LM-based agents excel when given high-level action APIs but struggle to ground language into low-level control. Prior work has LLMs generate skills or reward functions for RL, but these one-shot approaches lack feedback to correct specification errors. We introduce SCALAR, a bidirectional framework coupling LLM planning with RL through a learned skill library. The LLM proposes skills with preconditions and effects; RL trains policies for each skill and feeds back execution results to iteratively refine specifications, improving robustness to initial errors. Pivotal Trajectory Analysis corrects LLM priors by analyzing RL trajectories; Frontier Checkpointing optionally saves environment states at skill boundaries to improve sample efficiency. On Craftax, SCALAR achieves 88.2% diamond collection, a 1.9x improvement over the best baseline, and reaches the Gnomish Mines 9.1% of the time where prior methods fail entirely.
The 15-minute city is a powerful planning concept to counter car-dependence by promoting active mobility to amenities and fostering inclusive urban environments. However, this policy has challenges in amenity-poor urban peripheries. Public transport remains underexplored in this discourse despite its role in distant access. Here, we propose a framework that incorporates public transport into the 15-minute city model using openly available data. By comparing Helsinki, Madrid, and Budapest, we demonstrate that multimodal mobility substantially increases access to amenities and enhances socio-spatial integration within a 15-minute reach. Although urban periphery benefit significantly from radial or high-speed public transport lines in their social mixing potential, such lines alone do not improve their access to amenities. These findings underscore the need to optimize polycentric public transport networks that can improve inclusive urban accessibility and complement active mobility in polycentric cities.
Abstract For city planners, public experimentation has become an attractive tool to “look into the future”, increasingly including novel technologies: Actors test novel mobility options, such as autonomous driving on urban roads to receive real-world feedback on their prototypes; and digital technologies are used to create virtual spaces of experimentation to explore interventions in urban space before implementation. Paying explicit attention to the performative character of experiments and the mechanisms by which they make envisioned futures more plausible than others, we build on the concept of “techniques of futuring” (ToF) to better understand the role of experiments in urban transformations. We ask: How do urban experiments perform mobility futures and how does the performance make these futures plausible? We provide empirical insights on two cases of experimental environments in Munich: a living lab for autonomous driving and an urban digital twin for novel bicycle infrastructure design. We identify three core performative mechanisms by which urban experimentation contributes to making certain futures plausible: picturing the vision, preparing the city, and persuading the public. These mechanisms show how experiments involving novel technologies can become powerful in underpinning the presented visions of future mobility. At the same time, they call for caution when the allure of these mechanisms outplays alternative ways of deliberating and creating mobility futures.
Aesthetics of cities. City planning and beautifying, Cities. Urban geography
Cities are some of the most intricate and advanced creations of humanity. Most objects in cities are perfectly synchronised to coordinate activities such as jobs, education, transportation, entertainment, and waste management. Although each city has its own characteristics, some commonalities can be observed across most cities, such as issues related to noise, pollution, segregation, and others. Further, some of these issues might be accentuated in larger or smaller cities. For example, with more people, a city might experience more competition for space, so rents would be higher. The urban scaling theory provides a framework for analysing cities in terms of their size. New data for analysing urban scaling theory allow for an understanding of how urban metrics change with population size, whether they apply across most regions, or whether patterns correspond only to some countries or regions. Yet, reducing a city and all its complexity to a single indicator might simplify urban areas to the extent that their disparities and variations are overlooked. Often, the differences in living conditions across different parts of the same city are greater than the degree of variation observed between cities. For example, in terms of rent or crime, within-city variations might be more significant than between cities. Here, we review some urban scaling principles and explore ways to analyse variations within the same city.
This study proposes the first demand-driven, multi-objective planning model for optimizing city-scale capacity allocation of EV charging infrastructure. The model employs a bottom-up approach to estimate charging demand differentiated by vehicle type-battery electric vehicles (BEVs), extended-range electric vehicles (EREVs), and plug-in hybrid electric vehicles (PHEVs). Chongqing, a rapidly expanding EV industry cluster in China with a strong industrial base, supportive policies, and diverse urban morphologies, is selected as the case study. The results show that (1) monthly EV electricity consumption in Chongqing rose from 18.9 gigawatt-hours (GWh) in June 2022 to 57.5 GWh in December 2024, with associated carbon emissions increasing from 9.9 kilotons of carbon dioxide (ktCO2) to 30 ktCO2; (2) 181,622 additional charging piles were installed between 2022 and 2024, with the fastest growth observed in Yubei, reflecting a demand-responsive strategy that prioritizes areas with higher population density, higher income levels, and adequate land availability for pile deployment, rather than broad geographic coverage; and (3) between 2025 and 2030, EV electricity demand is projected to reach 1940 GWh, with the number of charging piles exceeding 1.4 million, and charging demand from EREVs and PHEVs expected to overtake BEVs later in the period. While Chongqing serves as the pilot area, the proposed planning platform is adaptable for application in cities worldwide, enabling cross-regional comparisons under diverse socio-economic, geographic, and policy conditions. Overall, this work offers policymakers a versatile tool to support sustainable, cost-effective EV infrastructure deployment aligned with low-carbon electrification targets in the transportation sector.
Deploying spatio-temporal forecasting models across many cities is difficult: traffic networks differ in size and topology, data availability can vary by orders of magnitude, and new cities may provide only a short history of logs. Existing deep traffic models are typically trained per city and backbone, creating high maintenance cost and poor transfer to data-scarce cities. We ask whether a single, backbone-agnostic layer can condition on "which city this sequence comes from", improve accuracy in full- and low-data regimes, and support better cross-city adaptation with minimal code changes. We propose CityCond, a light-weight city-conditioned memory layer that augments existing spatio-temporal backbones. CityCond combines a city-ID encoder with an optional shared memory bank (CityMem). Given a city index and backbone hidden states, it produces city-conditioned features fused through gated residual connections. We attach CityCond to five representative backbones (GRU, TCN, Transformer, GNN, STGCN) and evaluate three regimes: full-data, low-data, and cross-city few-shot transfer on METR-LA and PEMS-BAY. We also run auxiliary experiments on SIND, a drone-based multi-agent trajectory dataset from a signalized intersection in Tianjin (we focus on pedestrian tracks). Across more than fourteen model variants and three random seeds, CityCond yields consistent improvements, with the largest gains for high-capacity backbones such as Transformers and STGCNs. CityMem reduces Transformer error by roughly one third in full-data settings and brings substantial gains in low-data and cross-city transfer. On SIND, simple city-ID conditioning modestly improves low-data LSTM performance. CityCond can therefore serve as a reusable design pattern for scalable, multi-city forecasting under realistic data constraints.
The religious topography of German cities diversifies in terms of both the social spaces of faith and the built presence of religions and denominations. This challenges established Christian congregations to preserve architectural places and Christian spaces while simultaneously advancing interreligious interaction with the city and society. This paper summarises and discusses insight from a recently completed research project. By synthesising quantitative and qualitative data, it analysed churches that have undergone architectural or functional changes over the last decades. Cases range from interior design changes to the abandonment and even demolition of buildings. We found a wide variety of approaches to balancing the spatial and social needs of congregations. The paper presents four cases of re-ordering parish functions, both spatially and architecturally. The communities all face the challenge of maintaining post-war structures on the one hand, and declining funding and participation in church service, on the other. The different solutions chosen allow for discussion of the role of parish centres beyond architectural questions alone, considering the broader picture of urban space and social networks.
Architectural drawing and design, Aesthetics of cities. City planning and beautifying
Spatial planning, recognized as a systematic policy instrument for regional development and governance, plays a crucial role in achieving carbon peak and carbon neutrality. This study establishes a framework for carbon sources/sinks estimation and carbon compensation optimization and conducts empirical research in a representative coal resource-based city. We analyzed the spatial–temporal distribution characteristics of net carbon emissions in Huaibei from 2006 to 2020 using a spatial correlation model and an improved Carnegie–Ames–Stanford approach (CASA). Then, we applied the normalized revealed comparative advantage (NRCA) index and the SOM-K-means clustering model to categorize the carbon pattern into payment, balance, and compensation areas. These areas were further integrated with the “Three-zones and Three-lines” to reclassify nine spatial partition optimization types. Finally, we proposed a targeted emission reduction and sink enhancement optimization scheme. We found that urban carbon emissions and carbon sinks exhibit a significant mismatch, with the net carbon emission intensity reaching 166.76–383.27 t·hm<sup>−2</sup> from 2006 to 2020, showing a rapid increase followed by stabilization. The high-value area, centered in Xiangshan District, exhibits a circularly decreasing spatial characteristic, gradually extending to the central city of Suixi County. In the optimized payment area, the level of the carbon emission contributive coefficient surpasses the ecological support coefficient (3.92 < ECC < 6.04, 2.09 < ESC < 3.58). The optimized space in the balance area type is primarily situated in mining subsidence areas, leading to a lower overall level (0.42 < ECC < 0.57, 0.49 < ESC < 1.13). The optimized space in the compensation area type (2.24 < ECC < 3.25, 4.59 < ESC < 5.69) requires economic or non-economic compensation from the payment area. The study combines the “Three-zones and Three-lines” with the results of carbon compensation to formulate an urban emission reduction and sink enhancement program, which not only helps to consolidate the theory of low-carbon cities but also effectively promotes the realization of the regional carbon peak goal.
Carlos Núñez-Molina, Juan Fernández-Olivares, Raúl Pérez
In this work we propose a planning and acting architecture endowed with a module which learns to select subgoals with Deep Q-Learning. This allows us to decrease the load of a planner when faced with scenarios with real-time restrictions. We have trained this architecture on a video game environment used as a standard test-bed for intelligent systems applications, testing it on different levels of the same game to evaluate its generalization abilities. We have measured the performance of our approach as more training data is made available, as well as compared it with both a state-of-the-art, classical planner and the standard Deep Q-Learning algorithm. The results obtained show our model performs better than the alternative methods considered, when both plan quality (plan length) and time requirements are taken into account. On the one hand, it is more sample-efficient than standard Deep Q-Learning, and it is able to generalize better across levels. On the other hand, it reduces problem-solving time when compared with a state-of-the-art automated planner, at the expense of obtaining plans with only 9% more actions.
The idea of smart cities (SCs) has gained substantial attention in recent years. The SC paradigm aims to improve citizens' quality of life and protect the city's environment. As we enter the age of next-generation SCs, it is important to explore all relevant aspects of the SC paradigm. In recent years, the advancement of Information and Communication Technologies (ICT) has produced a trend of supporting daily objects with smartness, targeting to make human life easier and more comfortable. The paradigm of SCs appears as a response to the purpose of building the city of the future with advanced features. SCs still face many challenges in their implementation, but increasingly more studies regarding SCs are implemented. Nowadays, different cities are employing SC features to enhance services or the residents quality of life. This work provides readers with useful and important information about Amman Smart City.
The grid plays a prominent role in architecture, aiding in space organization and influencing all aspects of planning, ranging from urban design to intricate building details. This paper posits that the grid receives heightened emphasis in Brutalism, particularly in constructivist Brutalism, where materials and construction are intentionally exposed. A question arises regarding the grid’s characteristics—despite its subtle appearance, the grid can sometimes be deceptive, ambiguous, and manipulative. The paper analyzes the merits and drawbacks of employing the grid in architecture, shedding light on its contributions to both structural and perceptual comprehensibility, as well as its role in increasing usefulness. To illustrate the application and perception of the grid, the paper examines two primary planning levels: urban planning and building design. The case studies focus on examples from New York City housing developments, specifically those constructed between the 1950s and the 1970s, and projects by architect I. M. Pei, which offer valuable insights into practical implementation. The paper concludes that while the grid can establish order, it may also engender an “uncanny” feeling.
Dalila García Hernández, Salvador Adame Martínez, Carlos Alberto Pérez Ramírez
et al.
La sociedad actual enfrenta una susceptibilidad que demerita la compleja construcción de la percepción del riesgo ante situaciones que, por su frecuencia, se normalizan. Estas se interiorizan hasta dejar de ser consideradas como negativas o peligrosas, ya que forman parte de la cotidianidad. Por ello, el objetivo de este artículo es analizar la transformación de la realidad desde los principios de la antropología, sugiriendo cómo la carga moral se presenta en la percepción que cada sistema social tiene sobre el riesgo. Esto se explica a través de la revisión y análisis derivados del mapeo sistemático disponible sobre la percepción de las inundaciones. El enfoque del estudio es analítico-reflexivo, a partir de la argumentación de los aspectos claves que inciden en la transformación del entorno. La aceptación o rechazo generado mediante el ejercicio de la percepción, independientemente del grado de vulnerabilidad que la sociedad ha concebido. Este análisis se centra en el contexto del sureste mexicano, donde se destaca cómo las inundaciones pueden generar pérdidas socioeconómicas de alto impacto. Desde la antropología, se logra profundizar en el argumento de la dinámica social real mediante, ejercicios analíticos. Esto se plantea a partir de la necesidad de responder a la pregunta: ¿Cómo se desarrolla la percepción del riesgo en un escenario de vulnerabilidad real, vinculándola con la incidencia de la realidad empírica?
Aesthetics of cities. City planning and beautifying, Urban groups. The city. Urban sociology
Does the national innovation city and smart city pilot policy, as an important institutional design to promote the transformation of old and new dynamics, have an important impact on the digital economy? What are the intrinsic mechanisms? Based on the theoretical analysis of whether smart city and national innovation city policies promote urban digital economy, this paper constructs a multi-temporal double difference model based on a quasi-natural experiment with urban dual pilot policies and systematically investigates the impact of dual pilot policies on the development of digital economy. It is found that both smart cities and national innovation cities can promote the development of digital economy, while there is a synergistic effect between the policies. The mechanism test shows that the smart city construction and national innovation city construction mainly affect the digital economy through talent agglomeration effect, technology agglomeration effect and financial agglomeration effect.
The recent surge in interest in city layout generation underscores its significance in urban planning and smart city development. The task involves procedurally or automatically generating spatial arrangements for urban elements such as roads, buildings, water, and vegetation. Previous methods, whether procedural modeling or deep learning-based approaches like VAEs and GANs, rely on complex priors, expert guidance, or initial layouts, and often lack diversity and interactivity. In this paper, we present CityGen, an end-to-end framework for infinite, diverse, and controllable city layout generation. Our framework introduces an infinite expansion module to extend local layouts to city-scale layouts and a multi-scale refinement module to upsample and refine them. We also designed a user-friendly control scheme, allowing users to guide generation through simple sketching. Additionally, we convert the 2D layout to 3D by synthesizing a height field, facilitating downstream applications. Extensive experiments demonstrate CityGen's state-of-the-art performance across various metrics, making it suitable for a wide range of downstream applications.
Antika Fardilla, Rifta Septiavi, Ratna Juwita T
et al.
Land use change is an important issue for urban and regional planners and policy makers, but it is also very useful in conservation planning, food security, and hydrological modeling. Data, information and analysis tools become obstacles in detecting changes in land use. With increasing access to data and current technology, it is hoped that land use observations can be carried out in a simple way but have more accurate results. This study aimed to analyze land cover changes in Padang City 2018-2022, using Landsat Imagery and Geographic Information System (GIS) analysis. Firstly, observations on ESRI Land Cover which displays a global map of land use or land cover (LULC) derived from ESA Sentinel-2 Imagery at a resolution of 10 m. The results showed that the area of forest cover has decreased and the built-up area has increased in the 2017-2018 and 2021-2022. Secondly, using the EO Browser, namely Sentinel-2, that was done in one to search for and compare images using high resolution at these sources, there were 19 land cover changes, such as increasing residential land use, while forest land allotment decreased.
Abstract Effective urban planning and management rely on accurate land cover mapping, which can be achieved through the combination of remote sensing data and machine learning algorithms. This study aimed to explore and demonstrate the potential benefits of integrating Sentinel-1 SAR and Sentinel-2 MSI satellite imagery for urban land cover classification in Gondar city, Ethiopia. Synthetic Aperture Radar (SAR) data from Sentinel-1A and Multispectral Instrument (MSI) data from Sentinel-2B for the year 2023 were utilized for this research work. Support Vector Machine (SVM) and Random Forest (RF) machine learning algorithms were utilized for the classification process. Google Earth Engine (GEE) was used for the processing, classification, and validation of the remote sensing data. The findings of the research provided valuable insights into the performance evaluation of the Support Vector Machine (SVM) and Random Forest (RF) algorithms for image classification using different datasets, namely Sentinel 2B Multispectral Instrument (MSI) and Sentinel 1A Synthetic Aperture Radar (SAR) data. When applied to the Sentinel 2B MSI dataset, both SVM and RF achieved an overall accuracy (OA) of 0.69, with a moderate level of agreement indicated by the Kappa score of 0.357. For the Sentinel 1A SAR data, SVM maintained the same OA of 0.69 but showed an improved Kappa score of 0.67, indicating its suitability for SAR image classification. In contrast, RF achieved a slightly lower OA of 0.66 with Sentinel 1A SAR data. However, when the datasets of Sentinel 2B MSI and Sentinel 1A SAR were combined, SVM achieved an impressive OA of 0.91 with a high Kappa score of 0.80, while RF achieved an OA of 0.81 with a Kappa score of 0.809. These findings highlight the potential of fusing satellite data from multiple sources to enhance the accuracy and effectiveness of image classification algorithms, making them valuable tools for various applications, including land use mapping and environmental monitoring.
Nicos Makris, Gholamreza Moghimi, Eric Godat
et al.
Motivated from the increasing need to develop a quantitative, science-based, predictive understanding of the dynamics and response of cities when subjected to hazards, in this paper we apply concepts from statistical mechanics and microrheology to develop mechanical analogs for cities with predictive capabilities. We envision a city to be a matrix where people (cell-phone users) are driven by the economy of the city and other associated incentives while using the collection of its infrastructure networks in a similar way that thermally driven Brownian probe particles are moving within a complex viscoelastic material. Mean-square displacements (ensemble averages) of thousands of cell-phone users are computed from GPS location data to establish the creep compliance and the resulting impulse response function of a city. The derivation of these time-response functions allows the synthesis of simple mechanical analogs that model satisfactorily the behavior of the city under normal conditions. Our study concentrates on predicting the response of cities to acute shocks (natural hazards that stress the entire urban area) that are approximated with a rectangular pulse with finite duration; and we show that the solid-like mechanical analogs for cities that we derived predict that cities revert immediately to their pre-event response suggesting that they are inherently resilient. Our findings are in remarkable good agreement with the recorded response of the Dallas metroplex following the February 2021 North American winter storm which happened at a time for which we have dependable GPS location data.