouts are well established, the evidence regarding pedestrian safety remains limited.
Although roundabouts intuitively offer benefits for pedestrians—such as fewer vehicle-pedestrian conflict points, lower speeds at crossings, improved sight lines, and refuge opportunities in splitter islands—the available scientific data on crash reductions for pedestrians is sparse. The few existing pedestrian crash modification factors (CMFs) are generally rated low in quality, with limited sample sizes, methodological shortcomings, and little attention to design variations such as multilane roundabouts, bypass lanes, or roundabout interchanges. As a result, agencies currently lack high-quality, generalizable CMFs to reliably assess the pedestrian safety performance of roundabouts.
This project addresses this critical gap by seeking to develop CMFs to quantify the change in pedestrian crash frequency associated with installing a roundabout at an intersection. The analysis will focus on sites previously controlled by stop signs or traffic signals. An observational Empirical Bayes before–after study design will be employed, as it is the state-of-the-art approach for roadway safety evaluations and appropriately accounts for regression-to-the-mean, crash trends, and changes in traffic volumes. In addition to controlling for traditional roadway infrastructure and traffic volume impacts, the reseach team will also attempt to incorporate pedestrian exposure through surrogate measures, such as demographic and socioeconomic characteristics, land use diversity, network connectivity, and pedestrian infrastructure, drawing on models and metrics the research team has recently developed for NCDOT. It is anticipated that the resulting CMFs may be in the form of Crash Modification Functions that provide an estimate of the safety impacts of roundabouts as a function of several design variables, if it is determined that any of these impact pedestrian safety performance at roundabouts. Candidate variables that the research team intends to explore the impact on pedestrian safety as a part of this project include:

•	Roundabout radius
•	Approach speed limit
•	Level of pedestrian activity (either direct exposure estimates or exposure surrogates) 
•	Number of travel lanes
•	Traffic volume
•	Speed limit
•	Functional classification
•	Pedestrian crossing or pedestrian infrastructure at roundabout locations 
•	Presence of pedestrian crossing aids (e.g., Rectangular Rapid Flashing Beacons or Pedestrian Hybrid Beacons) or signage

Anticipated products include a final report documenting the CMF development process, the CMFs themselves, and their interpretation. An implementation plan will also be provided, outlining how the CMFs should be applied, the level of confidence associated with the results, and any data quality limitations. The plan will include recommendations for updating NCDOT policies, guidelines, and training resources to ensure that the findings are effectively integrated into planning and design practices. By providing robust, evidence-based CMFs for pedestrian crashes at roundabouts, this research will enable NCDOT to make more informed decisions about the safety performance of roundabout projects, particularly in pedestrian-rich environments.
]]></description>
      <pubDate>Wed, 19 Aug 2026 17:02:41 GMT</pubDate>
      <guid>https://rip.trb.org/View/2762930</guid>
    </item>
    <item>
      <title>Effectiveness of School Zone Speed Limits as a Traffic Calming Strategy</title>
      <link>https://rip.trb.org/View/2762674</link>
      <description><![CDATA[Minnesota statutes grant local road agencies the authority to establish school zone speed limits (SZSLs). However, there is limited research as to the actual impacts of these SZSLs on driver speeds in school zones. This has resulted in the implementation of SZSLs without a clear understanding of the likely effects on travel speeds or the resultant impacts on pedestrian and bicyclist safety. To that end, this project will examine the effectiveness of SZSLs as a traffic calming strategy across various roadway contexts in Minnesota. The study will also examine the role of supplemental measures, which may include law enforcement and various types of traffic control devices, such as flashing beacons and enhanced signage/pavement markings. Field data collection will deploy radar sensors to collect speed data during different times of day at a diverse cross-section of schools throughout Minnesota, including those with and without SZSLs. These data will enable the research team to quantify speed reductions and to determine the effectiveness of SZSL and the supplementary measures in reducing travel speeds. The research will also synthesize current practices to clarify how school zones are defined and applied across jurisdictions. The results will provide evidence-based recommendations to help Minnesota Department of Transportation (MnDOT) and local agencies implement SZSLs more effectively and consistently, thereby enhancing safety for children walking and biking to school.]]></description>
      <pubDate>Wed, 19 Aug 2026 16:30:17 GMT</pubDate>
      <guid>https://rip.trb.org/View/2762674</guid>
    </item>
    <item>
      <title> 
Developing a Web Interface for Pipe Deterioration Models
</title>
      <link>https://rip.trb.org/View/2762052</link>
      <description><![CDATA[Pipeline infrastructure is critical to energy transportation and urban utility systems, yet aging pipelines are increasingly vulnerable to corrosion, material degradation, and environmental damage. Traditional integrity assessment relies on periodic inspection and offline analysis, providing only discrete snapshots of system condition and failing to capture the continuous, uncertain evolution of deterioration, leaving maintenance and risk-mitigation decisions reactive rather than predictive.
This project develops a dynamic probabilistic deterioration model and an interactive web-based decision-support interface. The predictive engine is a Dynamic Bayesian Network (DBN) constructed in the GeNIe modeling environment, mapping causal dependencies among pipeline condition factors (coating degradation, soil corrosivity, operating parameters) and defining time-slice transitions to capture corrosion evolution. Conditional probability tables are parameterized from historical inspection records, empirical corrosion models, and structured expert elicitation. The DBN is exported and embedded in a backend inference module that applies Bayesian updating to user inputs and computes posterior risk probabilities in real time. An interactive frontend translates the probabilistic forecasts into intuitive time-series risk curves and visual dashboards supporting 'what-if' scenario analyses.
The result is a fully functional web-based framework bridging probabilistic engineering modeling and practical pipeline integrity management. Operators and field engineers can integrate inspection data, simulate risk trajectories, forecast deterioration trends, optimize maintenance schedules, and mitigate corrosion risk before structural failures occur — shifting asset management from reactive to proactive, reducing costly emergency repairs, extending infrastructure lifespan, and mitigating risks to public safety and ecological health. Deliverables include the calibrated DBN model, open-source web interface code, implementation guidelines, a technical report, and peer-reviewed publication.
]]></description>
      <pubDate>Wed, 19 Aug 2026 16:19:45 GMT</pubDate>
      <guid>https://rip.trb.org/View/2762052</guid>
    </item>
    <item>
      <title>Quantum Bayesian Networks for Cyber-Physical Threat Prediction in Pipeline SCADA Systems</title>
      <link>https://rip.trb.org/View/2762051</link>
      <description><![CDATA[Pipeline supervisory control and data acquisition (SCADA) systems are cyber-physical systems whose expanding connectivity to enterprise networks and remote services has broadened the cyber-attack surface. Intrusions can propagate beyond digital networks to manipulate physical processes, producing operational disruption, safety hazards, environmental damage, and financial loss. Predicting such cross-domain threats is difficult because attacks are rare and governed by complex causal dependencies, and classical sampling-based inference scales poorly under rare-evidence conditions.
This project constructs a Bayesian network representing causal relationships among cyber-attacks, physical impacts, and sensor observations in a pipeline SCADA environment, informed by a hardware-in-the-loop pipeline testbed with conditional probability tables derived from operational data and engineering judgment. After establishing a classical inference baseline for posterior estimation under normal and rare-evidence queries, the network is encoded as a quantum circuit that preserves its causal structure. Three quantum inference families are implemented and evaluated: quantum rejection sampling, quantum likelihood weighting, and a variational Born-machine approach for posterior attack-probability estimation. Methods are benchmarked on a noiseless simulator and on superconducting quantum hardware, with comparison across accuracy, sample efficiency, and circuit-depth requirements.
Expected outputs include a structured probabilistic model of cyber-physical threat propagation, a quantum-circuit implementation suited to near-term hardware, a comparative evaluation report, open-source code, and peer-reviewed publications. Findings will clarify where quantum inference can improve sampling efficiency for rare-event queries in critical infrastructure security and where current hardware constrains deployment, informing risk-assessment workflows for pipeline operators, the Pipeline and Hazardous Materials Safety Administration (PHMSA), and state DOTs, and laying groundwork for quantum-enhanced digital twin and resilience frameworks.
]]></description>
      <pubDate>Wed, 19 Aug 2026 16:17:33 GMT</pubDate>
      <guid>https://rip.trb.org/View/2762051</guid>
    </item>
    <item>
      <title>Bayesian Network Modeling of Cyber-Physical Risk and System Dependencies in Railway Control Systems</title>
      <link>https://rip.trb.org/View/2762050</link>
      <description><![CDATA[Railway systems increasingly depend on integrated control, communications, and operational technologies, and this growing connectivity exposes them to cyber disruptions that can cascade across interdependent subsystems and compromise operations and safety. Existing cybersecurity assessment methods are largely static or deterministic and poorly suited to representing cascading effects in complex railway environments, motivating a probabilistic framework for understanding disruption propagation, critical vulnerabilities, and the resilience benefits of alternative mitigation strategies.
This project develops a railway-specific Bayesian network model representing major cyber-physical components, dependencies, and disruption pathways in railway control systems. The methodology first identifies and characterizes critical railway subsystems and their cyber-physical interdependencies, then defines the model structure and key variables capturing system states, disruption pathways, and operational consequences. Conditional probabilities are developed from available empirical evidence, published literature, and expert-informed estimates. Bayesian inference is applied to quantify how cyber events propagate through the system and influence safety and operational outcomes, and scenario-based resilience analysis identifies critical bottlenecks, assesses mitigation effectiveness, and compares risk-reduction strategies.
Expected products include a validated probabilistic modeling framework for cyber-physical risk in railway control systems, quantitative risk metrics for operational disruption and safety impact, scenario-based analysis tools, and methods for identifying high-impact components and vulnerability concentrations. Outputs will support railway operators, infrastructure owners, and regulators in cyber risk management, infrastructure hardening, investment prioritization, and operational resilience planning, and may inform future guidance and regulatory discussions on railway cybersecurity.
]]></description>
      <pubDate>Wed, 19 Aug 2026 16:14:19 GMT</pubDate>
      <guid>https://rip.trb.org/View/2762050</guid>
    </item>
    <item>
      <title>Analysis and Risk Management of Motorcycle, Bicycle, and Pedestrian Crashes in Minnesota</title>
      <link>https://rip.trb.org/View/2762499</link>
      <description><![CDATA[In Minnesota, the reduction in overall roadway fatalities have plateaued, as motorcycle, bicycle, and pedestrian crashes and fatalities have plateaued relative to historical trends and pedestrian fatalities have increased in 2024. The causes of the increase in frequency and severity of these crashes have not been fully accounted for, resulting in little actionable information for local agencies to address the problem. The Minnesota Strategic Highway Safety Plan (SHSP) notes that the highest proportion of fatality and serious injury crash proportions are motorcyclists, pedestrians, and bicyclists. The research team intends to examine available data to examine this issue in detail on the basis of prior research and will lead a risk assessment using the information obtained in the previous analysis to better clarify the mechanisms behind the increases in vulnerable road user and motorcycle severe injury crashes and fatalities.]]></description>
      <pubDate>Wed, 19 Aug 2026 16:13:51 GMT</pubDate>
      <guid>https://rip.trb.org/View/2762499</guid>
    </item>
    <item>
      <title>Application of Machine Learning to Investigate Grinding Parameters of a Novel Railway Grinding Machine</title>
      <link>https://rip.trb.org/View/2762049</link>
      <description><![CDATA[Rail surface defects develop through repeated wheel–rail interactions and require timely grinding intervention to maintain rail integrity, enhance safety, and avoid costly rail replacement. Although prior studies have optimized individual grinding parameters such as rotational speed, stone granularity, and feed rate, the combined influence of multiple interacting parameters on metal removal rate and surface quality remains insufficiently understood, hindering development of optimized grinding patterns.
This project develops an uncertainty-aware machine learning framework tailored to the limited size of laboratory-generated railway grinding datasets. A Gaussian Process Regression (GPR) model will predict key performance indicators — Metal Removal Rate and Surface Roughness- from inputs such as rotational speed and grinding wheel orientation while quantifying predictive uncertainty. Leave-one-out cross-validation maximizes data utilization, GridSearch optimization identifies hyperparameters, and a fixed random state ensures reproducibility; performance is evaluated with the coefficient of determination and Root Mean Squared Error. A complementary Support Vector Classifier is developed to classify grinding burn conditions, with data-imbalance mitigation via class weighting and Random Minority Oversampling, evaluated by accuracy and F1-score.
Interpretability is central to the effort: SHAP analyses will rank feature importance so railway management can identify the most influential grinding parameters, and Partial Dependence Plots will clarify individual parameter effects. Expected outcomes include unified grinding patterns that achieve target rail profiles while maximizing material removal efficiency and surface quality, supporting the industry shift from corrective to preventive grinding, extending rail service life, reducing replacement costs, and strengthening railway operational safety.
]]></description>
      <pubDate>Wed, 19 Aug 2026 16:05:23 GMT</pubDate>
      <guid>https://rip.trb.org/View/2762049</guid>
    </item>
    <item>
      <title> 
Impacts of Left-Turn Traffic Signal Phasing Strategies on Safety Outcomes
</title>
      <link>https://rip.trb.org/View/2762048</link>
      <description><![CDATA[Left-turn (LT) maneuvers are among the most challenging and highest-risk movements at signalized intersections. Transportation agencies across the United States have adopted varied signal phasing and operational strategies to manage LT movements, such as Flashing Yellow Arrow (FYA) displays in lieu of the Solid Green Ball for permissive turns, and FYA used in place of or in combination with protected-only phasing. While prior studies have examined safety performance at regional or whole-intersection levels, limited information exists at the intersection-approach level, where phasing decisions are actually implemented.
This project develops a methodology for approach-level safety analyses at signalized intersections based on left-turn signal phasing and operational strategies, focusing on crash involvement and outcomes in the vicinity of signalized intersections. The methodology quantifies and compares safety outcomes across locations, time periods, and operational strategies (e.g., before-and-after conditions; peak versus off-peak periods) using crash frequencies, severity outcomes, and exposure-adjusted rates, with analytical techniques including non-parametric statistical tests. Application is demonstrated through a Las Vegas metropolitan area case study, drawing on crash records from Nevada Department of Transportation (NDOT) and local agencies; signal, roadway, and traffic characteristics from the RTC of Southern Nevada; and vehicle trajectory and operational data (e.g., Streetlight, INRIX) from regional partners.
Results will advance the state of the science and practice for approach-level LT safety analysis and provide transportation agencies a basis to refine policies and procedures that enhance safety and operational efficiency at signalized intersections.
]]></description>
      <pubDate>Wed, 19 Aug 2026 16:02:53 GMT</pubDate>
      <guid>https://rip.trb.org/View/2762048</guid>
    </item>
    <item>
      <title> 
Research Experience for Undergraduates (REU): Smart Cities Supplement Summer 2027
</title>
      <link>https://rip.trb.org/View/2762047</link>
      <description><![CDATA[Since 2020, the University of Nevada, Las Vegas (UNLV) has hosted an NSF Research Experiences for Undergraduates (REU) Site on Smart Cities focused on advanced mobility technologies, including Intelligent Transportation Systems (ITS), Connected and Automated Vehicles (CAVs), and vehicle-to-everything (V2X) communication. The program attracts more than 100 applicants annually and has trained over 45 students, producing more than one peer-reviewed publication per summer. This project supplements the existing site to support two additional undergraduate researchers working on Research and Education for Promoting Safety University Transportation Center (REPS UTC) safety-related projects.
The project will recruit two undergraduate students from relevant disciplines (electrical and computer engineering, civil engineering, computer science), with emphasis on outreach to institutions with varied student populations. Selected students will complete a ten-week summer research experience under faculty mentorship across the ITS, CAV, and V2X focus areas, complemented by co-curricular training including weekly cohort meetings, enrichment activities, and cohort-building. Key tasks include developing compelling safety research projects with UNLV mentors, national advertising and recruitment, candidate evaluation and selection, mentored summer research, and final reporting through the UNLV Undergraduate Research Symposium and academic publication venues.
Expected outcomes include expanded undergraduate participation in transportation safety research, student projects aligned with REPS UTC initiatives, student reports, presentations, and potential peer-reviewed publications. The project strengthens the pipeline of future transportation professionals and UNLV's research infrastructure in traffic safety, helping develop the next generation of researchers and practitioners equipped to address emerging safety challenges in transportation systems.
]]></description>
      <pubDate>Wed, 19 Aug 2026 15:59:14 GMT</pubDate>
      <guid>https://rip.trb.org/View/2762047</guid>
    </item>
    <item>
      <title>Tools for Improving Visibility for Snow Plowing</title>
      <link>https://rip.trb.org/View/2762132</link>
      <description><![CDATA[The ability for snowplow operators to see ahead and around their equipment is crucial to the safe and efficient clearing and treatment of roadways during storm events. Under poor visibility, a plow operator may lose situational awareness of their surroundings, which includes other vehicles, the edge or center of the roadway, ditches, curbs, and so forth. Aside from the direct safety impacts on the travelling public in the form of a potential crash, the loss of visibility by plow operators means a need to pay even closer attention to their environment, leading to increased fatigue, stress, etc. Low visibility also increases the likelihood of striking minor objects, such as concrete curbing, leading to the potential for operator injuries to occur, not to mention infrastructure damage.
There is a need for research to investigate the available technologies that can be employed in-vehicle to improve the visibility of the roadway environment for plow operators. As autonomous vehicles become more sophisticated, the technologies they employ to view the roadway are likely transferable to activities like snow plowing. The application of these technologies (i.e., a Global Positioning System [GPS], different types of cameras, other sensors) to snowplows would provide operators with improved visibility of the road ahead. However, until an investigation is made into what technologies are available, their capabilities and costs, what agencies within Minnesota and nationally may be already using them, among other questions, the potential for widespread application of such devices remains largely untapped. The primary benefit of this research will be an understanding of the technologies and products that are available to assist plow operators in seeing the road ahead and the associated costs of those technologies. ]]></description>
      <pubDate>Wed, 19 Aug 2026 15:18:12 GMT</pubDate>
      <guid>https://rip.trb.org/View/2762132</guid>
    </item>
    <item>
      <title>Observational Intersection Traffic Safety Analysis II: Vision-Based Surrogate Safety Analysis</title>
      <link>https://rip.trb.org/View/2762043</link>
      <description><![CDATA[Intersections are among the most safety-critical locations in roadway networks, yet detailed data to support proactive intersection safety analysis remain limited beyond the basic detection used for signal controller operations. Building on Phase I advances in efficient vehicle detection and tracking, this project leverages existing camera-based detection deployments, including those in Clark County, Nevada, to move vision-based surrogate safety analysis toward technology transfer and operational deployment.
The project uses road-user trajectories produced by Phase I models to compute surrogate safety indicators, principally time-to-collision (TTC) and post-encroachment time (PET), and develops a comprehensive intersection safety score that combines surrogate measure distributions with exposure. Key tasks include obtaining access to live intersection video feeds; implementing trajectory-based safety analysis code; developing the composite safety score; deploying cloud-based, GPU-enabled computing infrastructure with backend database storage for live video analysis; and building a web-based front-end platform presenting maps, safety scores, heat maps, and detailed analytics.
Anticipated outcomes include computer vision algorithms for detecting and tracking all road users; trajectory-based indicators covering turning movement counts, pedestrian and bicyclist activity, TTC, and PET; and an operational web platform for storing, visualizing, and reporting safety measurements. Open-source code and datasets will be released publicly, and the team will coordinate with the Regional Transportation Commission of Southern Nevada to ensure tool utility, enabling agencies to identify safety risks earlier and target interventions proactively.
]]></description>
      <pubDate>Wed, 19 Aug 2026 11:52:42 GMT</pubDate>
      <guid>https://rip.trb.org/View/2762043</guid>
    </item>
    <item>
      <title>Innovating TOD Operationalization: A Safety-Oriented, Technology-Based Framework Leveraging Street-Level Imagery and Computer Vision</title>
      <link>https://rip.trb.org/View/2762042</link>
      <description><![CDATA[Traditional transit-oriented development (TOD) metrics rely on coarse, static datasets that overlook street-level safety elements such as lighting, visibility, and natural surveillance, limiting planners' ability to assess how the built environment influences transit use and perceived safety. This project closes that gap by developing and operationalizing an artificial intelligence (AI)-based TOD measurement framework that integrates Google Street View (GSV) imagery with machine-learning image segmentation to extract fine-grained, pedestrian-experience-oriented indicators of walkability and safety around transit stations.
The methodology is demonstrated for all BART stations in the San Francisco Bay Area. Conventional TOD levels are calculated from established indicators (EPA Smart Location Database, WalkScore, parcel-level land use, and accessibility measures), while a parallel image-based measurement is constructed from GSV imagery within 400–800 meter pedestrian catchment areas using PSPNet semantic segmentation to classify streetscape elements. A SafeTOD typology is then developed to classify stations based on measurable safety-related features including lighting, pedestrian and cyclist presence, visibility, and natural surveillance. Traditional and image-based measures are compared using descriptive statistics and spatial analysis (Moran's I, hotspot analysis), and both are incorporated into mode choice and mode shift models to evaluate relationships among SafeTOD scores, crime, and transit ridership.
Expected products include a validated, scalable, station-level image-based TOD and SafeTOD dataset, a documented PSPNet-based segmentation workflow and analytical codebase, a SafeTOD classification framework, and empirical evidence comparing image-based and conventional TOD indicators. Outputs are intended for direct use by the California Department of Transportation (Caltrans), metropolitan planning organizations (MPOs), transit agencies, and local planning departments to prioritize station-area investments and strengthen first- and last-mile safety strategies.
]]></description>
      <pubDate>Wed, 19 Aug 2026 11:50:46 GMT</pubDate>
      <guid>https://rip.trb.org/View/2762042</guid>
    </item>
    <item>
      <title>Exploring Role-Based Human-Centered AI Alerts with Novel Wearable Sensors and Multi-Sensory Cues</title>
      <link>https://rip.trb.org/View/2762040</link>
      <description><![CDATA[As vehicles transition between manual control and varying levels of automation in response to roadway, weather, traffic, and system conditions, drivers frequently lack a clear understanding of the vehicle's current capability boundaries, the actions expected of them, and the urgency of required responses. These gaps elevate risk during safety-critical control transitions. This project designs and empirically evaluates role-based, human-centered artificial intelligence (AI) safety alerts delivered through novel wearable sensors (a smart ring and a haptic glove) and in-cabin interfaces (seat vibration, windshield visuals, and odor cues where feasible) to improve alert meaning, driver comprehension, and driving performance.
The research proceeds in two phases using a driving simulator with conditional automation. Phase I employs repeated-measures, counterbalanced design to compare smart ring, glove-based, and seat-vibration cues during standardized, time-limited takeover events, including periods of driver distraction, with urgency manipulated through available lead time. Dependent measures include takeover and response timing, minimum time-to-collision, lane-keeping and speed stability, braking and steering profiles, control smoothness, and subjective ratings of workload, trust, clarity, and comfort. Phase II examines how AI role communication, information framing and tone (neutral, supportive, stern) affect driver understanding, workload, and trust calibration when delivered through multi-sensory interfaces. Mixed-effects models will account for repeated events within drivers.
The project will deliver implemented alert prototypes, empirical evidence on whether wearable sensors provide measurable safety benefits over traditional seat vibration, a validated approach for communicating AI support roles and tone, and design guidance for transparent, higher-meaning driver alerts. Results will support agencies, vehicle manufacturers, and suppliers in refining alert strategies that reduce confusion, improve response quality, and promote appropriate trust calibration during automated driving.
]]></description>
      <pubDate>Wed, 19 Aug 2026 11:47:24 GMT</pubDate>
      <guid>https://rip.trb.org/View/2762040</guid>
    </item>
    <item>
      <title>LIFT: Localization-Informed Flight Trajectories for UAV Resilience against GPS Loss  </title>
      <link>https://rip.trb.org/View/2762044</link>
      <description><![CDATA[Autonomous vehicles—such as autonomous cars and drones—require highly accurate localization to operate safely, yet global positioning system (GPS) alone often cannot provide the needed precision. Furthermore, many real-world applications—such as disaster response—often require localization in GPS-denied environments. Cooperative localization offers a promising alternative to GPS: connected vehicles share information to localize more accurately than any one vehicle can. The accuracy of this localization depends on how they move and the sensing geometry they create. These challenges motivate new approaches that integrate localization, sensing, and planning, which aligns closely with priorities from national agencies—such as the U.S. Department of Transportation—that are exploring sensing-based PNT solutions (link).  The research team proposes Localization-Informed Flight Trajectory (LIFT) planning to enable drone and other autonomous vehicle the ability to achieve precise and reliable localization across diverse environments, from dense urban areas to remote rural regions. This capability will improve resilience against GPS loss, which is particularly a concern in high-latitude locations (such as northern Minnesota) during geomagnetic storms. LIFT leverages sensing-enabled localization, where cooperative autonomous vehicles estimate their position relative to mobile or stationary anchors using sensor measurements (e.g., range and line-of-sight) instead of relying on GPS. Such sensing-based localization is lightweight and power-efficient, but its performance is highly sensitive to the geometric configuration of the autonomous vehicles and the anchors. This vehicle–anchor geometry directly affects localization accuracy, which in turn affects trajectory planning. Conversely, the planned trajectories influence future sensing geometry, creating a closed feedback loop among motion, sensing, and localization. This project develops LIFT, a unified framework that jointly reasoning over localization, sensing, and planning to generate UAV trajectories that maintain high localization accuracy. The team will introduce localization-informed constraints, develop real-time trajectory optimization algorithms, and validate the proposed LIFT framework experimentally in realistic GPS-denied environments. ]]></description>
      <pubDate>Wed, 19 Aug 2026 11:26:11 GMT</pubDate>
      <guid>https://rip.trb.org/View/2762044</guid>
    </item>
  </channel>
</rss>