Urban Currents

Using Connected Vehicle Data to Quantify Speeding Behavior on Large Networks

This study analyzed speeding behavior across more than 120,000 highway segments in a statewide network using 3.5TB of Wejo connected vehicle trajectory data. The researchers extracted maximum hourly speeds, compared them against 13 speed limit thresholds from 35 to 90 mph, and developed a speeding index to pinpoint unsafe road segments. The analysis also evaluated temporal variations across seasons, holidays, and off-peak hours, validating the results against INRIX speed data.

Why it matters — It demonstrates how high-resolution connected vehicle data can replace or supplement traditional, localized traffic monitoring to identify speeding hotspots dynamically across an entire state's road network.

Caveat: The accuracy of the speeding index and its alignment with benchmark data depends on having a sufficiently large sample size of connected vehicles on any given segment.

Transportation Research Record Journal of the Transportation Research Board · doi

URBAN CLIMATE GOVERNANCE AND THE UNEVENNESS OF CITY NETWORKS: The Trajectory and Perspectives of the Global Covenant of Mayors for Climate and Energy in Chile

This study investigates the integration of Chilean local governments into the Global Covenant of Mayors for Climate and Energy (GCoM), examining their motivations for joining and the practical challenges they face. The research reveals that while membership provides visibility, technical assistance, and knowledge exchange, local administrations struggle with severe financial and human resource constraints. These global commitments often impose administrative burdens that compete with existing national environmental initiatives and lack the practical instruments required for local execution.

Why it matters — It demonstrates that joining global climate networks is insufficient for local action without strong mayoral leadership and the capacity to leverage private capital. This exposes how global climate governance can function as a neoliberal mechanism that shifts responsibilities to municipalities without transferring the resources needed to fulfill them.

Caveat: The findings are based on a qualitative analysis of a single national context, Chile, which may limit generalizability to countries with different national-to-local governance structures.

International Journal of Urban and Regional Research · doi

GHOSTLY ENCOUNTERS IN LAGOS: Virtual Reality between Urban Marketing and Media Activism

This study analyzes the political role of virtual reality (VR) in shaping Lagos's urban future by contrasting its use in top-down and bottom-up contexts. Based on fieldwork in Nigeria, it compares the exclusive, high-end marketing of the Eko Atlantic City megaproject with 360-degree videos produced by local media activists to document the daily struggles of residents. The analysis introduces a theoretical framework combining Walter Benjamin's concept of phantasmagoria with the literary insights of Amos Tutuola to explain how immersive media can both mystify and democratize urban space.

Why it matters — It demonstrates how immersive technologies, often reserved for elite real estate marketing, can be subverted by local activists as tools for sensory counter-narratives and urban resistance.

Caveat: The study relies on qualitative fieldwork and theoretical synthesis rather than quantitative metrics of VR's reach or impact on the general public in Lagos.

International Journal of Urban and Regional Research · doi

Exploring the relationship between e-cycling and additional moderate to vigorous intensity physical activity – A series of N-of-1 studies using accelerometer data

This study analyzed the relationship between e-cycling and overall physical activity using thigh-mounted accelerometer data from 18 university employees in Norway over 52 to 85 consecutive days. Using individual time-series dynamic regression modeling, the research tracked whether commuting via e-bike replaced or added to other moderate-to-vigorous physical activity. The analysis revealed that for almost all participants, e-cycling did not reduce other physical activities, while total cycling showed mixed individual variations with three participants showing negative and two showing positive associations.

Why it matters — It provides objective, device-based longitudinal evidence supporting the independence hypothesis, demonstrating that e-cycling acts as an additional source of physical activity rather than displacing existing exercise habits.

Caveat: The findings are based on a very small sample of 18 employees from a single Norwegian university, which may limit how well these behavioral patterns generalize to broader populations.

Journal of Cycling and Micromobility Research · doi

Small talk in a small vehicle: a qualitative study of a shared automated vehicle service integrated in public transport

This study analyzes the user experience of early adopters using an SAE Level 2 shared automated vehicle service integrated into a public transport system. Based on 12 semi-structured interviews, the researchers applied reflexive thematic analysis, the SAVA model, and Social Practice Theory to examine how trust, utility, and social comfort intersect with users' daily routines.

Why it matters — It shifts the understanding of autonomous vehicle acceptance from isolated individual attitudes to a set of interdependent practices, demonstrating that acceptance depends on how well the technology fits into existing physical infrastructures, social norms of sharing, and institutional trust.

Caveat: The findings are based on a small qualitative sample of 12 early adopters of a service that still utilized a visible safety driver, which may not represent the broader public's reaction to fully driverless vehicles.

Travel Behaviour and Society · pdf · doi

Urban revitalization between proper and improper consumption: insights from Downtown Rio de Janeiro

This study analyzes the consumption-led revitalization of Downtown Rio de Janeiro by examining policy documents and public statements from local state representatives and stakeholders. It traces how the city's revitalization campaign uses curated commercial and cultural activities to destigmatize the area, establishing moral geographies that categorize consumption into 'proper' and 'improper' circuits. The analysis details how this categorization justifies the heightened regulation, surveillance, and marginalization of informal street vendors by associating them with crime and disorder.

Why it matters — It extends critical urban scholarship on consumption-led development beyond Northern contexts, demonstrating how Southern revitalization strategies use moral classifications of commerce to actively police and exclude informal workers from public space.

Caveat: The findings are based on a qualitative analysis of policy documents and stakeholder discourse rather than empirical measurements of vendor displacement or economic impacts.

Urban Geography · doi

We will stay-put or die remaining: why some communities choose active immobility in conflict zones in Nigeria

This study analyzes why communities in Nigeria's Middle Belt choose to remain on their ancestral lands despite facing recurring armed attacks and a lack of state protection. Using focus group discussions conducted in 2022 across Benue and Nasarawa States, the research applies reflexive thematic analysis within a resilience-accessibility framework to examine the material and symbolic drivers of this active immobility.

Why it matters — It provides empirical evidence of voluntary immobility in active conflict zones, demonstrating that the decision to stay is driven by a collective desire to preserve local lifescapes and maintain access to ancestral land resources rather than a simple lack of resources to flee.

Caveat: The findings are based on qualitative focus group data from two specific states in Nigeria, which may reflect localized cultural and land-tenure dynamics that differ from other conflict zones.

Mobilities · pdf · doi

Thermal mobilities: heat stress and the right to the city amongst Phnom Penh’s climate vulnerable paratransit workers

This study investigates the physical and social impacts of extreme heat on informal paratransit workers, specifically tuk-tuk and motodop drivers, in Phnom Penh. Using qualitative interviews, observations, and a longitudinal cohort study with wearable physiological sensors, the research tracks how heat stress causes physical symptoms like dizziness and fatigue. A quantitative intervention using wearable devices demonstrates that providing access to cooled spaces reduces the duration of elevated core body temperatures by over 50%.

Why it matters — It establishes the concept of 'thermal mobilities' to show how heat exposure is socially and spatially structured, demonstrating that physical cooling interventions must be paired with overcoming social stigmas that bar informal workers from air-conditioned urban spaces.

Caveat: The physiological findings are based on a specific cohort of paratransit workers within a single city, which may limit direct generalizability to other informal labor sectors or geographic regions.

Mobilities · pdf · doi

Adaptive traffic signal control using deep learning and metaheuristic-optimized Kernel Extreme Learning Machine

This study develops an adaptive traffic signal control framework that combines YOLOv8 vehicle detection with a Kernel Extreme Learning Machine classifier optimized using metaheuristic algorithms. Tested in a custom simulation environment over 30 independent runs, the YOLOv8 model achieved a 92% mAP@50 for vehicle detection, while the Genetic Algorithm-optimized classifier reached a peak classification accuracy of 91.733% for determining adaptive signal timings.

Why it matters — The framework demonstrates that dynamic signal-time reallocation can be achieved by leveraging existing roadside camera infrastructure, offering a highly accurate, data-driven alternative to traditional fixed-time traffic control systems.

Caveat: The system's performance and signal reallocation capabilities were evaluated entirely within a custom simulation environment rather than on real-world physical intersections.

Transportation Letters · doi

Development of a Permanent Deformation Model to Predict Rutting Performance in Substandard Airfield Pavements

A permanent deformation model was developed to predict rutting in substandard airfield asphalt pavements using data from 34 test items trafficked with a heavy vehicle simulator and a deployable load-cart. The model simulates mechanistic responses within an unsaturated poroelastic multilayered structure under heavy aircraft loads, including the C-17, C-130, and P-8, and links these to progressive rutting using an incremental-recursive approach. Evaluation against the standard PCASE 7.0 software revealed that the existing design methodology underpredicts passes to failure for approximately 75% of the test cases.

Why it matters — The new model provides a more accurate, mechanistic-based prediction of airfield pavement degradation across varying lifespans, correcting a systematic underestimation of pavement durability found in current design tools.

Caveat: The model's performance was verified using the same historical full-scale pavement testing experiments that provided the development data.

Transportation Research Record Journal of the Transportation Research Board · doi

Central Europe: Application of the Unified Theory of Acceptance and Use of Technology Framework

This study analyzed the factors driving ride-hailing adoption among Generation Z in Slovakia and the Czech Republic using an updated 2022 version of the Unified Theory of Acceptance and Use of Technology (UTAUT) framework. Survey data collected between May and September 2024 were analyzed using partial least squares structural equation modeling (PLS-SEM). The results show that while performance expectancy, habit, and social influence shaped behavioral intentions in both countries, effort expectancy was only significant for Czech respondents, and compatibility and personal innovativeness were only significant for Slovak respondents.

Why it matters — It establishes empirical evidence for ride-hailing adoption patterns specifically within Central European youth, revealing that regional nuances exist even between closely related neighboring countries, which challenges the assumption of a uniform regional market.

Caveat: The findings are based on self-reported survey data from a specific demographic cohort, which may not reflect actual travel behaviors or generalize to older populations in the region.

Transportation Research Record Journal of the Transportation Research Board · doi

Machine-Learning-Based Traffic State Prediction in Car–Bicycle Mixed Traffic Using Synthetic Data

This study evaluated random forest, multi-layer perceptron, and linear regression models to predict traffic output flow, delay, and density in mixed car-bicycle environments. The models were trained and tested on a synthetic dataset generated from numerical evaluations of traffic flow theory. The analysis compared model performance across different data splits, revealing that while random forest excels on previously observed conditions, the multi-layer perceptron generalizes better to unseen, high-flow, and high-density scenarios.

Why it matters — It demonstrates that synthetic datasets derived from traffic theory can train machine learning models for multimodal traffic estimation, while establishing that neural networks generalize better than tree-based models when encountering novel traffic volumes.

Caveat: The models were trained and evaluated entirely on synthetic data generated from theoretical traffic flow equations rather than real-world empirical traffic observations.

Transportation Research Record Journal of the Transportation Research Board · doi

Assessing Functional Performance of Asphalt Pavements under Data Sparsity: A Probabilistic-Deterministic Approach

The study develops a hybrid probabilistic-deterministic modeling framework to evaluate asphalt pavement deterioration using the International Roughness Index (IRI). A Markov chain model categorizes pavement into five functional states for short-term forecasting without historical data, while an exponential regression model uses estimated pavement age to predict long-term IRI and remaining service life. Validation against field observations showed absolute state proportion differences between 0.0002 and 0.1116 for the short-term model, and an R-squared value between 0.84 and 0.86 for the long-term model.

Why it matters — This approach enables municipal and regional transport agencies to reliably forecast road deterioration and schedule maintenance even when they lack historical pavement performance datasets.

Transportation Research Record Journal of the Transportation Research Board · doi

Framework and Multi-modal Dataset for Roadwork Zone Detection and Geo-localization

This study introduces the Roadwork Zone Detection and Geo-localization (RZDG) dataset, a multimodal sensor dataset containing both simulated and real-world data annotated for semantic segmentation, 3D object detection, and geo-localization. The researchers also developed a tracker-based pipeline, extending the AB3DMOT model, to detect roadwork zones and map their local coordinates to global positions. Evaluated on a one-meter tolerance threshold for true positives, the pipeline achieved an F1-score of 0.597 on real-world data (0.565 precision, 0.898 recall) and 0.665 on simulated data (0.615 precision, 0.809 recall).

Why it matters — It provides the first publicly available benchmark dataset and localization pipeline specifically designed to map temporary, semi-static roadwork zones, enabling autonomous vehicle navigation systems to update high-definition maps with precise global coordinates.

arXiv · pdf

Beyond travel mode: urban context shapes active mobility's mental health effects over time

Using causal machine learning and causal deep learning on a dataset of 264,168 UK adults, this study analyzed how urban environments shape the long-term mental health impacts of active mobility. The analysis revealed that individual anxiety risk changes range from a 40.6% reduction to a 10.1% increase over time, with the greatest mental health benefits concentrated in greener, safer, less polluted, and less deprived neighborhoods. In these supportive environments, mean anxiety risk fell by 26.4%, compared to just a 7.4% reduction in the least supportive urban contexts.

Why it matters — It demonstrates that the mental health benefits of active travel are not universal but are highly unequal and contingent on neighborhood quality, proving that urban compact form only amplifies these benefits when paired with supportive environmental conditions.

arXiv · pdf

Environmental Drivers of Respiratory Disease: A District Level Analysis

Using an 11-year panel dataset (2014-2024) across all 25 administrative districts in Sri Lanka, this study built two temporally validated XGBoost models to predict annual district-level respiratory admission rates (R^2 = 0.937) and monthly PM2.5 concentrations (R^2 = 0.976). SHAP analysis revealed that cumulative air quality burden accounts for 80.1% of the variance in respiratory rates, followed by forest degradation at 15.6% and fire activity at 4.3%. A newly developed Forest-Air-Health Risk Index identified Colombo, Gampaha, and Kalutara as the highest-risk districts.

Why it matters — It provides the first district-level quantitative framework in Sri Lanka that links environmental degradation to respiratory health, resolving a historical paradox where declining hospital admissions masked rising pollution due to uneven healthcare access.

Caveat: The predictive models generalized well for most of the country but exceeded a 20% Mean Absolute Percentage Error in 4 out of the 25 districts.

arXiv · pdf

LegalFarePlan: A Label-Setting Framework for Fare-Transparent Urban Rail Route Planning under Non-Additive Fare Rules

The LegalFarePlan framework models urban rail route planning under non-additive fare rules by treating legal exit-and-reentry actions as explicit constraints. It implements Dijkstra baselines, a greedy split heuristic, bounded exact label-setting, and Pareto-frontier search, which were evaluated on a 57-station semi-synthetic benchmark with 360 origin-destination pairs. Under a 45-minute extra-time budget, the bounded exact search identified fare reductions for 71.11% of the pairs, yielding a mean reduction of 3.78 and a maximum of 9.0 synthetic fare units.

Why it matters — It provides a systematic method to identify and exploit legal fare-saving opportunities through strategic mid-journey exits and re-entries, a capability previously unaddressed by standard shortest-path transit routers.

Caveat: The performance and fare-saving metrics are demonstrated on synthetic and semi-synthetic benchmarks rather than empirical data from an active transit operator.

arXiv · pdf