2026-09-02

7 arXiv categories· 96 journals· 2622 candidates — 6 worth your time· 114 without an open abstract

Headline

AI governance, Chicago mobility scenarios, LLM mobility predictions, and GPS calibration

Today, together

tag shift
no tag ran above its 30-day average
shares references
no items share references today counted over the 6 papers published above
on the same shoulders
6 of the day's papers cite Accessibility evaluation of land-use and transport strategies: review and research directions (Karst Geurs, Bert van Wee 2003); 4 cite Local Indicators of Spatial Association—LISA (Luc Anselin 1995) counted over the 110 of this day's 120 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
Ritsumeikan University; The State Bank of Vietnam; University of Louisville none of them on a paper we published in the 82 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.

"Beyond Provincial Averages" analyzes national upper secondary graduation examination scores across all 63 provinces in Vietnam from 2017 to 2024 to diagnose spatial educational inequality. "Assessing urban flood resilience" develops a suite of indices to evaluate flood exposure, sensitivity, adaptive capacity, and potential economic impacts in Melbourne. "Adaptive, ethical and responsible AI governance" conceptualizes smart city AI governance and develops a six-step framework integrating risk management, responsiveness, and ethics-by-design.

published: Discover Cities

Z. R. M. Abdullah Kaiser
Discover Cities

University of Louisville

Smart Cities and Technologies · Ethics and Social Impacts of AI · Digital Economy and Work Transformation · New York City · Singapore · University of Louisville

Adaptive, ethical and responsible AI governance for smart cities and nations

This study conceptualizes AI governance and develops a six-step adaptive, ethical, and responsible framework that integrates risk management, responsiveness, and ethics-by-design across the AI lifecycle. The framework distinguishes between the regulation of AI systems and decision-making by AI systems, identifying key cross-sector risks such as algorithmic bias, vendor dependence, and regulatory gaps. The authors illustratively apply this framework to analyze the decentralized governance structure of New York City and the centralized system of Singapore.

Why it matters — It provides a structured, adaptable framework for smart cities and nations to manage the evolving risks of public-sector AI deployment, offering concrete design principles to balance technological efficiency with public trust and accountability.

Caveat: The framework's real-world utility is demonstrated only through illustrative, high-level applications to two specific administrative contexts rather than empirical testing of its implementation outcomes.

Discover Cities · pdf · doi

published: Applied Spatial Analysis and Policy

Anh Ton Pham
Applied Spatial Analysis and Policy

Ritsumeikan University

School Choice and Performance · Intergenerational and Educational Inequality Studies · Spatial and Panel Data Analysis · Ritsumeikan University · The State Bank of Vietnam

Beyond Provincial Averages: A Spatial Diagnostic Framework for Educational Inequality in Vietnam’s National Upper Secondary Graduation Examination

The study analyzes national upper secondary graduation examination scores across all 63 provinces in Vietnam from 2017 to 2024 for Mathematics, Literature, and Foreign Language. Using Theil decomposition and a sequential spatial diagnostic framework, it reveals that provincial average scores account for only 6% to 18% of total score inequality, leaving 82% to 94% of the variation within provinces. A 3x3 Spatial Markov analysis further details subject-specific patterns, showing that Mathematics has the highest rank stability and spatial clustering, while Literature exhibits higher mobility but growing spatial differentiation.

Why it matters — It demonstrates that relying on provincial averages for educational monitoring hides the vast majority of student performance inequality, providing policymakers with a framework to diagnose sub-provincial disparities and spatial performance clusters.

Caveat: Some transition profiles estimated via the Spatial Markov analysis remain imprecise due to small conditional origin counts in specific neighborhood contexts.

Applied Spatial Analysis and Policy · pdf · doi

published: Urban Informatics

Prasan Lama, Qian Sun
Urban Informatics

RMIT University

Flood Risk Assessment and Management · Urban Stormwater Management Solutions · Infrastructure Resilience and Vulnerability Analysis · spatial analysis · Melbourne · RMIT University

Assessing urban flood resilience: a comprehensive framework with evidence from Melbourne

This study develops a Flood Resilience Index (FRI), an Integrated Flood Resilience Index (IFRI), and a Damage Index to evaluate flood exposure, sensitivity, adaptive capacity, and potential economic impacts. The framework was applied across 474 urban units in the Moonee Valley and Maribyrnong local government areas of Melbourne, combining hydrological modeling, spatial analysis, and socio-economic data. Multi-scale Geographically Weighted Regression (MGWR) was then used to analyze how the relationships between flood resilience and its driving factors vary across different locations.

Why it matters — It provides urban planners with a localized, spatially explicit tool to identify specific neighborhoods with low adaptive capacity and high physical vulnerability, moving beyond uniform city-wide assumptions to enable targeted infrastructure and policy interventions.

Caveat: The assessment is constrained by the scope of the available data and does not incorporate future climate or urbanization scenarios.

Urban Informatics · pdf · doi

preprint

Md Rakibul Alam, Omer Verbas, Taner Cokyasar, et al.
arXiv physics.soc-ph

agent-based model · Chicago

Modeling of Mobility and Energy Policies in an Agent-Based Framework: Case Studies for Chicago Region in 2050

The study simulated nine policy scenarios for the seven-county Chicago metropolitan region in 2050 using POLARIS, an agent-based modeling framework calibrated to 2019 conditions. The model co-simulates passenger demand, freight generation, traffic assignment, and transit operations to evaluate the combined impacts of electrification, road pricing, parking reform, and transit expansion. Key findings show that high electrification reduces total fuel mass by 68% but increases charging energy by 4 to 8 times, creating a peak power demand of nearly 4 GW in the urban core, while transit expansion increases ridership by 18% compared to the business-as-usual baseline.

Why it matters — It provides the first integrated, agent-based scenario framework for Chicago that jointly evaluates multiple transportation and energy interventions, allowing planners to see how specific policies like road pricing and freight management interact to affect grid infrastructure and vehicle miles traveled.

Caveat: The simulation relies on a model calibrated to 2019 conditions to project behaviors and infrastructure needs for the year 2050.

arXiv · pdf

preprint

Saad Mohammad Abrar, Eesha Kurella, Arnav Dadarya, et al.
arXiv cs.LG

machine learning · United States

Do LLMs Know Your Neighborhood? Auditing LLM Priors for Neighborhood-Level Mobility Prediction and Structural Alignment

This study evaluated zero-shot Large Language Models (LLMs) on predicting Census Block Group-level mobility outcomes across four U.S. metropolitan areas using anonymized Cuebiq trajectory data. The researchers compared LLM predictions against supervised baselines across point-level, trajectory-level, and temporal metrics, finding that supervised models achieved 0.580 average accuracy compared to 0.435 for the best-performing LLM. A directional alignment analysis revealed that LLMs rely on coarse, stable predictor-level priors that remain uniform across different cities and outcomes, including asymmetric treatment of protected-group predictors.

Why it matters — It establishes the performance boundaries and systematic biases of LLMs when used as zero-shot generators of aggregate neighborhood mobility, proving that their internal priors do not yet align structurally with empirical urban dynamics.

Caveat: The evaluation is restricted to four metropolitan areas in a single country, and the underlying mobility patterns are derived from a single proprietary mobile device dataset.

arXiv · pdf

preprint

Nicolò Alessandro Girardini, Unchitta Kan, Eduardo López, et al.
arXiv physics.soc-ph

Human Mobility and Location-Based Analysis · Opportunistic and Delay-Tolerant Networks · Impact of Light on Environment and Health · mobile phone data · census data · travel survey

Behavioral calibration of mobile-phone GPS data for population-representative analyses

The Behavioral Population (BePop) framework was developed to calibrate mobile phone GPS data by aligning both demographic and behavioral distributions using census data and time-use surveys. Tested across three U.S. metropolitan areas, the framework embeds mobility sequences into behavioral profiles and estimates person-level weights to correct daily activity patterns, including time allocation, activity transitions, and mobility motifs.

Why it matters — It addresses behavioral biases in GPS data that demographic-only calibration leaves uncorrected, preventing these discrepancies from distorting downstream mobility indicators and population-level inferences.

arXiv · pdf

Still cited

Luc Anselin (1995), Local Indicators of Spatial Association—LISA 1 of today's items cite it · 41 of 3240 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.