Chaos Computer Club - recent events feed

## OpenStreetMap Sound Demo _by Taro Matsuzawa (@smellman)_ ## ArcGIS Living Atlas of the World: Open Data Content _by Deniz Karagulle_ Creative Commons Attribution 3.0 Unported https://creativecommons.org/licenses/by/3.0/ about this event: https://2025.stateofthemap.org/sessions/P3FWCW/

EUthMappers is an ERASMUS+ initiative that promotes STEAM education in secondary schools throughout the European Union, enhancing students' digital skills and fostering environmental civic engagement. The project includes three universities, students and teachers from five European schools, working for two years. The project has three main phases: development of training materials, local mapping projects and humanitarian mapping collaboration. This presentation outlines the project's implementation steps and showcases the remarkable results achieved by not only participating students but also organizers. EUthMappers is an ERASMUS+ initiative that promotes STEAM education in secondary schools throughout the European Union, enhancing students' digital skills and fostering environmental civic engagement. The project bridges theoretical learning with practical applications using open-source geospatial tools and collaborative mapping on the OpenStreetMap (OSM) platform. The idea is to establish an European mapping network similar to YouthMappers. This presentation outlines the project's implementation steps and showcases the remarkable results achieved by participating students. The project involves three universities (Politecnico di Milano, Universidad Politécnica de Madrid, and Presovska Univerzita V Presov) and five secondary schools located in Italy, Spain, Slovakia, Romania, and Portugal, engaging approximately 160 students in total. The management of the project is done by Euronike, which is an association expert in capacity building activities and EU policies. Until now, the project has been running for two years, and was implemented in three phases. 1. Development of training materials: a comprehensive training package on open geospatial tools and data analysis was written and made available in six languages (English, Italian, Spanish, Portuguese, Slovakian and Romanian). This training was first delivered to teachers, who then transferred the knowledge to their students. 2. Local mapping projects: under the guidance of their teachers and with university collaboration, students developed local mapping initiatives from ideation to implementation, focusing on data acquisition and visualization techniques. Across five European cities, pupils carried out mapping-focused projects that combined local engagement with practical outcomes. In Rovereto (Italy), students identified drainage channels on forest roads in the “Bosco della Città” as a central issue and mapped them using GPS and drone imagery. Since such features had never been mapped on OSM, they engaged with the OSM community to propose and share a new mapping methodology. The resulting maps and database now support both routine and extraordinary maintenance by the Rovereto Forestry Service, helping mitigate hydrogeological risks. In Madrid (Spain), pupils mapped key elements related to mobility, leisure spaces, and climate shelters in the Arganzuela neighborhood. Working in groups, they identified and analyzed public space features, and then proposed concrete improvements to enhance young people’s quality of life, which they presented at local events and forums. In Prešov (Slovakia), students focused on tree mapping in central city parks. After ecological training, they collected detailed data—such as species, trunk width, and height—using practical tools and mobile apps. Their results contributed to urban ecological databases and were showcased to the public during their school’s open-door day. In Lisbon (Portugal), pupils explored how graffiti and street art reflect local identity by mapping urban artworks around their school. Through documentation and critical analysis, they created a record of artistic interventions with social and historical value, culminating in a community event promoting dialogue on youth expression in public space. In Pitești (Romania), students diagnosed low public awareness of recycling facilities and mapped the location, accessibility, and types of waste accepted at collection points. Their interactive map aimed to make recycling easier and more visible, encouraging environmental responsibility across the city. These projects show how pupil-led mapping activities can generate innovative tools, influence local planning, and foster active citizenship. 3. Humanitarian mapping collaboration: as the final activity, students participated in a humanitarian mapping project issued by the United Nations Global Service Centre (UNGSC). There were two workshops about humanitarian mapping and Sustainable Development Goals (SDGs) in order to provide students basic knowledge about humanitarian mapping and their contribution to common goods. Students are then trained on humanitarian mapping simulation projects before actually participating in a real-world scenario proposed by UN Maps, a programme in UNGSC to enhance UN peacekeeping missions operational capabilities through open geospatial information. The training project on

This talk examines the dynamics of collective intelligence in humanitarian mapping projects coordinated through the HOT Tasking Manager, using a dataset of 746 projects and 312,289 tasks to evaluate participation, collaboration, and evidence of intelligent group behavior. Humanitarian mapping projects coordinated through the Humanitarian OpenStreetMap Team Tasking Manager (HOT-TM) represent a paradigmatic case of large-scale digital collaboration. Yet, while their practical utility in disaster response and preparedness is increasingly evident, the underlying collective dynamics that allow these efforts to function effectively remain under-explored. This talk builds on our recent article published in ACM Transactions on Computer-Human Interaction (TOCHI) [1] that presents a comprehensive analysis of HOT-TM projects through the lens of collective intelligence. Collective Intelligence is defined as “groups of individuals acting collectively in ways that seem intelligent [2].” Following this definition, we structured our analysis around three guiding research questions: (RQ1) What characterizes the group of individuals collaborating in HOT-TM mapping projects?, (RQ2) How is collective action organized within these projects?, and (RQ3) What evidence of intelligent action can be identified in this setting? To answer these questions, we constructed and analyzed a dataset encompassing 746 HOT-TM projects executed between December 2021 and November 2023. The dataset includes 312,289 mapping tasks performed by 38,893 contributors, as well as detailed records of over 1.8 million task states. Additionally, we incorporated spatial information on the area of mapped buildings using data extracted from the OpenStreetMap database. Our analysis proceeds in three stages. First, we profile the mapping community. Results show that the vast majority of contributors are beginners, who typically participate in a single project. However, a small group of highly experienced mappers—classified by HOT-TM as "advanced"—contribute to dozens of projects and assume more complex tasks. Notably, only 29% of contributors declare their country, but among those who do, the majority are based outside the regions affected by the mapping projects (see Figure 1). This reinforces existing concerns about the limited presence of local knowledge in humanitarian Volunteered Geographic Information (VGI) initiatives [3]. Second, we investigate the organization of collective action using process mining techniques applied to task state logs. Most mapping tasks follow a simple trajectory: a task is mapped and then validated without being split or invalidated (see Figure 2). However, tasks that involve higher complexity or suffer errors require more contributors and longer processing times. Roles within the mapping system are clearly stratified: while beginners dominate mapping in simpler projects, advanced mappers take the lead in complex cases and are responsible for nearly all validations. Despite the potential for collaboration in the mapping phase—through the sequential editing of tasks—true interdependence among contributors is limited. Most tasks are executed by a single mapper, and where collaboration occurs, it is often sequential and uncoordinated. This suggests that HOT-TM's microtasking design promotes a form of "collection" rather than "collaboration" [4]. Third, we assess the presence of intelligent group behavior by analyzing task validation outcomes through logistic regression. We find that advanced contributors are significantly more likely to produce validated outputs, especially when working alone. However, involving multiple contributors in a task—especially when they are less experienced—decreases the probability of successful validation. Furthermore, tasks with larger building areas or those requiring extensive validation times are less likely to be validated, suggesting that complexity and ambiguity remain major challenges. These findings highlight a paradox at the heart of humanitarian mapping: although the system succeeds in rapidly mobilizing volunteers to produce useful geographic data, its collective intelligence is unevenly distributed and relies heavily on a small core of experienced contributors. The wisdom of the crowd is therefore not uniformly distributed; rather, it is the wisdom of a few that sustains the productivity and reliability of the system. Moreover, the absence of strong collaborative mechanisms and the limited engagement of local mappers constrain the potential for adaptive and context-aware mapping. We conclude by reflecting on possible design improvements for platforms like HOT-TM. These include: (1) enhancing onboarding and mentorship to accelerate the transition from beginner to advanced contributor; (2) incentivizing meaningful collaboration beyond sequential task handovers; and (3) integrating local knowledge more effectively by prioritizing and rewarding contributions from

Mapping in iD is designed to be a welcoming experience that should require little to no required knowledge to get started. Additionally, the editor does also have some additional features up its sleave for more advanced mapping tasks. Regardless if you are a new, intermediate or advanced mapper: it can never hurt to know a trick or two that make your mapping more efficient! This talk will go through common mapping workflows in iD and how they can elevate your mapping experience. There will be sections dedicated for beginners, intermediate and advanced mappers. For starters, it will be shown how one can get up to speed most efficiently with iD through its built-in walkthough and other help functionality. For intermediate mappers, the talk will cover topics such as photo-mapping, efficient use of keyboard shortcuts, integrated quality assurance tools, useful browser extensions, etc. The talk concludes with advanced techniques such as adding custom background imagery or map data. The showcased mapping workflows shall give a good overview of the spectrum of different mapping tasks one might encounter as a mapper on an regular basis, and will also highlight tools other than iD that allow to dive even further into the respective topics. Creative Commons Attribution 3.0 Unported https://creativecommons.org/licenses/by/3.0/ about this event: https://2025.stateofthemap.org/sessions/H3NM7X/

Humanitarian OpenStreetMap Team proposed the fAIr project (https://fair.hotosm.org/) - an open-source AI-assisted mapping tool. This study describes our user testing organized to compare AI-assisted mapping of buildings in the fAIr tool and classic manual mapping of buildings in the JOSM editor without AI assistance. 26 participants took part in the experiment. Efficiency (number of buildings mapped per minute), effectiveness (proportion of buildings mapped correctly), and satisfaction (feedback from participants) were analyzed. Introduction Humanitarian organizations need maps for their activities. However, there is a lack of quality maps in many countries. The Humanitarian OpenStreetMap Team (HOT) is coordinating a global effort for humanitarian mapping, where volunteers create maps of remote locations. They use satellite imagery to search for roads and buildings and draw them into the OpenStreetMap. Despite all the efforts of volunteers, many regions remain insufficiently mapped (Herfort et al., 2021). Artificial intelligence (AI) is already used to analyze satellite imagery. There is thus an opportunity to use AI for humanitarian mapping as well. HOT proposed the fAIr project (https://fair.hotosm.org/) – an open-source AI-assisted mapping tool (HOT, 2025). In a presentation from SOTM 2024, Anna Zanchetta (Zanchetta, 2024) analyzed the performance of fAIr in detecting buildings in different conditions. Our group, located in the Department of Geography, Faculty of Science, Masaryk University, Czechia, is part of the humanitarian mapping community Missing Maps Czechia and Slovakia. We were in contact with the fAIr developers from HOT – Omran Najjar, Kshitij Sharma, and Anna Zanchetta. They asked us if we could organize user testing of the first version of fAIr. The goal was to compare AI-assisted mapping buildings with the fAIr tool and classic manual mapping buildings without AI assistance. The popular JOSM editor, which offers experienced tools for manual building mapping, was used for comparison. Methodology The experiment took place in an online environment during November 2024. The experiment took place in four different sessions, so each participant could choose the day and time that suited them best. Training datasets were prepared for 24 different localities around the globe. 26 participants took part in the experiment – 14 beginners and 12 experienced contributors. The participants were divided into groups A and B, so there were similar numbers of experienced contributors and beginners in both groups. Each participant was assigned one location for mapping in fAIr and one location for mapping in JOSM. Each session was structured as follows: -An explanation of the rules for correctly and incorrectly mapped buildings. -45 minutes: group A mapped in fAIr, group B mapped in JOSM -5 minutes: break -45 minutes: group B mapped in fAIr, group A mapped in JOSM -Participants filled in the feedback questionnaire. Validators evaluated the mapped buildings, and the results were analyzed. Efficiency (number of buildings mapped per minute), effectiveness (proportion of buildings mapped correctly), and satisfaction (feedback from participants) were analyzed. Results Mapping in JOSM (86.08 %) was significantly more accurate than the fAIr tool (78.22 %). Mapping in JOSM (4.23 buildings/min) was significantly faster compared to the fAIr tool (1.97 buildings/min). Results of experienced users and beginners were also compared. Beginners mapped faster in JOSM (2.89 buildings/min) than in fAIr (2.08 buildings/min), but the mapping quality was almost the same in JOSM (79.28 %) and fAIr (80.41 %). Experienced contributors mapped much faster (5.69 buildings/min) and with higher quality (93.45 %) in JOSM than in fAIr (1.84 buildings/min and 75.38 %). Interestingly, it means that beginners have better results of mapping quality (80.41%) and mapping speed in fAIr (2.08 buildings/min) than experienced contributors (75.38 % and 1.84 buildings/min). On the contrary, experienced contributors have much better results in mapping quality (93.45 %) and mapping speed (5.69 buildings/min) in JOSM than beginners (79.28 % and 2.89 buildings/min). We also analyzed the influence of the complexity of mapping locality, e.g., the density of buildings in the mapped site and the quality of the imagery. For easy localities, experienced contributors generally have higher mapped building counts, indicating better fAIr performance on simpler tasks. For complex localities, the number of buildings mapped by experienced contributors drops significantly, suggesting that complex localities pose a challenge even for experienced contributors. A higher number of errors was found in complex localities, where AI likely generated lower-quality building outline designs. Interestingly, for complex localities, beginners mapped more buildings than experienced contributors, which may suggest that they focus on quantity. This could indicate potential quality issues.

This talk explores the extrinsic quality of OpenStreetMap data in Brno by comparing the city’s most frequently mapped amenities against a custom, field-collected reference dataset. The findings highlight relatively high attribute accuracy in OSM but reveal gaps in feature completeness, with only about 34.94% of features matched with the reference dataset. OpenStreetMap is a notable example of a database created by volunteers. Due to its open approach to data collection, establishing trust in the data is essential. Three key factors must be considered to evaluate this trust: completeness, correctness, and positional accuracy. The most common method in recent years for assessing large datasets like OpenStreetMap is to examine intrinsic data quality [1-3], which relies on metadata. However, this approach does not allow for a thorough analysis of the mapped features, resulting in only a rough estimate of the data's trustworthiness. To provide a more detailed understanding of the data, extrinsic data quality is evaluated by comparing OpenStreetMap with a reference dataset. This method is effective for evaluating feature completeness. However, when assessing attribute accuracy, a similarly detailed dataset for comparison is often unavailable. In these instances, the evaluator must gather their own reference dataset, which can be expensive and time-consuming. Because of that, previous studies mainly focused on assessing attribute accuracy by utilizing intrinsic data quality. In our study, we focused on assessing the extrinsic data quality of the city of Brno in the Czech Republic. We wanted to know how much we can trust OpenStreetMap in our city and if there is some correlation between attribute accuracy and metadata of the features. A secondary objective was to determine how well ISO 19157 can be used to assess the attribute quality of the OpenStreetMap. We chose the city’s ten most mapped amenity features and gathered a reference dataset for these features. Since we knew what we would be evaluating, we have acquired a dataset perfect for evaluating OpenStreetMap. Thus, there was no need for major compromises in data evaluation. Over the course of several months, we traveled over 1,000 km on foot and gathered a few thousand reference features using the Locus GIS app. Each feature contained a list of evaluated attributes together with a photo of the object for further evaluation. Evaluated amenities were bench, waste_basket, recycling, restaurant, bicycle_parking, cafe, vending_machine, post_box, pub and fast_food – the most mapped node features in Brno. Our assessment of OpenStreetMap's attribute accuracy revealed generally positive results. Several attributes in our sample achieved 100% accuracy, particularly those with boolean values. However, the most significant issues arose with string attributes that lack defined value lists, such as opening_hours. Ultimately, the primary concern identified was the completeness of the data. We found that the completeness of feature occurrence is inadequate; we were only able to match 34.94% of all reference features with those in OpenStreetMap. This completeness varied significantly across evaluated amenities, with waste_baskets and benches being notably underrepresented. We also assessed the positional accuracy of the data. Each amenity was evaluated separately, revealing average positional errors ranging from 2.63 meters to 4.02 meters. The median error was found to be between 1.83 meters and 3.13 meters. Overall, OpenStreetMap appears to be a relatively accurate positional database for the city of Brno, despite some isolated deviations (outliers). When examining the relationship between attribute accuracy and feature metadata, we assumed that more users editing a feature would lead to more accurate data. This concept is known as the “many eyes principle” [4, 5]. However, the correlations between metadata (such as the number of contributors, versions, and days since the last edit) and attribute correctness are typically not statistically significant. As a result, no explicit dependency can be determined, and no clear patterns emerge from the statistically significant values. Our work also shows that evaluating OpenStreetMap using ISO 19157 can be problematic because this standard does not consider multiple correct values or the varying degrees of attribute correctness (“level of detail” of attributes). Additionally, automating the evaluation of specific attributes, such as opening_hours, is challenging since these attributes can contain different yet correct values. Furthermore, missing or incomplete documentation significantly impacts evaluation, as there should be clear rules indicating which values are correct or not. However, achieving this is difficult for projects that rely on the folksonomy principle, which encourages users to create new values that often lack documentation. OpenStreetMap offers a comprehensive database, but its data quality varies signific

This is a story about the last 7 years of OSMTracker and how, from an academic space in a Costa Rican public university, we managed to keep alive the app with computer engineering undergrad students. We want to share what we have learned: the challenges and contributions during this process, and how we plan to continue. OSMTracker is one of the oldies free software data capture tools in the OSM ecosystem. Its simplicity, low technical requirements, easy customization, and ability to use the exported data in various applications make it a valuable resource for mappers. The app became part of the methodological resources used by the Laboratorio Experimental de Computación y Comunidades (LabComún) for development of university extension projects together with communities like Erizo Juan Santamaría Informal settlements and Alajuela en Cleta urban cycling collective. Naturally, the usage of OSMTracker in this context showed us improvements needed in the app, and consequently, we contributed with code that were incorporated in the tool. In 2018, nguillaumin, the original developer of the app, transferred the maintenance OSMTracker to LabComún. Since then, LabComún has been managing the challenge of developing free software from a (global south) public university: scarcity of resources, bureaucratic difficulties, and how to align the process of developing software while offering a sustainable educational experience for students. The ongoing focus of development of OSMTracker highlights the importance of engagement of undergrad computer science students in projects that are related to territories and people. This offers a real perspective on how their contributions to software could improve the quality of life. From a technical and academic perspective, the opportunity to be part of a bigger free software and open data community is unique. Finally, we want to open discussion on how to enhance the sustainability of OSMTracker, involving other actors in collaborations and keeping the app useful for thematic mapping projects. Creative Commons Attribution 3.0 Unported https://creativecommons.org/licenses/by/3.0/ about this event: https://2025.stateofthemap.org/sessions/9HN9SX/

OSMlanduse is the first EU-wide 10 m land-use map integrating 3.2 million OpenStreetMap geometries with Sentinel-2 imagery through an open deep-learning workflow. Delivering CORINE-level thematic detail with finer spatial resolution, it achieves 89% accuracy, providinng wall-to-wall coverage while retaining OSM’s sub-metre detail where available. Released under open licences with reproducible scripts, it supports applications from climate modelling and biodiversity surveys to urban planning and policy monitoring. By uniting crowdsourced mapping and Earth observation, OSMlanduse demonstrates a scalable, transparent approach to producing reliable, high-resolution land-use information at continental scale. Spatially resolved information on present-day land use (LU) is fundamental for climate-mitigation tracking, food-system monitoring and spatial planning, yet Europe still relies on inventories such as CORINE Land Cover (CLC) that are updated quinquennially at 100 m and under-represent the urban micro-mosaic. Meanwhile, new 10 m remote-sensing maps excel at spectral land-cover separation but lack the thematic richness contributed by citizens through OpenStreetMap (OSM). We introduce OSMlanduse, the first European Union-wide LU map at 10 m resolution that fuses 3.2 million OSM geometries with Sentinel-2 multispectral composites by means of a completely open workflow. The product supplies CLC-level thematic detail while matching the spatial grain of Copernicus imagery. Our objective was to demonstrate that globally recurrent, freely licensed data streams can be combined through deep learning to overcome the spatial incompleteness of volunteered geographic information. We converted the March 2020 OSM planet snapshot into 13 CLC classes, directly labelling 61.8 % of EU28 territory. For the unlabeled remainder we trained country-specific residual convolutional neural networks (ResNets) on medoid composites of cloud-masked Sentinel-2 top-of-atmosphere reflectance, then mosaicked predictions and original labels into a seamless raster–vector hybrid. Accuracy was assessed with 4 616 stratified reference points interpreted independently on sub-metre Bing and Google imagery. The map attains 89 % overall accuracy (95 % CI ± 2 %); producer’s accuracies range from 77 % (shrub/herbaceous vegetation) to 99 % (water bodies), while user’s accuracies exceed 93 % for agricultural strata. Most confusion arises between spectrally similar urban greens and semi-natural grasslands, or between early construction sites and bare soil, reflecting both spectral ambiguity and occasional tag noise. OSMlanduse inherits sub-metre geometric detail wherever OSM mapping is dense—Dutch canal parcels, German allotment gardens, Romanian farmyards—yet guarantees wall-to-wall coverage through 10 m raster infill. GeoTIFF tiles, training rasters and the OSM-to-CLC translation table are released under CC-BY 4.0 (DOI: 10.11588/data/IUTCDN) and visualised at https://osmlanduse.org; GPL-3.0 scripts ensure full reproducibility. Three methodological insights emerge. First, the current density and thematic granularity of OSM LU tags suffice to train deep networks that generalise across divergent biogeographic regions without external annotation campaigns. Second, multi-temporal medoid compositing plus per-country modelling dampens atmospheric noise and phenological divergence, enabling continental consistency from uncalibrated top-of-atmosphere data. Third, open-science principles—public code, permissive licences, cloud execution—place high-resolution mapping within reach of resource-constrained institutions. The dataset opens new avenues for investigating LU dynamics, from crop-rotation detection and peri-urban sprawl quantification (SDG 11.3.1) to habitat-specific sampling frames for biodiversity surveys and downscaling of economic statistics. Moreover, its billions of labelled pixels address the chronic scarcity of public training corpora highlighted by recent computer-vision studies. Limitations persist. Crowdsourced tags remain temporally asynchronous; the March 2020 snapshot necessarily precedes pandemic-era peri-urban expansion. Spectral ambiguity endures in arid shrublands and gravel pits, and the 10 m grid misses features narrower than one pixel such as hedgerows. Nevertheless, by releasing not only the final map but all processing scripts under GPL-3.0 we invite replication, auditing and regional adaptation. Training was performed on the FAO SEPAL cloud using Nesterov-accelerated Adam, batch size 64 and early stopping on 20 % held-out OSM tiles per country. Managing the 1.2 TB Sentinel-2 archive through orbit-wise partitioning and raster caching allowed execution of the full continental workflow within 96 GPU-hours on commodity instances. Sampling 5 000 patches per class and tile balanced the abundant yet noisy training labels while preserving rare categories such as mines and wetlands. While earlier studies either harvested OSM tags to create fragme

This study presents a geospatial framework that combines NLP, machine learning, and GIScience to extract and georeference tweets related to the November 2016 Haifa wildfire, enabling near real-time insights into urban fire dynamics. Using OpenStreetMap and GeoNames to geocode over 16,000 tweets, the researchers demonstrated strong spatial and temporal alignment with official fire incident reports, highlighting social media’s potential as a supplementary data source for disaster response. The approach offers a scalable model for leveraging crowdsourced and user-generated data in emergency informatics, especially in data-scarce regions. The increasing frequency and severity of urban wildfires demand new sources of near real-time information to support emergency response and disaster risk reduction [1-4]. In this study, we present an application for extracting and georeferencing the spatiotemporal distribution of tweets associated with the November 2016 wildfire in Haifa, Israel. The unprecedented urban fire influenced densely populated neighbourhoods, caused extensive infrastructural damage, and led to the evacuation of thousands of residents [5]. However, because of the nature and extent of the fire that lasted nearly 3 days, complete and reliable information concerning the emergence and development of new fire locations at the sub-city scale was only partial. Accordingly, management and decision-making procedures were complicated and some of the cascading events along the occurrence of the catastrophe were hard to be detected and addressed. The purpose of the study was to analyse tweets as a potential source of near real-time information and examine to what degree Twitter can be used to assist decision-making during occurrences on urban catastrophes. The implemented research combined Natural Language Processing (NLP), Machine Learning (ML), and Geographic Information Science (GIScience) to filter, classify, and precisely geolocate tweets at the city, neighbourhood or street-level resolution. One of the main components of the established geospatial framework was OpenStreetMap (OSM, https://www.openstreetmap.org, accessed July 2019), used in conjunction with the GeoNames gazetteer (http://www.geonames.org/, accessed July 2019) to construct a comprehensive spatial reference corpus. This enabled the geocoding of both explicitly and implicitly localized tweets that lack GPS metadata—an essential challenge given that only 1–3% of the tweets are geotagged with reliable geographic coordinates [6, 7]. We have collected approximately 2.4 million tweets using keywords related to the wildfire (in the Hebrew, Arabic, and English languages) between November 24th –27th, 2016. After classification using topic modelling and RCNN (Recurrent Convolutional Neural Networks) [8], around 114,000 tweets were labelled as relevant to the event. Of these, only 31 tweets were geotagged with geographic coordinates which is obviously an insufficient number of observations to perform spatial analysis. To overcome this shortcoming, we implemented a text-based georeferencing approach leveraging gazetteer data extracted from the OpenStreetMap and GeoNames databases. Accordingly, we converted 18 OSM shapefiles into a unified point dataset containing a wide variety of geographic features—ranging from neighbourhoods and roads to public buildings and natural landmarks. This dataset was merged with a version of the GeoNames corpus to create a point-based localized gazetteer representing the Haifa metropolitan area and its environs. For the purpose of geocoding, each point was associated with its place name in Hebrew, Arabic, or English. Following, the geocoding pipeline consisted of the following key steps: (1) NLP techniques including tokenization, stemming, and stop-word removal to extract named entities and spatial references from the tweet’s metadata [9]; (2) FuzzyWuzzy-based string matching algorithm that computes the Levenshtein distances between strings extracted from the tweet tokens and the place names in our gazetteer [10]; and (3) matchings were assigned a confidence score, which allowed us to filter or weight the credibility and accuracy of the spatial data. For example, georeferenced tweets with scores above 90% were deemed highly reliable for spatial trend detection. The process yielded 16,672 georeferenced tweets distributed across 130 unique localities within Haifa and its close vicinity. Following, we conducted a spatiotemporal inspection by aggregating tweets into 5×5 km grid cells and 8-hour intervals—bins that were informed by the density of localities extracted from the OSM/GeoNames hybrid gazetteer. We used Esri ArcGIS Pro to generate Kernel Density Estimation (KDE) maps [11] and 3D visualizations of the tweets’ distribution. The results presented strong temporal (Figure 1) and spatial (Figure 2) correspondence between georeferenced tweets and the officially reported fire incidents by the Israel Fire and Rescue Services (IF

The world is not just made of streets, buildings, and zones; it is shaped by how people engage and interact with places in their everyday lives. This abstract presents a web-based geospatial tool that supports the mapping of these lived places and locales named PlaceCrafter. PlaceCrafter supports researchers in identifying platial regions: functional, human-centred areas that cross administrative and formal boundaries. The framework is built on OpenStreetMap, combining (near) real-time clustering, analysis, and statistical validation of these platial regions. PlaceCrafter supports researchers in exploring the subjective experiences of place through existing datasets and city structures. Contemporary urban analysis requires tools and analytical software that can not only capture the physical structure of the city, but also the dynamic and human-centred places that can emerge from everyday interactions. While these technologies, such as existing OpenStreetMap (OSM) [1] views and Geographic Information Systems (GIS) [2] are effective for spatial tasks, these abstractions fail to understand the notions of place when compared to space [3]. Recent work [4, 5] has sought to shift the focus towards platial information systems, tools which situate the human experience, subjective knowledge, and fuzzy representation as a comparison to existing spatial systems. These fuzzy, subjective, and personal representations attempt to model ‘place’ as a contrast to space, which may not align with traditional geometries or formal administrative zones. PlaceCrafter responds to this challenge by integrating a spatial-platial [6] approach to identifying regions that are functionally cohesive, representing dense and meaningful concentrations of specific points of interest (POI). The POIs are traditional representations of space within GIS [3], such as cafes, restaurants, museums, pathways, and places of worship. Rather than relying on the top-down designation of locations, PlaceCrafter supports users, analysts, and researchers curating clusters that represent how space is used as opposed to administratively divided. The approach aligns with recent calls in GIScience to shift from ‘space’ to ‘place’ in smart city analysis [7] and to continue building on work which has operationalised the sense of place in urban contexts [8]. PlaceCrafter is designed not just to analyse space, but to make its platial structure visible, explorable, and comprehensible to analysts, researchers, and planners. Figure 1 presents the web-based application, developed in TypeScript using React, Vite, Leaflet, Turf.js, and D3.js. The software uses the Overpass API [9] to retrieve (near, depending on number of POIs loaded) real-time user-filtered POI data. The datasets are organised into semantic categories based upon the existing OSM semantic structures [10]; these filters can be customised dependent on the analytical task. This functionality supports the broad purpose of the web-based application, which is to enable researchers and practitioners to understand city form and structure through a platial lens. PlaceCrafter is structured around four phases guided by the OSM filtering approach: the initial phase (1) focuses on filtering and selection of relevant OSM categories, building upon existing work, such as POI Pulse [11] which classifies regions using semantic signatures, and user behaviour to generate profiles of locations in the Los Angeles area and ClusterRadar [12] which supports comparative spatial clustering and parameter tuning through interactive visualisation to examine how clusters change temporally; the second phase (2) is where the fuzzy clustering approaches are applied interactively and include K-Means [13] for compact cluster formation, DBSCAN [14] for spatial structures, and hierarchical clustering for multi-level spatial structures. These clustering methods are applied to the filtered POI data to reveal platial regions. The penultimate phase (3) focuses on statistical validation, where each clustered region is evaluated using established spatial metrics. These evaluations include the nearest neighbour index to assess spatial clustering, silhouette scores [15] for understanding cluster coherence, and spatial autocorrelation is measured using a simplified Moran’s I statistic for insights into category-based dependency [16]; The final phase is (4) visualisation, which explores the concept of platial readability, where each region is presented not just spatially but semantically, with data supported by POI type, diversity score, and density metrics. Additionally, the platial visualisation techniques used support the emerging approaches to conveying ambiguity, overlap, and functional gradients [3, 17]. The visualisation subsystem is modular for a wide array of end-user requirements, supporting fuzzy spray can visualisation and region influence grids as presented in Figure 2, in addition to convex hulls, kernel density heatmaps, and region quality

Between September 2024 and August 2025, we conducted a comprehensive street-level survey of Milano, Italy, capturing approximately one million 360° panoramic images using a monopod-mounted camera setup. These images were uploaded to Mapillary, contributing to open-access urban geospatial data. This presentation shares practical insights into continuous data collection methods and analyzes urban characteristics discernible from the imagery, such as graffiti prevalence, urban greenery distribution, and the potential of these visuals as foundational data for 3D digital twin models. I will discuss the current capabilities and limitations of using crowdsourced street-level imagery for urban analysis and planning. Between September 2024 and August 2025, we undertook a comprehensive street-level survey of Milano, Italy, capturing approximately one million 360° panoramic images using a monopod-mounted camera setup. These images were uploaded to Mapillary, contributing to open-access urban geospatial data. This presentation shares practical insights into continuous data collection methods and analyzes urban characteristics discernible from the imagery, such as graffiti prevalence, urban greenery distribution, and the potential of these visuals as foundational data for 3D digital twin models. I will discuss the current capabilities and limitations of using crowdsourced street-level imagery for urban analysis and planning. Advancements in consumer-grade 360° cameras have significantly enhanced image resolution, with modern devices achieving up to 11K. These improvements, coupled with enhanced low-light performance, have expanded the temporal window for effective data collection beyond daylight hours. Mapillary’s object detection capabilities can identify over 150 object classes, including traffic signs, poles, and vegetation. However, challenges remain in detecting certain features like graffiti and nuanced vertical urban greenery, highlighting areas for future development. By analyzing the Milano dataset, we can assess the efficacy of current detection algorithms and identify gaps where manual annotation or algorithmic refinement is necessary. This analysis informs strategies for leveraging 360° imagery in urban planning, such as monitoring infrastructure conditions and informing greening initiatives. The session will conclude with a discussion on the future of crowdsourced 360° street-level imagery, exploring how community-driven data collection can support comprehensive urban analysis and planning efforts. Creative Commons Attribution 3.0 Unported https://creativecommons.org/licenses/by/3.0/ about this event: https://2025.stateofthemap.org/sessions/NRJEVM/

OpenStreetMap and games feel like they go hand-in-hand and that's more than just coincidental. Both OSM and gaming have the power to bring people together, foster community engagement, and provide unique experiences. In fact, the OSM wiki has a page for games built using OSM data (https://wiki.openstreetmap.org/wiki/Games) and in recent years, we've seen the increase in the use of tools such as MapRoulette and StreetComplete that gamify the experience of contributing to OSM. While the latter is a very interesting topic in itself, this talk will focus on the former—games that use, but are not necessarily intended to contribute, OSM data. In this talk, we will explore the world of OSM-based/OSM-adjacent games to try and identify various game categories/genres and uses of OSM such as in location-based games (e.g. PokemonGO), serious and realistic simulation games, educational and trivia games, other niche/bespoke games, as well as both digital and tangible/tactile experiences. Furthermore, we will try to investigate the benefits and drawbacks of using OSM in games and look into other open source "games/game resources/gaming communities" (such as those in the Open Source Tabletop/RPG genre) to uncover possible intersections and opportunities. Whether you're a beginner or experienced OSM contributor, a game developer, or just a fellow gamer, this talk aims to spark new ideas and inspire further discussions, activities, and developments around the intersection of OpenStreetMap and games. Beyond the use of gamification for improving the OSM contributor experience, I feel that there is an opportunity to re-examine and revisit the broader topic of games using OSM data especially since OSM offers a rich and vibrant data source that can serve as the foundation for unique gaming experiences. How is/was OSM data used in games? What works/worked? What doesn't/didn't? Where (or where else) can OSM and games intersect? A lot of games use maps, can we use OSM there too? What other games can we create with OSM (e.g. Tactics? TTRPG? Board Games?). These are just some questions I'd like to ask and hopefully answer in the presentation. Creative Commons Attribution 3.0 Unported https://creativecommons.org/licenses/by/3.0/ about this event: https://2025.stateofthemap.org/sessions/LPDJXY/

The Humanitarian OpenStreetMap Team (HOT) celebrates its 15th anniversary in 2025. This presentation will explore HOT's evolution from a small group of mappers responding to the Haiti earthquake to a global organization at the forefront of humanitarian mapping. We'll delve into key milestones, technological advancements, the growth of the HOT community, and the increasing role of local mappers in leading humanitarian responses. The talk will also address the challenges and opportunities for the next 15 years, including sustainability, technological innovation, and expanding the impact of open mapping in a changing world. This presentation will examine how HOT has transformed disaster response, development initiatives, and community empowerment through collaborative mapping. It will also look at how HOT has fostered a global community of mappers and the impact of open data on humanitarian efforts. Furthermore, the presentation will discuss the strategic directions HOT is taking to ensure its continued relevance and effectiveness in an ever-changing technological and global landscape, emphasizing the importance of partnerships, innovation, and inclusivity. Creative Commons Attribution 3.0 Unported https://creativecommons.org/licenses/by/3.0/ about this event: https://2025.stateofthemap.org/sessions/U9MSCX/

Closing session of All Systems Go! 2025 Licensed to the public under https://creativecommons.org/licenses/by/4.0/de/ about this event: https://cfp.all-systems-go.io/all-systems-go-2025/talk/DR8ELH/

The UAPI Boot Loader Specification defines conventions that let multiple operating systems and bootloaders share boot config files. So far, only systemd-boot implements it - and it’s UEFI-only by design. As a result, hybrid UEFI/BIOS images require maintaining (and keeping in sync) two sets of bootloader configs: one for systemd-boot, and one for a legacy bootloader such as syslinux. I set out to fix that by building a BIOS bootloader that uses the UAPI Boot Loader Specification - allowing both UEFI and legacy boot to use a single shared set of config files. This talk is about why that matters, how I built it, and what comes next. In this talk, I’ll cover: - What the UAPI boot spec is - Why you'd want to use legacy boot instead of EFI/systemd-boot - *spoiler: you don't! but you might have to* - How I implemented UAPI boot support for legacy BIOS - What about UKIs? - A live demo of the bootloader in action - The current state of the project and what’s next https://uapi-group.org/specifications/specs/boot_loader_specification https://github.com/nkraetzschmar/bootloader Licensed to the public under https://creativecommons.org/licenses/by/4.0/de/ about this event: https://cfp.all-systems-go.io/all-systems-go-2025/talk/ANC879/

I overview how OS management is done at Meta. We run millions of Linux servers and we have to make sure that OS gets updated on all of them in a given period of time. To do that we developed several products: MetalOS (Image based version of CentOS), Antlir (image builder) and Rolling OS Update (a service that keeps a set of DNF repos in sync with upstream repos and uses them to update OS ) Licensed to the public under https://creativecommons.org/licenses/by/4.0/de/ about this event: https://cfp.all-systems-go.io/all-systems-go-2025/talk/VNCDRL/

Updates are a critical piece of managing your fleet of devices. Nowadays, Yocto-based distributions can utilize layers for well-established update mechanisms. But, did you know that recent releases of Yocto already come with a simple update mechanism? Enter systemd-sysupdate: a mechanism capable of automatically discovering, downloading, and installing A/B-style updates. By combining it with tools like systemd-boot, we can turn it into a comprehensive alternative for common scenarios. In this talk, we will briefly introduce systemd-sysupdate, show how it can be integrated with your Yocto distribution, and share thoughts on how it can be improved further. Licensed to the public under https://creativecommons.org/licenses/by/4.0/de/ about this event: https://cfp.all-systems-go.io/all-systems-go-2025/talk/MU7JM8/

`test.thing` is a VM runner which targets guests using an API defined by systemd. It started after a conversation at devconf about turning `mkosi qemu` into a library. A quick intro. ~~composefs is an approach to image-mode systems without the disk images. Files are stored in a de-duplicated content-addressed storage with integrity guaranteed through fs-verity. The last year has seen an acceleration of development on composefs-rs, a pure Rust implementation of the ideas behind composefs. Our goal is unification of the storage of bootable system images (via bootc), application Flatpaks, and traditional OCI container environments, bringing deduplication and integrity guarantees to all three. An overview.~~ Licensed to the public under https://creativecommons.org/licenses/by/4.0/de/ about this event: https://cfp.all-systems-go.io/all-systems-go-2025/talk/MLTTHW/

With Bootable Containers (bootc), we can place the operating system files inside a standard OCI container. This lets users modify the content of the operating system using familiar container tools and the Containerfile pattern. They can then share those container images using container registries and sign them using cosign. Using composefs and fs-verity, we can link a UKI to a complete read only filesystem tree, guaranteeing that every system file is verified on load. We integrate this in bootc by creating a reliable way to turn container images into composefs filesystem trees, and then including the UKI in the container image. We will share the progress on the integration of UKI and composefs in bootc and how we are going to enable remote attestation for those systems using trustee, notably for Confidential Computing use cases. https://github.com/containers/composefs-rs https://github.com/bootc-dev/bootc https://github.com/confidential-containers/trustee Licensed to the public under https://creativecommons.org/licenses/by/4.0/de/ about this event: https://cfp.all-systems-go.io/all-systems-go-2025/talk/TNKPQS/

S3NS is a trusted cloud operator that self-hosts Google Cloud infrastructure in France, targeting the SecNumCloud certification, the most stringent Cloud certification framework. SecNumCloud includes strict legal and operational constraints. To manage these systems securely and reproducibly, we’ve built a family of dedicated administration terminals based on the image based philosophy. These terminals rely on NixOS semantics and draw from the ParticleOS ecosystem: systemd-repart, and dm-verity, ensuring atomic updates, full immutability of the Nix store, and verifiable integrity of the boot chain and runtime system (measured boot), while using remote attestations by TPM2 when connecting to production assets. We will present the purpose of these terminals and what needs they serve along with their high level characteristics: partition layouts, provisioning and connection flow to the production assets. This talk will show an application of many of the concepts that were presented in the NixOS ecosystem and in All Systems Go itself by the systemd community. Licensed to the public under https://creativecommons.org/licenses/by/4.0/de/ about this event: https://cfp.all-systems-go.io/all-systems-go-2025/talk/TBDBDA/

In last year's ASG!, bootc and bootable containers were introduced. In this talk, we'll go over what changed since last year, and how Fedora CoreOS and RHEL CoreOS are leveraging bootable containers to reduce maintenance and increase sharing. Licensed to the public under https://creativecommons.org/licenses/by/4.0/de/ about this event: https://cfp.all-systems-go.io/all-systems-go-2025/talk/87TFB7/

Introduce ue-rs, a fresh project that aims to be a drop-in reimplementation of update engine, written in Rust. The goal of ue-rs is to have a minimal, secure and robust implementation of update engine, required by A/B update mechanism of Flatcar Container Linux. Just like the existing update engine, it downloads OS update payloads from a Nebraska server, parses its Omaha protocol, verifies signatures, etc. This project, however, is different from the original update engine in the following aspects. First, it aims to be minimal, by reducing heavyweight legacies in the update engine. Moreover, written in Rust, it brings a huge advantage for security, especially memory safety, in contrast to the original update engine, which is written mainly in C++ and bash. Finally, in addition to traditional OS update payloads, it supports systemd-sysext OEM, which is supported by Flatcar. Licensed to the public under https://creativecommons.org/licenses/by/4.0/de/ about this event: https://cfp.all-systems-go.io/all-systems-go-2025/talk/JAC3DH/

In the Linux world there are several tools and technologies to encrypt data on a hard drive, most falling into one of two categories: block device encryption (like LUKS) or stacked filesystem encryption (like EncFs or gocryptfs). This presentation will introduce Dirlock, a new tool that belongs to a third category: native filesystem encryption, using the kernel's fscrypt API. Dirlock is currently being developed and its aim is to provide a flexible way to encrypt files, suitable for both user accounts and arbitrary directories, with full PAM integration, support for hardware-backed mechanisms such as FIDO2 or TPM and with a D-Bus API for easy management. Licensed to the public under https://creativecommons.org/licenses/by/4.0/de/ about this event: https://cfp.all-systems-go.io/all-systems-go-2025/talk/AAWNQT/

Traditional package updates using tools like RPM or Zypper can introduce risks, such as incomplete updates or accidentally breaking the running system. To overcome these challenges, we developed **container-snap**, a prototype plugin designed to deliver atomic OS updates—updates that are fully applied or rolled back without compromising the system's state. container-snap leverages OCI images as the source for updates and integrates seamlessly with openSUSE’s [tukit](https://github.com/openSUSE/transactional-update) to enable transactional OS updates. By utilizing Podman’s btrfs storage driver, it creates btrfs subvolumes directly from OCI images, allowing systems to boot from the OCI image. This approach empowers users to construct their own OS images using familiar container image-building tools, like Docker or [Buildah](https://buildah.io/). In this session, we’ll dive into: - The architecture and technical implementation of container-snap - Challenges encountered during development and how we resolved them - Key lessons learned along the way - A live demo showcasing container-snap in action Come and join this session to learn more about how to boot from an OCI image without bricking your system! Licensed to the public under https://creativecommons.org/licenses/by/4.0/de/ about this event: https://cfp.all-systems-go.io/all-systems-go-2025/talk/YTCMSG/

Last year, Lennart Poettering of the systemd fame, [gave a presentation](https://media.ccc.de/v/all-systems-go-2024-276-varlink-now-) at this very same conference, where he introduced Varlink, a modern yet simple IPC mechanism. He presented a case for Varlink, rather than [D-Bus](https://en.wikipedia.org/wiki/D-Bus) to be the future of Inter-process communication on Linux. As someone who works on D-Bus, I took upon myself to prove him wrong, only to find out that I achieved exactly the opposite. It didn't take long before I got convinced of his vision. Since I was largely responsible for giving the world [an easy to use D-Bus Rust library](https://crates.io/crates/zbus), I thought it's only fitting that I do the same for Varlink. This talk will be the story of the creation of such a library, the challenges I faced, where Varlink fits the Rust idioms really well and where it does not and how all of this affected the development and the API. Licensed to the public under https://creativecommons.org/licenses/by/4.0/de/ about this event: https://cfp.all-systems-go.io/all-systems-go-2025/talk/SYGBNH/