The rapid urbanization of cities presents significant sustainability challenges, necessitating big data and digital tools as solutions for efficient resource management. A key advancement in this area is the Urban Digital Twin (UDT). UDTs aim to create dynamic virtual replicas of urban environments, enabling informed decision-making for city planners and policymakers. UDTs enable predictive modeling, resource optimization, and impact assessment of urban interventions. On the other hand, one of the globally accepted sustainable development goals (SDGs) to achieve by 2030 is SDG 11, which focuses specifically on Sustainable Cities and Communities. SDGs and SDG 11 consider the cities as a system that consists of the physical urban environment and social dynamics coming from governance, citizens and communities. However, current research on UDTs has primarily focused on technical aspects, leaving the potential of UDTs to support SDG 11 and its social dynamics underexplored. This study aims to understand whether UDTs can support the realization of SDG 11. Therefore, we explore how the capabilities of UDTs, such as monitoring, modelling and simulation, visualization, information provision and collection can support the SDG 11 principles of managing interconnected targets, inclusivity, multi-stakeholder collaboration, and monitoring of SDG 11 targets. We propose a socio-technical framework illustrating how UDTs can support SDG 11 and outline the key social and technical challenges to be addressed to fully realize UDTs’ potential. Finally, we discuss the conclusions and outlook for overcoming such challenges.

The current burgeoning of urban digital twin projects is leading to a proliferation, and sometimes confusion, about what exactly a digital twin is. Analyses tend to develop in two directions. First, efforts are being made to define typologies and debates are taking place on the twinness of digital twins’ projects: the specific state and qualities of something called a digital twin. Second, research has been carried out into the empirical analysis of the twinning process, not to judge what is or is not a 'twin', but rather how the process of creating something called a 'twin' actually takes place. We argue that these two directions are not sufficient to fully understand the current phenomenon of the development of digital twin projects and its implications for the transformation of urban governance in the digital age. We therefore introduce the notion of twinnation to capture the gradual process of broader transformations in the representation and governance of a city brought about by the development and use of digital twins. Using an illustrative empirical case, we show that the notion of twinnation allows us to grasp the role of a larger number of social and technical entities and raises new fundamental issues both for empirical analyses of digital twin projects and for reflections on the development of digital twins better adapted to their development and use contexts. This work should be an initial step toward fostering both transdisciplinary research and collaborations involving stakeholders in the development of data-driven infrastructures of urban governance.

An extension of the optimal transport problem is proposed, which includes a family of transformations incurring no transportation cost. This extension improves the co-registration among imagery datasets, where transformations such as rotations, displacements and changes of perspective, are a natural component of data acquisition. More generally, it provides a strategy for co-registration that blends the robustness of optimal transport with the interpretability of models.
The extended optimal transport problem pairs two distributions with minimal additional distortion, while identifying a cost-free, explainable component of the map.
A data-driven formulation is developed, as well as a methodology for its numerical solution. The latter complements gradient descent with a game-theory inspired approach, favoring collaborative moves between the cost-free and the unrestricted transformations.

The methodology is illustrated through its successful application to matching pairs of both synthetic and real images, which are
conceptualized as weighted samples from underlying distributions, and through the determination of treatment effects by co-registering treated and untreated populations in a synthetic example.

Encryption is crucial for smart city sensor data: It protects the confidentiality of the data collected by a vast array of sensors embedded throughout the city, such as those monitoring traffic, environmental conditions, utilities, and public safety. This data often includes sensitive information that could compromise individual privacy or city operations if accessed by unauthorized parties.

In addition, encryption can also ensure the integrity of the data, preventing unauthorized alterations that could lead to misinformation or manipulation of smart city systems. This is particularly important as tampering with sensor data could have severe consequencunles, such as disrupting traffic flow, causing malfunctions in public services, or hampering emergency response efforts.

Stream ciphers are ideal for encrypting smart city sensor data due to their ability to efficiently handle real-time data processing with minimal latency and computational overhead. These ciphers encrypt data on-the-fly, making them perfect for the continuous data streams generated by smart city sensors. Unlike block ciphers, stream ciphers do not require fixed-size blocks or data padding, allowing them to seamlessly handle diverse data sizes and formats typical in a smart city environment. Another often underestimated advantage of stream ciphers is that they – unlike block ciphers – do not require a mode of operation (which entails additional computational overhead and may introduce its own security issues). This flexibility enables security across a network of heterogeneous devices, safeguarding the confidentiality and integrity of information vital for applications ranging from traffic management to environmental monitoring.

Our contributions in this work are:

• We realised software-implementations of the recently published Lightweight Stream Cipher DRACO (Hamann et al., 2022) in two different programing languages (C++ and Rust), targeted at microcontrollers.
• Based on these implementations we conducted experiments on different types of resource-constrained Arduino/Raspberry nodes with random as well as real-word smart city sensor data, yielding numerous results.

To the best of our knowledge, our work presents the first performance analysis of the DRACO stream cipher implemented in Rust and C++ on microcontrollers.

We conducted a multitude of experiments on three microcontrollers: ESP32, ESP8266, and Raspberry Pi Pico. The ESP32 and ESP8266 are Wi-Fi-enabled microcontrollers often used in smart cities for applications like smart street lighting, air quality monitoring, and smart parking, while the Raspberry Pi Pico is usually used for local control tasks and sensor data acquisition, often in combination with other devices like the ESP32 for IoT solutions.

Our experiments measure the runtime performance of our two DRACO implementations (one implemented in C++, one implemented in Rust) on these resource-constrained devices. We benchmarked the performance of these implementations on the different microcontrollers for random data of various sizes as well as for more than 700 real-world smart city sensor measurements of various types (e.g. temperature, power units, volume flow rate) from cooling systems (valves, pumps, refrigeration units).

Nihonbashi, a commercial hub in Tokyo’s Chuo Ward, has a resident population of 170,000 and a daytime population of 650,000. Alongside other wards in metropolitan Tokyo, it confronts multiple urban risks, including floods, earthquakes, and heatwaves. According to Mainichi News (2019). 38% of evacuation centers across Tokyo’s 23 wards are located in severely flood-prone areas, particularly in the eastern low-lying regions. Chuo Ward’s proximity to Tokyo Bay heightens its susceptibility to these hazards. Tokyo has experienced a rising frequency of heatwaves, with 2022 marking the hottest summer since 1875, characterized by nine consecutive days of temperatures reaching 35°C. In 2024, 123 heat-related deaths were recorded. This vulnerability is intensified by the urban heat island effect, where anthropogenic heat in central Tokyo can exceed 400 W/m², a level classified as ‘extreme danger’ by Harlan et al. (2006). Tokyo’s Climate Adaptation Plan proposes heat countermeasures, including urban enhancements such as cool pavements, urban greening, and the establishment of cool spots; however, it lacks provisions for immediate responses to heatwave events. Official government guidance remains limited to recommendations for hydration and sheltering in place. In contrast, a technical report by the WHO Kobe Centre (2013) delineates three tiers of heat risk and corresponding responses, including the redistribution of vulnerable individuals to air-conditioned environments.

While adaptive and mitigative strategies are vital for managing the long-term impacts of urban risks, the immediate response behaviors of individuals in affected areas warrant equal attention. In Nihonbashi, the efficacy of evacuation plans is undermined by the significant proportion of evacuation centers situated in flood-prone zones, compounded by logistical challenges associated with navigating high-rise buildings and dense street networks. Previous research, such as Yamamoto and Li (2017), has employed the Ant Colony Optimization (ACO) algorithm to evaluate evacuation route safety in central Tokyo, focusing on Shibuya Ward near Nihonbashi. This study accounted for road blockage probabilities due to earthquake-induced building collapses and identified high-congestion routes to avoid, ultimately proposing safer alternatives.

The objective of this study is to assess the adequacy of Nihonbashi’s existing evacuation plans, encompassing both indoor movement within buildings and outdoor navigation to safe areas. With a focus on social vulnerability to urban risks, namely heatwaves, earthquakes, and flooding, the aim is to safeguard residents during emergencies. This study adopts a comprehensive, data-driven approach to evaluate and enhance evacuation strategies in Nihonbashi, integrating diverse data sources, advanced analytics, and predictive modeling tools. The research begins with a robust data collection process to capture the multifaceted nature of Nihonbashi’s urban environment. Spatial data, including building footprints, road networks, and open spaces, will be sourced from OpenStreetMap, Plateau and municipal records. These datasets provide the foundational geometry and topology needed to model the physical landscape accurately. Socioeconomic data, such as age distribution, income levels, and housing characteristics at the neighborhood level, will be gathered from census records and local government databases. This information is essential for understanding social vulnerability, as it reveals populations most at risk during emergencies (e.g., elderly residents or low-income households). Additionally, hazard maps detailing flood risks, earthquake vulnerabilities, and Urban Heat Island (UHI) effects will be incorporated from governmental and research institutions. These maps contextualize the environmental threats specific to Nihonbashi, enabling a risk-informed analysis.

The collected data will be analyzed using a combination of network analysis and spatial statistics to uncover patterns and vulnerabilities. For street network modeling, Python libraries such as NetworkX and OSMnx will be employed. NetworkX facilitates the assessment of connectivity by calculating metrics like node degree and betweenness centrality, identifying critical junctions in the road network. OSMnx complements this by extracting and analyzing street network topologies from OpenStreetMap, highlighting potential bottlenecks and high-traffic zones that could impede evacuation. For socioeconomic hotspot mapping, ArcGIS Pro is used to perform spatial autocorrelation analysis with Global Moran’s I, which measures whether vulnerability indicators (e.g., poverty, age) are clustered, dispersed, or randomly distributed across Nihonbashi. Additionally, Getis-Ord Gi* statistics identifies statistically significant clusters of high vulnerability, producing hotspot maps that pinpoint areas where physical risks (e.g., flooding) and social vulnerabilities (e.g., elderly populations) converge. These analyses will generate actionable insights into high-risk zones requiring prioritized attention.

The core of the methodology is the development of an urban digital twins of Nihonbashi using AnyLogic, a versatile simulation platform as a predictive model to inform decisions. This Nihonbashi digital twins will integrate spatial, socioeconomic, and hazard data into a dynamic, interactive model that replicates both indoor (e.g., building interiors) and outdoor (e.g., street networks) environments. The model will simulate various scenarios, including evacuation dynamics under heatwave, flood, and earthquake conditions, commuter patterns during peak hours, and UHI impacts on pedestrian movement. Key performance metrics, such as Required Safe Egress Time (RSET), will be tracked to assess evacuation plan effectiveness. AnyLogic’s multi-method simulation capabilities (e.g., agent-based and system dynamics modeling) enable the representation of individual behaviors (e.g., panicked movement) and systemic factors (e.g., traffic flow). Beyond simulation, the digital twins serves as an interactive visualization platform, supporting data representation, geospatial analytics, predictive modeling, and scenario-based decision-making. Stakeholders can use it to visualize risk scenarios, test infrastructure interventions (e.g., widened stairwells), and optimize evacuation routes.

The study’s outcomes include risk hotspot maps highlighting convergences of physical and social vulnerabilities, paired with recommendations for alternative evacuation routes. The urban digital twins will serve as a practical tool for stakeholders, enhancing Tokyo’s Climate Adaptation Plan by integrating immediate response strategies with long-term infrastructure solutions. This dual focus is critical for bolstering climate resilience in dense urban settings like Nihonbashi.

References

Harlan, S. L., Brazel, A. J., Prashad, L., Stefanov, W. L., & Larsen, L. (2006). Neighborhood microclimates and vulnerability to heat stress. Social Science & Medicine, 63(11), 2847–2863. https://doi.org/10.1016/j.socscimed.2006.07.030

Mainichi News. (2019). 38% of evacuation centers in central Tokyo in expected flood areas. https://mainichi.jp/english/articles/20190114/p2a/00m/0na/002000c

World Health Organization. (2013). Heat-health action plans: Lessons from the Western Pacific Region (Technical Report). WHO Kobe Centre. https://iris.who.int/handle/10665/208167

Yamamoto, K., & Li, X. (2017). Safety evaluation of evacuation routes in Central Tokyo assuming a large-scale evacuation in case of earthquake disasters. Journal of Risk and Financial Management, 10(3), 14.

Land use data, provided as open data by Japan’s National Land Numerical Information (NLNI), has long served as a fundamental resource across various fields such as urban planning and disaster prevention. The dataset divides the entire country into 100-meter square mesh units and classifies each unit according to its land use purpose, enabling spatially detailed understanding of land use patterns. However, maintaining this dataset requires significant time and cost, as human operators visually interpret land use for each mesh by overlaying satellite imagery with geospatial data. As a result, it is difficult to update the dataset rapidly on a nationwide scale, leading to insufficient responsiveness to changes in land use. To address this issue, this study developed a method to automate land use classification using satellite imagery and improve the efficiency of land use data maintenance. Specifically, the method utilizes high-resolution and high-frequency observation data from Sentinel-2 and employs machine learning to automatically classify land use into four categories: Residential land, Other inhabitable land, Water bodies, and Forests and wastelands. As a result, the method enables high-accuracy generation of land use data and achieves significantly improved operational efficiency compared to conventional approaches. Furthermore, by integrating time-series satellite imagery, the method shows potential for flexibly responding to changes in land use.

This study focuses on the growth of road networks in volunteered street view imagery (VSVI) data, using data from the Mapillary platform in Tokyo as a case study. The results demonstrate that the expansion of VSVI data is governed by two fundamental spatial processes, densification and exploration as observed for OpenStreetMap in previous studies. Furthermore, bivariate regression analyses indicated relationships between the number of image contributions and road coverage, as well as the growth rate of road networks for various road types. This study represents the first attempt to explore the links between the number of contributions and spatial distribution in VSVI, thereby providing new insights into its dynamic growth patterns and future development trends.

The integration of Building Information Modeling (BIM) and Geographic Information Systems (GIS) is increasingly critical for infrastructure planning, asset management, and smart city applications. While prior efforts have largely focused on buildings, the need for semantically rich, interoperable models for roads and transportation networks is becoming more prominent. This paper investigates the potential for unidirectional conversion between IFC 4.3-based road models and the Transportation model of CityGML 3.0. Through manual schema matching, the study classifies correspondences into full, partial, or no match categories and assesses the potential for semantic and geometric translation between the two domains. Findings reveal that while basic spatial and geometric entities can be mapped with partial fidelity, significant semantic details - such as alignment geometry, layered road structures, and road furniture - remain unmapped. The paper concludes by recommending the development of a unified road ontology or mediation layer to support lossless integration. These insights contribute to ongoing efforts in geoBIM interoperability, such as those led by buildingSMART International and the OGC's CityGML working groups.

Japan's geospatial information policy has evolved significantly since the 1995 Great Hanshin-Awaji Earthquake, marked by the enactment of the Basic Act on the Advancement of Utilization of Geospatial Information in 2007 and the recent "Project PLATEAU," a 3D urban model initiative launched in 2020. This paper analyses the 30-year history of Japan's geospatial information policy and focuses on the impact of PLATEAU on administration and industry.

The policy's journey, from the initial NSDI definition in 1999 to the contemporary PLATEAU Vision, showcases a shift from infrastructure establishment to broader social implementation. Key milestones include the 2007 Basic Act, which formalized geospatial information, and the Quasi-Zenith Satellite System (QZSS). PLATEAU, a nationwide 3D urban model project, stands out for its rapid data creation, open data approach, and diverse use case development across sectors like urban planning, disaster prevention, and mobility.

Analysis reveals a transition from infrastructure-centric policies to user-oriented strategies, with standardization efforts evolving from domestic rules to open standards like CityGML. PLATEAU's success stems from its "StandardsOps" methodology, emphasizing agile specification revisions and open community engagement. This approach, which balances open discussions with strict description rules, has fostered a dynamic standardization ecosystem.

PLATEAU's impact extends beyond data standardization, influencing business model innovation and industry productivity. Its adoption of open standards and agile methods sets a precedent for future geospatial information policies in Japan and globally, demonstrating the potential for rapid innovation through collaborative standardization.

Incorporating detailed 3D building data into national spatial data infrastructures (SDI) is associated with numerous technical and administrative challenges. This paper presents and discusses the challenges that have emerged within recent 3D mapping initiatives in Slovenia. The increasing availability and affordability of LiDAR and photogrammetric technologies have enabled the automated generation of comprehensive 3D building datasets. Despite these technological advancements, there are several difficulties in integrating these datasets with existing cadastral and topographic databases. Primary challenges include discrepancies in spatial accuracy, outdated building outlines, semantic inconsistencies, and varying classification methodologies among datasets. Moreover, maintaining these integrated datasets is complex due to differing update cycles across cadastral and topographic systems. Cadastral data requires rigorous administrative processes for updates, while 3D modelling of buildings typically relies on automated procedures. Additionally, the official status of cadastral data further complicates the integration. The challenges identified in the case of Slovenia can largely be generalized to systems in other countries, highlighting the necessity for strategic planning in data integration processes, considering both technical specifications and administrative frameworks, to use the full potential of 3D building data within national SDIs.

In this research, we propose a methodology to improve the performance of scan matching and point cloud segmentation for 3D mapping of urban river environments. We also focus on the integration of depth image-based scan matching and spatial segmentation using streaming LiDAR data embedded in GNSS/LiDAR-SLAM. Moreover, we conduct experiments using a waterborne mobile mapping system to verify that our methodology can improve the stability and scalability of point cloud processing and achieve high-speed processing even in measured environments that cause SLAM degeneration problems. In addition, we propose a fast object classification based on rule-based segmentation using streaming point clouds.

Forest inventories are essential for sustainable forest management and health monitoring. Image-based surveys using Unmanned Aerial Vehicles (UAVs) are increasingly adopted for this purpose. DeepForest, a deep learning model pre-trained on large annotated datasets, enables scalable and cost-efficient tree detection in the resulting imagery. While it is known that fine-tuning DeepForest to specific target sites improves its performance, optimal fine-tuning strategies remain unclear.
In this study, we investigate the fine-tuning of DeepForest on a small dataset from a temperate mixed forest. We examine the influence of four parameters on model performance: (1) the method used to label training data (manual labeling, manual correction of DeepForest predictions, and automatic labeling using 3D point cloud segmentation), (2) input image resolution, (3) the number of fine-tuning epochs, and (4) the size of the training dataset.
Our results show that the out-of-the-box DeepForest model performs poorly on the target dataset but improves significantly with fine-tuning, even when using limited data and few training epochs. Manual and automatic labeling achieve similarly high performance, both clearly outperforming the manual correction approach. While the model initially performs better on lower-resolution images, this difference diminishes after fine-tuning on higher-resolution inputs. We observe no clear benefit from increasing the training set size, possibly due to the low diversity of our test dataset.
Overall, our findings confirm that DeepForest is well-suited as a base model for tree detection in high-resolution UAV imagery and can be adapted to new environments with low effort.

This paper presents the development of a Web API that enables querying PLATEAU 3D city model attributes using Japan's Spatial ID referencing system. We implemented a three-tiered API architecture that provides flexible access to city model data for 250+ Japanese cities with median response times of 35ms. An interactive interface allows users to visually select and explore Spatial IDs. Our implementation revealed that the current Spatial ID specification's geoid-based vertical reference creates integration challenges with common 3D visualization platforms and positioning systems. We propose improvements including adopting ellipsoidal referencing or a more fundamental geocentric approach. This work advances Japan's digital-twin ecosystem by providing a standardized mechanism for integrating diverse urban datasets through universal spatial referencing.