Urban Currents

Understanding climate change vulnerability in Fiji’s informal coastal settlements

This study applies the Community Integrated Vulnerability Assessment framework to evaluate locally perceived climate vulnerabilities in Nanuku, an informal coastal settlement in Suva, Fiji. Through a participatory approach mapping community-defined human security concerns and livelihood assets, the assessment reveals how exposure to sea-level rise, flooding, and tropical cyclones is compounded by insecure land tenure, weak governance, limited financial capacity, and infrastructural deficits.

Why it matters — It provides localized, community-level evidence of how socio-institutional conditions directly constrain adaptive capacity in a high-risk Pacific Island informal settlement, moving beyond broad regional generalizations to offer actionable insights for community-led resilience planning.

Caveat: The findings are based on a participatory assessment of a single informal settlement, which may limit their direct generalizability to other coastal communities with different socio-institutional dynamics.

Discover Cities · pdf · doi

Mapping the invisible: A population synthesis approach to perceived healthcare accessibility

The study applied Iterative Proportional Updating (IPU) to harmonize survey and census data in South Carolina, synthesizing population-level indices of perceived transportation barriers at the census tract level. Regression analysis revealed that socioeconomic vulnerabilities are more strongly associated with perceived financial, time, and safety barriers to healthcare than objective spatial proximity. Notably, urban residents reported higher financial and time burdens than rural residents, and higher pedestrian network density in minority-concentrated tracts correlated with lower perceived safety.

Why it matters — It demonstrates that objective infrastructure metrics, like pedestrian network density or spatial proximity, fail to capture actual healthcare accessibility barriers, proving that infrastructure quantity cannot substitute for environmental quality and safety.

Caveat: The empirical findings and synthesized indices are limited to the state of South Carolina.

Travel Behaviour and Society · doi

Visual attention as a design variable: AI-integrated eye-tracking and semantic segmentation in urban perception analysis

This study develops a computational framework that combines wearable eye-tracking data from pedestrian routes in a mid-density American city with AI-driven semantic segmentation. The system categorizes gaze behavior into seven urban element classes to calculate a composite Urban Attention Index (UAI) based on fixation duration, visit frequency, and pupil diameter, while generating attention-weighted heatmaps from street-level imagery. Exploratory testing revealed that sidewalks, vegetation, and buildings consistently draw the highest visual engagement, while sky and signage remain peripheral.

Why it matters — It provides a reproducible, automated workflow that translates raw physiological gaze data into spatial design insights, allowing planners to test how specific streetscape configurations attract or deflect pedestrian attention.

Caveat: The framework's initial findings and attention-weighted heatmaps are based on an exploratory sample of pedestrian routes within a single mid-density American city.

Transportation Research Interdisciplinary Perspectives · doi

Evaluating Driver Behavior in Navigating Rotor Turbo Roundabout Features and Intersection Safety Performance

This study analyzed driver behavior and safety outcomes at the newly constructed rotor turbo roundabout in San Benito County, California, using descriptive analysis, statistical modeling, and extreme value theory. The analysis revealed that 5% of drivers made prohibited lane changes within the circular lanes, particularly during left turns, with lower speeds, light vehicle types, and specific vehicle headways increasing the likelihood of these violations. While the turbo roundabout eliminated speeding-related crashes and reduced overall crash severity, 76% of post-installation crashes were linked to driver unfamiliarity and right-of-way violations, and 22% were caused by navigational errors.

Why it matters — It provides some of the first empirical evidence on how US drivers adapt to the physical lane separators of turbo roundabouts, identifying specific behavioral failure points like improper lane changes and yielding failures that can be addressed through targeted signage and design adjustments.

Caveat: The findings are based on a single newly installed turbo roundabout in California, meaning the observed driver confusion and crash patterns may reflect initial novelty effects rather than long-term steady-state behavior.

Journal of Transportation Engineering Part A Systems · doi

Prospects for Green Aircraft Critical Technologies and Operational Aspects

This paper provides a comprehensive review of emerging aviation greening technologies, outlining a phased transition pathway that includes short-term fleet retrofitting, next-generation aircraft designs by 2035, and full fleet replacement to meet the ICAO Net Zero 2050 goal. It evaluates how electrification, hydrogen propulsion, and sustainable aviation fuels must be combined across different aircraft classes and mission profiles, alongside operational, regulatory, and infrastructural enablers.

Why it matters — It establishes that achieving aviation decarbonization cannot rely on a single technological breakthrough, but instead requires a coordinated mix of propulsion technologies paired with updated certification pathways and policy strategies.

Future Transportation · pdf · doi

João António de Aguiar and the Waterfront Avenue: The Seaside City Idea in the Last Phase of the Portuguese Empire

This study analyzes the urban composition and design principles of Portuguese architect-planner Joo Antnio de Aguiar, focusing on his coastal urban plans and waterfront avenues in former Portuguese colonies in Africa and Asia. Using a comparative and interpretative methodology, the research examines how Aguiar adapted European seaside city models to non-European colonial contexts under the framework of the 1934 legislative reform.

Why it matters — It shifts the scholarly focus from administrative planning inventories to the physical form of colonial cities, establishing how European coastal design models were structurally modified when deployed in African and Asian territories.

Caveat: The analysis relies on historical and comparative interpretation of planning documents rather than quantitative spatial metrics or post-occupancy evaluations.

Urban Science · doi

Reinforcement Learning-Based Traffic Signal Control for IoT-Enabled Intersections

The researchers developed a Proximal Policy Optimization (PPO) reinforcement learning controller for adaptive traffic signal control at a single intersection in Kuwait. Evaluated in a simulation environment calibrated with real-world hourly traffic volume data, the controller reduced average vehicle delays by 46% compared to fixed-time control and by 34% compared to vehicle-actuated control. It also decreased per-vehicle CO2 emissions by approximately 23%, maintaining these performance improvements under demand fluctuations of up to 15% across both weekday and weekend patterns.

Why it matters — It demonstrates that localized, edge-intelligent reinforcement learning can significantly optimize traffic flow and reduce emissions at individual intersections using only local observations, without requiring centralized coordination or future demand forecasts.

Caveat: The controller's performance was evaluated within a simulation environment of a single intersection rather than a real-world physical deployment.

arXiv · pdf

Distance from home matters: Investigation of a basic movement strategy

The study models human mobility using a simulated agent on a two-dimensional lattice who selects trip destinations based on the ratio of the destination's distance from home to the current starting point's distance from home. This simple movement strategy, tested with both uniform and realistic distributions of demands and suppliers, qualitatively reproduces empirical mobility patterns such as the power-law distribution of traveled distances.

Why it matters — It demonstrates that complex, heavy-tailed human travel patterns can emerge from a basic spatial decision rule centered on home distance, without needing to assume complex cognitive maps or memory-heavy exploration algorithms.

Caveat: The model's findings are validated qualitatively against statistical features of human mobility rather than quantitatively against a specific real-world empirical dataset.

arXiv · pdf

Perceiving exposure segregation with open urban imagery

The researchers developed VISAGE, a multi-modal model framework that analyzes open satellite and street-level imagery to identify physical features of the built environment that drive socioeconomic exposure segregation. Testing across 10,030 communities in 31 U.S. cities, the model mapped physical features to mobility-derived segregation patterns with a Pearson correlation of 0.770. The analysis revealed that defensible architectural forms like fences and gated enclosures, alongside monofunctional zoning, predict higher social isolation, while mixed-use infrastructure and inclusionary housing policies correlate with higher social mixing.

Why it matters — It establishes a direct, interpretable link between the visual, physical design of neighborhoods and actual behavioral segregation, proving that policy interventions like inclusionary housing leave legible physical signatures that foster social diversity.

arXiv · pdf