2026-10-27

7 arXiv categories· 96 journals· 3053 candidates — 9 worth your time· 100 without an open abstract

Headline

A national dataset of 6,551 pre-automobile downtown boundaries in the US

Today, together

tag shift
no tag ran above its 30-day average
shares references
no items share references today counted over the 9 papers published above
on the same shoulders
7 of the day's papers cite Travel and the Built Environment (Reid Ewing, Robert Cervero 2010) counted over the 101 of this day's 109 named papers whose reference lists we have — standing on the same older work is not the same as being about the same thing
first appearance
no institution here was new to the last 137 days none of them on a paper we published in the 137 days before this issue

Two counts, not one counted twice: every row here but “on the same shoulders” counts the papers published above, and that one counts every paper this day named — those and the ones under Also published today, whose abstracts we could not read.

An integrated evaluation framework assessed OpenStreetMap 3D building data quality by combining 3D geometric comparisons with machine learning predictions. Researchers created an open-access national dataset of 6,551 downtown business district polygons for pre-automobile U.S. cities and towns. Another project established a scalable, rule-based framework to classify anonymous truck GPS data into four distinct stop purposes across urban and non-urban settings.

codedatapublished: Environment and Planning B Urban Analytics and City Science

Andrew J. Van Leuven
Environment and Planning B Urban Analytics and City Science

University of Vermont

Regional Economics and Spatial Analysis · Urban Design and Spatial Analysis · Urbanization and City Planning · United States · University of Vermont

A national dataset of downtown business district boundaries for pre-automobile U.S. cities and towns

The authors created an open-access dataset of 6,551 downtown business district polygons covering pre-automobile cities and towns across 47 U.S. states. Using historical census data, modern population statistics, and Overture Maps Foundation point-of-interest data, an open-source R workflow uses kernel density estimation to delineate each place's primary commercial cluster. The resulting boundaries were validated against official Main Street program districts in two states.

Why it matters — Researchers studying small and mid-sized American towns now have a standardized, nationally consistent set of spatial boundaries to evaluate downtown revitalization and commercial activity without needing to draw local borders manually.

Caveat: Boundary validation was evaluated in only two states against Main Street districts, which may not capture commercial core patterns nationwide.

Environment and Planning B Urban Analytics and City Science · doi

codedatapublished: Transactions in GIS

Longtao Hu, Heshan Du
Transactions in GIS

University of Nottingham Ningbo China

Geographic Information Systems Studies · Automated Road and Building Extraction · 3D Modeling in Geospatial Applications · machine learning · openstreetmap · Beijing

Assessing the Quality of OSM 3D Building Data Using 3D Geometric Comparison and Machine Learning

An integrated evaluation framework was developed to assess OpenStreetMap 3D building data quality by pairing 3D geometric comparisons with machine learning predictions. Tested on over 115,000 buildings across six cities in Asia and Europe—Beijing, Ningbo, Osaka, Yokohama, Amsterdam, and Almere—the model uses footprint Intersection over Union and centroid distances to pair crowd-sourced structures with official reference datasets, evaluating spatial precision through height, location, surface area, and 3D Hausdorff distance metrics.

Why it matters — This framework moves beyond 2D footprint analyses to evaluate actual 3D building characteristics like facade geometry and height, allowing researchers to flag quality issues across incomplete mapping datasets where reference geometry is missing.

Transactions in GIS · doi · code

datapublished: Transportation Research Record Journal of the Transportation Research Board

Mehdi Zolali, Sarah V. Hernandez
Transportation Research Record Journal of the Transportation Research Board

University of Arkansas at Fayetteville

Urban and Freight Transport Logistics · Vehicle Routing Optimization Methods · Logistics and Transportation Systems · University of Arkansas at Fayetteville

Rule-Based Technique for Classifying Stop Purposes Using Anonymous Truck GPS Data Across Urban and Non-Urban Settings

Researchers created a scalable, rule-based framework that classifies anonymous truck GPS data into four stop categories: pickup/delivery, short rest, long rest, and staging. Evaluated through sensitivity analysis, the optimized model achieved an overall accuracy of 85%, performing at 94.5% for short rests, 89.4% in non-urban settings, and 81.1% in urban areas.

Why it matters — Public agencies can now reconstruct freight stop purposes from vendor location streams without needing proprietary algorithms or external contextual datasets.

Caveat: The classification rules struggle to reliably distinguish short rest stops from pickup and delivery activities due to overlapping duration and location profiles.

Transportation Research Record Journal of the Transportation Research Board · doi

published: Nature Cities

Kangning Huang, Mingzhen Lu
Nature Cities · was arXiv

New York University Shanghai

Urban Design and Spatial Analysis · Regional Economics and Spatial Analysis · Urban Planning and Governance · New York University Shanghai · New York University

Nested economies of scale in global city mass

An analysis of over 3,000 cities globally reveals that built environmental mass scales sublinearly with population both across cities (beta = 0.90) and across neighborhoods within a given city (delta = 0.75). The mathematical framework shows that city-level scaling efficiencies directly originate from the internal, uneven spatial distribution of residents.

Why it matters — This challenges the widespread assumption that building material requirements scale linearly with population growth, demonstrating that larger urban areas and denser neighborhoods naturally achieve higher material efficiency per capita.

Nature Cities · doi

published: GeoJournal

Muneera AlYahya, Mark Daniel, Saad AlSharrah
GeoJournal

Dasman Diabetes Institute

Urban and Rural Development Challenges · Urban Transport and Accessibility · Public Health Policies and Education · Dasman Diabetes Institute · Deakin University

Public health and rapid modernisation in Kuwait: a cautionary tale

This study traces the history of Kuwait's post-oil urban growth to analyze how imported planning frameworks and policy deficits reshaped the built environment. It documents how rapid modernization eradicated major infectious diseases while simultaneously fostering high rates of non-communicable conditions like obesity, diabetes, and cancers, alongside persistent infectious risks for marginalized populations.

Why it matters — It demonstrates how uncritically applying Western urban development paradigms to Middle Eastern contexts can degrade public health and worsen social health equities.

Caveat: The analysis relies on a historical and policy overview rather than a quantitative health or spatial dataset.

GeoJournal · doi

published: Computational Urban Science

Dong Hong
Computational Urban Science

Traffic Prediction and Management Techniques · Traffic control and management · Transportation Planning and Optimization

Smart city traffic flow and congestion prediction model based on spatio-temporal graph neural network

A spatio-temporal graph convolutional network combining graph convolutions with a gated recurrent unit was built to forecast urban traffic flow, paired with an entropy-weighted fuzzy C-means clustering algorithm to predict congestion. The flow forecasting component achieved a mean square error of 1.176 and 97.99% accuracy, while the congestion prediction component attained an F1 score of 98.56% and a recall rate of 97.88%.

Why it matters — Integrating entropy weighting into fuzzy clustering directly alongside graph-based spatial-temporal modeling raises baseline precision and recall for joint flow and congestion classification across smart city road networks.

Caveat: The abstract does not report the location, spatial resolution, or temporal duration of the evaluation dataset.

Computational Urban Science · pdf · doi

datapreprint

Xiaofei Yu, Kailai Sun, Yanyu Mao, et al.
arXiv cs.AI

AgenticCity-Bench: Longitudinal Multimodal Observations of Urban Dynamics with Agentic Reasoning

AgenticCity-Bench is a longitudinal, multi-layered urban benchmark that spans 16 cities across 12 countries and six continents between 2000 and 2025. It aligns 18 distinct data modalities covering environmental conditions, human activity, and urban function to a shared calendar index and spatial reference. The platform evaluates cross-modal translation alongside agentic reasoning tasks including urban resource analysis, planning, and forecasting.

Why it matters — Researchers can now evaluate AI agents on multi-decade, heterogeneous urban decision-making and cross-modal sensing within a single, unified spatial-temporal structure.

arXiv · pdf

preprint

Leandros Maglaras, George Boustras, Azidine Guezzaz, et al.
arXiv cs.NI

SUMO

Routing Around the Queue: An Evaluation of Eco-Routing Protocol on different Traffic Scenarios

The authors evaluated ErouVe, an RSU-based vehicular eco-routing protocol, using a Veins and SUMO simulation framework on a road network featuring a short signalized route and a longer signal-free bypass. Under light traffic, the protocol had minimal impact, but under dense traffic with active queuing, it achieved medium fleet-wide CO2 reductions and higher individual fuel savings for diverted vehicles. Based on these failure modes, the study designed three corrective mechanisms: a knowledge-gated cost rule, a diversion-locking rule, and a distance-normalized CO2 cost metric.

Why it matters — It demonstrates that crowdsourced eco-routing only yields material emissions reductions under congested queuing conditions and provides concrete algorithmic fixes for common failure modes like decision flapping and metric bias toward idling.

Caveat: The protocol evaluation was conducted on a synthetic road network with only two distinct route choices.

arXiv · pdf

preprint

Christian Mulomba Mukendi
arXiv cs.LG

Resolving the Mobility-Equity-Emissions Trilemma: A Causal Machine Learning Framework for Dynamic Tolling Policy

The authors developed a Double Machine Learning framework to evaluate dynamic congestion tolling across different urban zone types using a synthetic network. Dynamic tolls applied to commercial zones produced nearly double the delay reductions observed in residential zones. Implementing an adaptive quota mechanism increased residential toll coverage 34-fold while retaining 93% of overall delay savings and reducing emissions by 743.5 metric tons of CO2.

Why it matters — The framework identifies a stable toll price window that maintains effective delay and emission reductions despite fluctuating traffic demand, eliminating the need for continuous toll recalibration in real-world deployments.

Caveat: The findings are based entirely on synthetic traffic data rather than an empirical deployment on a physical road network.

arXiv · pdf · doi

Still cited

Bing Yu, Haoteng Yin (2018), Spatio-Temporal Graph Convolutional Networks: A Deep Learning Framework for Traffic Forecasting 1 of today's items cite it · 17 of 3384 in the archive stand on it

Also published today

These appeared today in journals we track. Their abstracts are not openly available, so we cannot summarise them.