openstreetmap

The closing session of the State of the Map 2022 conference. about this event: https://2022.stateofthemap.org/sessions/LX3EGF/

This proposal expands an understanding of humanitarian mapping from an ethnographic perspective, seeking to understand the complex mechanics behind this confluence of humanitarianism, technology, and crowdsourced labor. It seeks to scaffold a notion of the “open source mapping supply chain”, situating both humanitarian mapping and OpenStreetMap itself within a larger ecosystem of commercial, humanitarian, open source, government, and other actors in developing geospatial-related technologies. This presentation presents a selection of a MA dissertation project, pursued over the course of more than 1.5 years of immersive fieldwork on OpenStreetMap. This presentation will focus on humanitarian mapping through qualitative study, seeking to expand an understanding of humanitarian mapping (particularly that which has emerged from mappers associated with the Humanitarian OpenStreetMap Team - also known as HOT) through the use of ethnographic tools, seeking to understand the complex mechanics behind this confluence of humanitarianism, technology, and crowdsourced labor, asking how and why people contribute to open-source platforms like OSM, and what role humanitarian mapping plays within the wider ecosystem of geospatial and mapping technologies. Ultimately however, it seeks to scaffold a notion of the “open source mapping supply chain”, situating both humanitarian mapping and OpenStreetMap itself within a larger ecosystem of commercial, humanitarian, open source, government, and other actors in developing geospatial-related technologies. Founded in the aftermath of the 2010 earthquake in Haiti, the Humanitarian OpenStreetMap Team (HOT) helps both globally remote and local in-person volunteers to identify roads, buildings, and other features on the OpenStreetMap (OSM) platform. Created as a “free, editable map of the world,” OSM has enabled the mass-creation of volunteered geographical information (VGI) on a scale that is now more accurate than proprietary maps in many places, particularly as “crisis-mapping” has emerged as a means to gather real-time data on areas that have been affected by natural disasters or socio-political conflicts. OSM has also become also a site of resistance, where local and indigenous communities have engaged in mapping projects to reclaim autonomy, agency, and space through the historically contested practice of (digital) mapping. For these reasons, such crowdsourced maps have increasingly been used by humanitarian organisations to facilitate aid and disaster relief, and as open training data for algorithms learning how to automatically detect features through Artificial Intelligence (AI). As a key partner of humanitarian, corporate, and local actors, and having mobilised over 200,000 volunteers since 2010, HOT lies at the crux of these ongoing entanglements and contestations, both within and around the field of OSM. Previous studies of crowdsourced geographical information and crisis-mapping have generally revolved around quantitative analyses of OSM’s data, focusing on the credibility of the data itself, the makeup of the communities that contribute to it, the effects of “event-centric” crowdsourcing, or “newcomer retention” in humanitarian mapping (Dittus et al., 2016a, 2016b, 2017; Haklay, 2010; Haworth et al., 2018; Sui et al., 2013). Alternatively, they have also focused on the “spatial knowledge”, “hacker political imaginary”, and gender composition of mappers themselves (Brandusescu & Sieber, 2018; McConchie, 2015; Stephens, 2013). Parallel studies of other volunteer-driven communities like “Wikipedians” have taken similar approaches, analysing “user-generated content” and the motivations behind them (Nov, 2007; Yang & Lai, 2010). Both hacking and free and open source software (F/OSS) have also been explored ethnographically (Coleman, 2012; Kelty, 2008). While automated detection of features on OpenStreetMap has only recently become an important topic of research, ongoing studies have primarily focused on the accuracy or credibility of this endeavour (Brovelli et al., 2017; Resor, 2016). While existing studies of digital communities have focused on the socialities they engender or labor they require, they tend to forget the bureaucratic apparatuses that have emerged to govern them, both implicitly and explicitly (Coleman, 2012; Kelty, 2008). Similarly, studies of humanitarianism have focused on the ethics they operationalize, or the technologies that are mobilized in turn, but often at the expense of engaging in the wider spectrum of social and economic life that they enable (Cross, 2013; Redfield, 2012, 2016a; Scott-Smith, 2013, 2016a, 2019; Ticktin, 2014a). While this project draws upon these overlapping strains of research, it seeks to push the debate in an ethnographic direction, scaffolded by theories of bureaucratic technology, political economy, and humanitarianism. This research draws from participation in over 40 online events over 1.5 years, including mapathons, conferences and online lectures with OSM mappers, as well as semi-structured interviews conducted with 27 key-informants, alongside watching more conference videos, and reading blogs, mailing list emails, Twitter exchanges, and other internet archives. While empirically influenced by studies of hacking and open source software, this work ultimately focuses on the mechanisms and means through which this “free and open map” is created, and ultimately the ways of seeing and doing that it enables (Coleman, 2012; Kelty, 2008). Ultimately, it was the “supply chains” heuristic that emerged as a means to understand and illustrate this process. Similar to how supply chains “link ostensibly independent entrepreneurs, making it possible for commodity processes to span the globe”, the OSM project relies upon a series interconnected processes that enable the creation of the world’s crowdsourced map in a process that is far more precarious, and much less secure than promotional material might have one think (Tsing, 2009). Similar to how the satellite, computer, and software industries converged to create the conditions that allowed for OSM’s creation, so do people – and their associated institutions create map data through an almost miraculous collision of circumstances, assured precedent, and training. The Humanitarian OpenStreetMap Team, which was the initial entry point into open source mapping through, made its name by optimizing the mapping value chain: that is, by making it easier to contribute to OSM. But it also extended outwards: contributing to OSM was enabled not only by the wider socio-economic forces that coalesced to produce the project in the first place, but also by a series of digital value chains – both past and present. By delineating this supply-chains approach, this study hopes to scaffold a mental model of humanitarian mapping and the OpenStreetMap more broadly, to be employed in future studies – both quantitative and qualitative. Practically, it hopes to provide a heuristic and application of ethnographic tools, and present questions and queries directly to the community more broadly. about this event: https://2022.stateofthemap.org/sessions/NWB9QF/

Wikimedia Italia, the Italian OpenStreetMap Local Chapter of the OSM Foundation, presents its activities, online infrastructure developed to support OpenStreetMap in Italy and the Italian community. The talk will share the experience, situations and factors that have influenced agreat collaboration with the local contributors and institutions during the last years. The presentation will go through different areas of the Local Chapter’s activities. The recently updated infrastructure, composed by the Tasking Manager and the OSM extracts for Italy. Those tools are available and used by the Italian OSM community. Moreover, the official new Italian OSM website, the OSM licences tracking process and other tools developed to support the community will be presented. Other experiences that will be shared are the collaborations with local institutions, with the scope of strengthening local communities and increase the data in OSM. The keys to success are the volunteer coordinators. They are a point of contact important to establishing collaborations with individuals and institutions throughout the national territory about this event: https://2022.stateofthemap.org/sessions/MRK3C8/

Geo-literacy provides skills to read, interpret and use geospatial information, where little evidence exists regarding the potential and capacity of new education programs in advancing these skills. We present a citizen science project held in 13 high schools in Israel, where the students practice participatory mapping with OpenStreetMap to map features relevant to the navigation of visually impaired pedestrians. We show that students improve their geospatial thinking and reasoning skills, including their self-esteem. We believe that this research contributes to various pedagogic and education levels, in terms of theoretical knowledge about the integration of innovative geo-literacy programs. The 21st Century dictates that people have a good spatial and geographic understanding and knowledge. Geo-literacy is aimed to provide skills to read, interpret and use geospatial information. This is achieved by acquiring critical spatial thinking, reasoning, and analysis, and presenting understanding of the world using geographical terms and spatial language. Recent years have led to the development of new geo-literacy education programs specifically designed to nurture and promote these skills. These education programs build on geographical education that promotes spatial thinking and active citizenship. Still, little evidence exists regarding the potential and capacity of these programs in advancing civic and geographic skills and knowledge in the 21st Century, and on its contribution to - and advancing of - the individual and the society. The aim of this research is to gain a better understanding of the development of geo-literacy in the framework of a citizen science project in high schools. The citizen science project implemented in several schools in Israel – landmarks for accessible space, advances scientific research that aims to make the urban environment more accessible for visually impaired pedestrians. The participating high school students practice participatory mapping with OpenStreetMap (OSM) to map features relevant to the navigation of visually impaired pedestrians. These map features are used for the automatic calculation of optimal walking routes. The project combines social involvement, learning through geographic information systems, and familiarity with the field of urban accessibility for visually impaired people. The project includes the following stage: 1. Pre-stage that includes a) the design of the modular learning environment, b) the organizational and pedagogical preparation of the project integration in schools, and c) questionnaires examining the current level of geographic literacy of the participating students and their perspective regarding the integration of citizen science in schools. 2. Intervention program that includes guest seminars (including YouTube videos), lectures and learning activities, exposure to the world of visually impaired people, and the need for accessible environments and learning activities in the field of geoinformation with emphasis on OSM, crowdsourcing, and participatory mapping. 3. Mapping missing data into OSM. This stage is carried out in the field with a designated app developed for this project. The app - “Mundi” - allows the mapping of specific geographic features (mapping elements) used for the calculation of accessible routes designed specifically for visually impaired pedestrians. The features include, among others, sidewalks, crossings, accessibility aids, and handrails. The app includes gamification and tasks to encourage the students to map the missing features in their area of residence. 4. Post-stage questionnaires aimed to investigate and analyze the development of spatial skills in the context of participation in this program, examining whether students’ level of geographic literacy improved and whether they gained new knowledge on urban accessibility and the navigation proficiencies of visually impaired pedestrians. This stage also included a quantitative analysis of the students’ contributions in terms of OSM mapping, among others, the number of map edits, type of mapped features, the spatial coverage and temporal extent of their mapping activity. The study was conducted in the last two years in 13 high schools, including 25 classes and 460 students. The intervention model was implemented for three months in each class. The participating students implemented this project within their Cyber Geography studies, enabling them to learn through various geographic information systems. In total, close to 10,000 OSM edits were made by the students, which included more than 3,000 new crossings (and attributed tags), 400 new sidewalks, and 7,000 new street objects and obstacles (e.g., bus stations, light poles, trees, gates, bicycle parking). Preliminary analysis showed that participation in the citizen science project increased the students’ geospatial thinking and reasoning. For example, according to the questionnaire variables, on a score scale of 0-100, the geospatial thinking score has increased from 31 to 56, while the spatial awareness score has increased from 34 to 73 (p < .001). The geographic skills knowledge has increased from 3 to 3.9 (scale of 1-5). Moreover, the students' self-esteem with respect to their knowledge and use of geographic skills has improved considerably. In addition, results show that the broad and in-depth intervention model increased the students' appreciation of the scientists' contribution to the project, the contribution of the program in general, and the satisfaction with their participation in the project. This is the first project to introduce the use of OSM-based learning to the study of geography in Israel. Based on its outcome and analysis, we believe that this research contributes to various pedagogic and education levels, in terms of theoretical knowledge about the integration of innovative geo-literacy programs. These promote the drawing of operational and applicable actions regarding the planning of future projects and serving stakeholders in academia and the education system in terms of integrating scientific projects that increase students’ involvement in science and society and promote geo-literacy. about this event: https://2022.stateofthemap.org/sessions/3YQRDX/

Follow the OSM journey of the Médecins Sans Frontières (MSF) from a mapathon in Berlin in 2014 to creating and contributing geodata for numerous MSF operations through the Missing Maps project. This talk will be about how MSF uses OpenStreetMap internally and how we contribute through remote and field mapping. We will also share the lessons learned and reflect on the biggest challenges for MSF in creating and using the OSM data. The OpenStreetMap journey of Médecins Sans Frontières (MSF) started in 2014 with a mapping party in Berlin and field mapping in Lubumbashi. Eight years later, OpenStreetMap is the reference geographical dataset for most of MSF operations on the ground. OpenStreetMap has been used in public health interventions, disease outbreaks, mortality studies and to support large logistical operations. Still every day we are learning how geographical information can support us in doing our job better, in reaching more people, in saving more lives. MSF is not only using OpenStreetMap, but it also actively contributes to the map through the Missing Maps project launched in 2014 together with the American Red Cross, the British Red Cross, and the Humanitarian OpenStreetMap Team. Since then, MSF has trained dozens of Missing Maps champions, co-organized hundreds mapathons and involved thousands of volunteers in some 30 countries. In this presentation we will talk about how MSF uses OpenStreetMap internally and how we contribute through remote and field mapping. We will also share the lessons learned and reflect on the biggest challenges for MSF in creating and using the OSM data. about this event: https://2022.stateofthemap.org/sessions/XJZBFH/

OpenIndoor is an open source SaaS that uses OpenStreetMap indoor data to display a 3D graphical rendering of building interiors. The resulting map offers a gamified experience to meet different types of needs: indoor navigation, data representation, immersive tour etc. We will discuss how we use the available open data and the Maplibre engine to address these different use cases. about this event: https://2022.stateofthemap.org/sessions/HEKRDY/

This talk explores the effects of OpenStreetMapping on the mappers. These effects also infer that OSM mapping can be used as a tool for skill-building. OpenStreetMap, the crowdsourced geospatial database, currently has over eight million registered members [1]. This makes it one of the largest VGI projects with proven multifaceted use cases e.g. post-disaster response, combating female genital mutilation, app development, and navigation. The database is wholly made and maintained by its contributors, making all decisions without a top-down governing authority. People in OSM contribute in multiple ways, extending databases, onboarding newcomers, building community, exchanging information, and providing public benefit. Within the OSM community, OSM mapping is regarded as a form of volunteering to create freely accessible geodata. However, recent studies suggest that the experience a mapper gains through the mapping process could be equally important as well [2-4]. Building on the existing body of knowledge, in this talk, we will share the findings our research on how mapping in OSM affects the mapper. Being a quality OSM mapper requires training and practice. The act of OSM mapping requires transitioning from having an interest in mapping to creating an OSM account, learning how to use the application, developing an understanding of the technical and theoretical dimensions of mapping, and then applying these skills and knowledge to accurately convert satellite imagery into map data. Such a process engages mappers in multiple decision-making processes and continuously exposes them to buildings, topographies, and features of satellite imagery. We suggest that such experiences affect the mapper in multiple ways. We studied a youth mapping internship called Digital Internship and Leadership (DIAL) Program conducted in three cohorts. We chose this internship program for its inclusiveness in terms of academia, gender, and the geographical locations that the participants came from. Participant mappers were called through an open invitation on social media. Recent high school graduates and undergraduate students participated in the mapping internship. They were from diverse academic backgrounds (geomatics engineering, architecture, crisis management, management, forestry, geomatics engineering, computer science and engineering, electronics engineering, management, public health, mechanical engineering). The internship aimed to reduce OSM data gaps in rural Nepal through the involvement of Nepali high school graduates. The program was designed and executed by Kathmandu Living Labs (KLL). We studied the self-assessed experiences of the participant mappers at two different points of time: (i) during the mapping program (ii) after two years for Cohorts II and III, and three years for Cohort I. Short-term effects were studied through grounded theory coding of reports and blogs documented during the internship period. For long-term impacts, an online survey administered to identify if the effects persisted. Results show OSM mapping helps the mappers develop a number of vital skill sets and expand their knowledge in a variety of areas. Some of them are: deepening of civic engagement, development of social identity, expansion of geographic knowledge, spatial awareness, increase in happiness and satisfaction. They retain most of these skills even in the long run, irrespective of differences in gender, academic, or professional backgrounds. Surprisingly, 44.8% of the participants cited ‌they considered being a professional mapper or cartographer at some point in time because of their experience in DIAL. The same people report that OSM mapping increased their belief in their ability to help society. Apart from these individual benefits, we also sense collective benefits. Collective benefits such as network development and an increased sense of civic responsibility hold potential to facilitate broader public good. These might also be applicable for youth mobilization, team building, and collective work. It is difficult to pinpoint the exact root cause of this development, however, the benefits of OSM mapping may be in part related to the continuous exposure to satellite imagery, continuous use of technology, the requirement of multiple layers of decision-making, humanitarian aspects of OSM, and the growing global OSM community encouraging conversations around it. Our findings build upon the studies of the use of OSM in high schools, which was noted to increase creativity and spatial awareness among the students [5, 3, 2]. When compared to Minghini et al.’s (2016) study with ten-year-olds, the similarity in findings suggests ‌these developments might be similar across ages. These developments suggest new directions toward the use of OSM as a tool for youth skill building and youth community engagement and design newer incentive mechanisms for people to join and retain in OSM. There is still a huge scope of investigation left in this area, ideally through a longitudinal study with a bigger and more diverse sample and comparisons between different program designs, to fully understand the wide array of effects of OSM mapping on the mappers, as well as the potential to deepen positive outcomes via associated youth learning and leadership programs. There are undoubtedly other categories of benefits of OSM mapping that are yet to be identified. Hence, it is worthwhile to reconsider the idea of participatory mapping and related programs, and their effects on the contributing mappers. about this event: https://2022.stateofthemap.org/sessions/HSSWBD/

OpenStreetMap data represents a valuable source of information for public green areas in large urban centers and effectively measures the United Nations' Sustainable Development Goal 11.7. Our study provides a threefold contribution in this direction. First, we validate land-use-related tags in OpenStreetMap, through a comparison with satellite data from the European Urban Atlas. We then propose a framework and an interactive tool to measure access to public green areas through several established indices. Finally, we show how the framework can be used to simulate the impact of new green areas and help policymakers design effective interventions. As of 2020, around 55% of the worldwide population lives in urban areas and the World Bank estimates forecast an increase of around 1.5 times in the urban population by 2045. Cities are also major contributors to the climate-change, with a consumption of about 78% of the worldwide energy and a production of 60% of greenhouse gas emissions. A transition toward greener cities is often called as one of the solutions to reduce the environmental impact of cities, but also to make the urban environments more liveable, with positive spillovers on the mental and physical health of their population. In this context, the United Nations' Sustainable Development Goals 11.7 [1] indicates the need to make cities more inclusive and safe, but also environmentally sustainable, calling for the universal provision of safe, inclusive, and accessible, green and public spaces. A proper evaluation of this target requires complementing standard average metrics, looking for instance at the surface of green areas per capita within an urban area, with more sophisticated metrics, that are able to capture the interplay between the spatial distribution of both the population and green areas within a city. A few studies on selected cities worldwide highlighted the importance of considering this interplay [2-7]. A recent study on the city of Seoul [3] shows that vast portions of the parks in the city are located in outer areas so that frequent opportunities to visit them are relatively minimal. In general, urban green areas in Seoul are inadequately distributed in relation to population, land use, and development density. By contrast, in the case of Shanghai [6], the degree of accessibility to green areas appears to decrease as we move from the city core to the urban periphery. The authors also found a negative association between the degree of accessibility to green areas and the housing prices, which translates directly into a large environmental inequality, wherein wealthier communities benefit more from green space accessibility than disadvantaged communities. A similar socio-economic, but also ethnic, stratification is observed in the city of Chicago, where white-majority census tracts generally enjoy a significantly higher degree of accessibility to green areas than minority-dominated census tracts [7]. The former ethnic group also presents a lower income-based green-areas accessibility inequity compared to the other racial-ethnic groups. Efforts to move beyond case studies and provide more accurate cross-country indicators have led to the construction of the 'generalised potential access to green areas’ from the European Commission, which is provided as one of the city-level indicators of the Global Human Settlement - Urban Centers Database [8]. The metric measures the proportion of the urban population for urban centers included in the atlas living in high green areas. Based on satellite data on the Normalized Difference Vegetation Index, the metric is however agnostic with respect to the characteristics of these high green areas - for instance, whether these are public or private green areas - and any accessibility notion, since the metric does not consider that people can move from their residential location. These limitations are accounted for in a recent study for the European Environmental Agency [9], whose geographical coverage is however limited to specific urban hotspots in Europe, for which high-resolution land use data from the Urban Atlas (https://land.copernicus.eu/local/urban-atlas) is available. With its worldwide coverage and detailed mapping, the use of land use and street network data from OpenStreetMap [10] allows to expand the analysis beyond the European boundary. Our study provides a threefold contribution in this direction. First, we compare detailed high-resolution land use data on green uses for European hotspots included in the Urban Atlas with land use-related tags in OpenStreetMap for similar geographical areas. We use similarity indices to assess the degree of completeness of the OSM tags of natural land uses in urban environments and show how the quality varies according to the type of natural use and the size as well as the geographical area of the urban center under consideration. Second, we propose a framework for the monitoring of the target for large urban centers worldwide. In particular, by leveraging data from OpenStreetMap and population estimates from the Global Human Settlement [11], we develop a framework to measure accessibility to public green in large urban centers worldwide at a high resolution. For each urban center, we identify natural green areas using OSM tags on ‘land use’, ’natural’ and ‘leisure’ (e.g.: ‘leisure’:’park’) and extract the walkable street network to measure walking distances. Accessibility indices are then constructed for each populated cell of the population grid. The framework is also used to build an interactive tool to navigate our results, which can be customized to select the type of green of interest, as well as the size of the green area. Following the academic literature on urban accessibility, we build several accessibility indices, from a minimum distance index to exposure metrics. The resulting database represents a valuable source of information for policymakers to identify cities that are missing out and direct attention to those subareas within otherwise well-performing cities where the degree of accessibility is still insufficient. The constructed indices are then used to study the relationship between the measured level of accessibility and the structural characteristics of the cities and unveil the role of small green areas as accessibility enhancers, particularly in densely inhabited urban centers. Thirdly, we show how the framework can be used to simulate the impact of different urban interventions, from the addition of a new public green area to infrastructural interventions to the street network, to help policymakers to shape transitions toward more sustainable and accessible urban environments. about this event: https://2022.stateofthemap.org/sessions/TA9VAF/

Cities worldwide encourage urban active mobility by advocating policy and planning. Although contribution is evident, in practice, these actions disregard population parts that have mobility impairments. This research suggests using OpenStreetMap data in customized analytical models to assess the accessibility level of the urban environment for visually impaired pedestrians. Models results show the existence and spatial distribution of existing accessibility problems, including challenging street network connectivity and dangerous walking areas. These models can be used to enable decision makers, city stakeholders and practitioners to enrich management, monitoring and development of their cities, and support sustainable, livable lifestyles and walkability equality. Many efforts that include city policy and planning strategies are implemented to encourage urban active mobility. The outcome of these actions is measured by how transportable and accessible the city is. Although contribution is evident, in practice, the commonly used measures mostly disregard a huge part of the population that have mobility impairments, which require specific accessibility needs, preventing them to be an equal part of the sustainable city vision. This research suggests using OpenStreetMap (OSM) data in customized analytical models to assess the accessibility level of the urban environment for visually impaired pedestrians. In principle, the models analyze the city on two levels: routing and accessibility. These are evaluated, correspondingly, based on possible routes, e.g., how long the optimal route is for visually impaired pedestrians compared to the shortest one, and on area, e.g., what is the overall accessibility and safety of a predefined urban extent. The play of both measures enables us to quantify the level of mobility and accessibility of the analyzed city. To do so, we implement the following steps: 1. We examine the navigation preferences of visually impaired pedestrians in the urban space. This allows a better understanding of the various environmental and morphological factors and characteristics of the urban form that promote safe and accessible navigation. These are translated into spatial and temporal criterion: a) Way Type, which quantifies how suitable the path is in terms of usage and safety; b) the existence of Vision Impairment Assistive Landmarks that support safe wayfinding and navigation; c) Way Complexity, which measures the level of linearity of the path; and d) Crowdedness, which measures the overall pedestrian traffic volume. 2. We transform OSM’s street network into a weighted graph, where for each graph edge we calculate the cost according to the above criteria. Cost is derived from segments that facilitate safe and accessible walking for visually impaired pedestrians (e.g., separated sidewalks and straight paths), and segments that hinder safe and accessible walking for visually impaired pedestrians (e.g., shared and overcrowded streets). 3. We develop three analytical models that measure the accessibility level of the urban environment for visually impaired pedestrians: a) street-based, which relies on averaging the costs of all graph edges for a given area, hence it can be implemented for different urban levels (spatial extents); b) centrality-based, which adds on the street-based the centrality indices betweenness and closeness that consider the significance of each graph edge in the street network in respect to all other edges (high centrality values mostly signify streets that attract large pedestrian traffic flow); c) route-based, a navigational method, in which numerous routes are generated on the graph for location tuples, and then the weight ratio of the optimal route for visually impaired pedestrians and the shortest route (commonly used for seeing pedestrians) is evaluated. The smaller the weight value, the more accessible the route. The developed models are evaluated for Greater London, the UK. 33 boroughs with their wards are analyzed, resulting in processing 421,107 streets, 377,164 OSM nodes and 634, 871 OSM ways. Results show the existence and spatial distribution of accessibility problems for visually impaired pedestrians. The street-based model highlights the fact that urban nature and green spaces, which are typically considered as contributing to wellbeing and encourage walking, are less accessible for visually impaired people, mostly due to the existing road types, e.g., gravel and dirt roads or shared spaces (bikes and pedestrians that share the same path), which are less accessible for this population. The centrality-based model shows that central streets are mostly more accessible, meaning that borough centers are considered in general as accessible, but as distance from city centers grows, the urban environment becomes less accessible. The route-based model, where more than 1,500,000 routes (with length shorter than 1,000 meters) were calculated, showed that on average the optimized routes are 11% longer and 17.5% more accessible than the shortest ones. Some optimal walking routes are twice as long as the shortest ones, where some impose safety issues that critically endanger visually impaired pedestrians. Wards that have a large proportion of street segments with poor accessibility evenly distributed throughout the ward tend to show less efficient route planning in terms of optimal routes that are considerably longer. In general, the route-based model produces clearer results to understanding the city’s morphology in terms of accessibility for visually impaired pedestrians. To a large extent, these models depend on the quality of OSM data, such that feature completeness and tag correctness should be investigated. In terms of completeness, we found that sidewalks and crossings, which are two important model features, are not always mapped in OSM, mostly in the outskirts of London. One solution is to use learning methods and prediction models to complete missing data. In terms of tag correctness, we found that some inconsistencies exist with certain tags. One solution can be to make tag definitions in, e.g., OSM Wiki, more inclusive and clear, with a focus on accessibility aspects. Results show how various accessibility levels for visually impaired pedestrians might be assessed and where they are found in the city, pointing to the existing problems this community faces today when navigating. These include challenging street network connectivity and dangerous walking areas. The results also demonstrate that the current practice of urban planning and design worldwide still suffers from lack of democratization, limiting the mobility and navigation of certain groups. The accessibility models developed in this research can be used for better city planning and design, enhancing the city mobility and walkability equality and improving quality of life for these vulnerable road users. Our findings provide analytical tools to enable decision makers, city stakeholders and practitioners to enrich management, monitoring and development of their cities, and support sustainable, livable lifestyles and walkability equality. These, in turn, will ease navigation and mobility of visually impaired pedestrians, overall improving health outcomes and their integration into society. about this event: https://2022.stateofthemap.org/sessions/MXS9R8/

As part of a larger endeavour to make floor plan representations from building models available for indoor map and navigation services, we study the integration of IFC and OSM. # Introduction, background, motivation Official geo data is increasingly published not only in the form of 2D maps, but also in 3D, mainly as city models in CityGML. Usually the outer shell of buildings is captured in such models, but they may also involve more intricate detail. Even more detailed building models are generated during the planning process for new buildings and renovations. These are nowadays produced in digital form, archived in as-built phase by owners and operators for the life time of a building and, in the future, may even be required to be submitted for building permits. At the same time there is an increasing public interest in detailed information about public and semi-public interior spaces, for example about their accessibility, localization of barriers or targets (e.g. contact persons in public administration, shops in malls, booths on fairs, markets or larger info events, departments or hospital wards) or resources (e.g. books in libraries, charging stations, fire equipment or defibrillators) or to get a first impression in advance (e.g. virtual open day). The interest and the points of interest may be temporary or permanent. Since the context of creating and capturing geo data and building data is fundamentally different, there is hardly any integration. Indoor data for maps and navigation models is manually captured or at best derived in undocumented multi-step semi-automatic workflows. # Aim and purpose of the study The project "Level Out" sets out to develop automated methods and services to make detailed indoor data from digital building models selectively available for the population of city models, map and navigation services (in the form of 2,5 D floorplans). Towards this end, we are developing a platform to check building models whether they are suitable and contain required data, extract selected and simplified indoor data and convert it into various formats: CityGML LOD0 (Indoor), IndoorGML and OSM Indoor. As input we rely on data in the format IFC (Industry Foundation Classes), the most widespread standard format for digital building models. Indoor OSM, in particular geometry with Simple Indoor Tagging, is one of the various extraction targets. The data created may not be directly fed into OpenStreetMap, but serve as a viable base for further mapping. There are already older solutions, e.g. BIMServerOsmSerializer (), which are only built for a version of IFC, which has been a long time standard version, but currently approaches towards its end of life: IFC2x3. There are also solutions under active development, e.g. the JOSM plugin "Indoor Helper" (), which, however, lack some general approach on the IFC side and coverage of the heterogeneous options to represent geometry in the IFC schema. With this research and development we aim to provide a workflow and software to systematically access floorplan data in IFC. # Methodology We start from both ends of integration by looking at the detailed structures of the source and target models in parallel. From the group of target models, we derive a common model, which will have, at best, near-trivial mappings to OSM Indoor, CityGML, IndoorGML. Although not strictly necessary for the IFC-to-OSM conversion case or any other bilateral integration, the intermediate model will not only allow to tackle integration of IFC with multiple targets besides OSM, but also integration of OSM with multiple sources besides IFC. Next, we identify relevant information in the source model. IFC exposes a wide variety of geometry modelling constructions from CAD software, mainly following the modelling paradigm of constructive solid geometry (CSG). So far, we found the following principle representation options: a) Direct floorplan representation in 2.5D: Here we have 2D representations located in 3D space, usually located at the level of the floor finish for a particular storey. There are two versions to be distinguished: space boundaries versus abstract representations of space-defining elements. b) Extraction from CSG: Spaces (as well as constructive elements) are often represented as solids resulting from extrusion of a planar shape. If extruded in z-direction, the base shape can be extracted and used as 2.5-D representation. c) Projection onto floor level: If the geometry is not in CSG-form with extrusions, but in BREP (boundary representation), then projection followed by a simplification of the projection result is a possible way to extract. In addition to the geometric elements, there are semantic elements connected to the geometry that are connected themselves and can be used to charge the geometric model elements with meaning. Depending on the geometry extraction method, correlation and consideration of semantic elements is more evident or complicated - hence possible to different degrees. The paper will discuss these implications. After identification of the relevant entities, we are developing a three stage process for the actual population of target models from IFC. 1. Building model enrichment: Information that can be represented in IFC will be played back to the building model instead of being promoted to the generic model only. 2. Building to intermedite model: This essential step is coved with a flexible rule-based mapping. 3. Intermediate model to target models: Following a careful design of the generic model, this step should be simple. We are testing the processes with data from public buildings, two sets of university campus buildings as well as one newly built municipal administration centre. From assessment of the original building data, we will also develop modelling and export guidelines for BIM software. As far as possible, the demo data will be made available publicly as open data. More important, the conversion procedures will be published open source and a respective conversion service will be offered online. # Discussion In summary, our work provides practical benefit in terms of tools to support the mapping process as well as a scientific contribution in terms of spatial data integration and expert involvement via domain-specific languages. The practical benefit of the conversion seems obvious: Building owners can publish data of their publicly accessible spaces to help with volunteer mapping work. In the future we will also tackle update, checking and comparison with existing OSM indoor data. Scientific contributions are also made in different ways: First, an application scenario for the OGC Indoor Feature Model is provided and - interesting for the audience of this conference - evaluation of how OSM data fits with the generalized model. Further, we explore methods for flexible data integration with domain specialist and expert community involvement. Finally, but beyond the scope of this conference, the applicability of integration methods for bidirectional integration with multiple sources and targets via intermediary formats is evaluated. about this event: https://2022.stateofthemap.org/sessions/ZUXTN8/

The lack of data on the distribution of the water resources, possess a great challenge for the water resource investment and AI/ML-enabled advancements in the water sector compared to all other sectors like heath. This paper describes the methodology for combining different water mapping schemas to create comprehensive multi-platform water infrastructure data and enhance rapid updates to support a suite of water resource analytics and extended advanced technology explorations towards improved decision-making. Access to clean and safe drinking water is critical to public health and socioeconomic prosperity, yet an estimated quarter of the world’s population lacks such. This was evidenced by the unprecedented outbreak of the COVID-19 pandemic, which left communities extremely vulnerable to fatal illnesses due to the limited access to water for handwashing or lack of knowledge of the existence of the utility. Subsequently, the lack of data on the distribution of the water resources poses a great challenge to the water resource investment and AI/ML-enabled advancements in the water sector compared to all other sectors like heath. Influencing the frequency of water point data collection through crowdsourcing and volunteered geographic information, would greatly improve the availability of water point data, and contribute to the extended roles of water resource distribution, monitoring, and management especially in rural communities. Therefore, this paper describes the methodology for combining different water mapping schemas to create comprehensive multi-platform water infrastructure data and enhance rapid updates to support a suite of water resource analytics and extended advanced technology explorations towards improved decision-making. The recent technological advances including the web 2.0, cameras, smartphones and sensor networks continue to empower the development of empirical methods as well as the generation of big data and analytical platforms that provide predictive performance on the various socioeconomic needs for sustainable development. OpenStreetMap (OSM) is a crowdsourcing platform which offers a collaborative experience through its database, community, and wiki platforms, to create and update data relevant to support or transform various data deficiencies whether humanitarian or planning. However, the project’s data quality shortcomings often hinder simultaneous data integration with other analytical platforms such as the Water Point Data Exchange (WPdx) that would explicitly maximize the usage and application of these crowdsourced data. Through a project dubbed ‘Water Infrastructure Mapping Uganda’, a data model based upon open mapping methods and survey tools was developed to facilitate the mapping of water infrastructure data points and simultaneous updates of both the WPdx and OSM databases. The project engaged a comprehensive review of the OSM water tag, rural water infrastructure data standards and the WPdx database to generate a survey data form that supported one-time collection of a water point for both OSM and WPdx databases. Underlying the development of the data model/schema in the overall project, a design criterion was established which guided and justified the overall selection of the most relevant factors to include in the process that would eventually become detailed to communicate water infrastructure and functionality. The criteria were followed by an assessment of the; compliance [agreement of the tag], consistency [temporal and spatial representation of the tag], completeness [attribute description of the tag], and granularity [quality of the event information] of the OSM tag to support the development of the. osm language in the Kobo toolbox. Gulu district, located in the North of Uganda, was identified as a potential pilot area for improving the approach created by the project based on its rich WPdx footprint as well as a well-established OSM community of YouthMappers. Up to date satellite imagery of up to 50cm spatial resolution was acquired through the USAID GeoCentre to facilitate any visual detection of water points, and the digitization of base map data including, buildings, roads and waterways, to be employed in the field mapping exercise. A field mapping workflow was designed to facilitate the field-data collection employing the developed water infrastructure data model and Kobo toolbox. An API link was developed that simultaneously tapped the open-source field collected data into the WPdx database. Through the project, more than 15000 buildings, 1400square kilometres of roads and over 500 water data points were added to OSM as well as the WPdx database for the later data. Also, from the project, several observations were made regarding the improvement of such processes and the extension of the data model beyond one geographical area. The developed workflows characterized and provided a general improvement in the water infrastructure data quality especially for OSM based on WASH indicators used to officially report on the sustainable development agenda. The workflow development waivered the interoperability gap in geospatial data sharing platforms which often results from unharmonized data structures. It was established that the designed methodology cannot be applied to water data updates but rather to freshwater point data collection. This would lead to exponential water point data increase, however, the workflow may be revised to include the framework for data updates without having to engage the full field mapping process. As well, the data model design was mainly based on the African water infrastructure and open mapping reviews, hence, the transfer of the data model from one continent to another may require a review of some data factors to create better insights of the water indicators in a place of that given continent. about this event: https://2022.stateofthemap.org/sessions/JNCVKY/

OpenStreetMap (OSM) data has the potential to facilitate bottom-up approach to transport planning which is essential for localized data-driven policy interventions. Given this, OpenInfra project is exploring the potential of OSM data in transport research with a focus on active travel. The exploration showed that currently missing data limits the applicability of OSM data. Nevertheless, we argue that the potential and relevance of OSM data can be demonstrated by recategorizing OSM data to provide more actionable insights to policy-makers. This, therefore, could encourage the uptake of open data leading to more transparent, reproducible, and participatory transport planning. One of the key domains in which OpenSteetMap (OSM) data has been utilized is transport research [1]. OSM has been used in agent-based transport simulation [2] and routing [3], including cycling [4], walking [5], wheeling [6], and blind pedestrian routing [7]. Another application of OSM data is in transport infrastructure planning. Nelson et al. [8] argue that OSM has the potential to become a primary source of data on infrastructure across the globe. Regardless of OSM’s potential to become a primary source of data on infrastructure, its potential in active travel infrastructure planning is yet to be realized. One of the potential reasons behind this lag might be linked to the perceived unreliability of open-access crowdsourced data [9]. The quality of OSM has received extensive examination [1] in which the question concerning data completeness plays a significant role because, it is argued, the mappers are not coordinated to guarantee systematic coverage [10]. To address this issue, Barrington-Leigh and Millard-Ball [11] assessed OSM road completeness and found that globally over 80% of roads are mapped. Problematically, however, their assessment focused on roads designed for motor traffic, thus excluding other modes of transport. This gap has been partially addressed by Ferster et al. [12]who examined and compared OSM cycling infrastructure in Canada. They have not, however, considered the infrastructure from the perspective of accessibility. Moreover, there seems to exist no equivalent study using OSM data in the context of pedestrian infrastructure planning. Nevertheless, open-access crowdsourced data, such as OSM, can support an increasing need for local evidence to inform transport policies. This is important in the context of the UK in which a shift from provision for motorised modes towards more sustainable active modes of travel, such as walking, wheeling, and cycling, takes place [13]. The importance of localizing interventions to meet the needs of local communities has been outlined in both policy [15] and academic [16] papers. A potential way to engage citizens in the decision-making is to encourage “produsage” – a model in which citizens both produce and use data [17]. Acknowledging the potential of OSM to boost citizen participation, OpenInfra project, run at the University of Leeds (UK), aims to address the gap of literature regarding the potential OpenStreetMap in transport research. The project started by examining the existing OSM tags relevant to active travel infrastructure in England with a focus on West Yorkshire, Greater Manchester, Greater London, and Merseyside. The data has been queried using osmextract [18], a package in R, and explored using exploratory data analysis (EDA) approach. A reproducible code containing all the figures discussed here can be found on GitHub: https://github.com/udsleeds/openinfra/tree/main/sotm2022 Given the extensive use of OSM data in transport research, it is not surprising that OSM provides a comprehensive active travel network, yet there is a lack of specification concerning the type of infrastructure that is present (e.g. is it a cycle lane or a cycle track?). For instance, cycleways and footways constitute about 1/3 of all the mapped highways on which one can legally walk, wheel or cycle but only a few percent of the cycleways and footways have tags detailing their type. The data gets even scarcer in the context of accessible infrastructure planning. For example, there is a lot of missing information on the presence and type of kerbs – a street element that might make the movement of a wheelchair user more challenging [19]. The missing data currently limits the use of OSM data in active travel planning, however this does mean that the use of OSM data should be dismissed. Following Nelson et al.’s [8] argument that it is important to make crowdsourced data more actionable, we decided to recategorize OSM data based on Inclusive Mobility (IM) [15], a guide that outlines the best practices in creating inclusive pedestrian infrastructure in the UK. For this, a function has been written (documentation can be found here: https://udsleeds.github.io/openinfra/articles/im_get.html). It takes an OSM dataframe, recategorizes its tags based on the definitions outlined in the guide, and returns an OSM dataframe with new columns to use in further analysis. However, the function provides a simplification of the IM guide for a couple of reasons. The first one could be considered in terms of definitional discrepancies. For instance, the guide defines footways as “pavements adjacent to roads”, yet this is not easily extracted from the OSM in which highway=footway is a generic tag and often there is no further refinement (e.g., sidewalk=*) to determine if it is a pavement adjacent to a road. Another reason is linked to assigned values. For example, the guide identifies six tactile paving surfaces but OSM focuses on the presence/absence of tactile paving, thus limiting how much information can be extracted from the data. One potential application of the IM function could be to explore the existence and geographic distribution of accessibility indicators, such as the presence of a flush kerb. Yet, more interesting results can be produced by using recategorised OSM data in conjunction with other datasets that would help to improve the understanding of the accessibility of streets. As an illustration for this, an open-access Leeds Central Council Footfall data was used [20]. We reasoned that the locations at which footfall data were collected are heavily used by pedestrians, thus demonstrating the need to ensure inclusive spaces. 5 unique streets were identified, which resulted in 35 linestrings in OSM. Then, a basic index of accessibility, ranging from 0 to 5, was created. For example, if a linestring is classified as a footway, footpath, or implied footway based on the IM guide, then it received 1, otherwise 0. If a flush kerb is mapped, it received 1, otherwise (e.g., not flush or NA), 0 is given. Finally, the values were added and a final index produced. Following this, the highest index score is 2 (19 linestrings), while the rest scored 1. This example does not necessarily show that the streets are inaccessible because the missing data make it hard to make a fair judgement (e.g., in this case not a single linestring has data on kerbs). However, we would argue that this is a space for OSM to produce more readily actionable insights regarding transport infrastructure, especially if joined with other (open) datasets that would help to overcome some of its current data limitations. The following steps of the OpenInfra project are focused on scaling up. The goal is to produce ‘OSM transport infrastructure data packs’ for transport authorities in England to support the uptake of open-access data, such as OSM, in transport planning. We believe that the utilization of open-access data could make transport planning more transparent, reproducible, and participatory which, consequently, would support an uptake of sustainable modes of travel. OSM specifically has the potential to provide localized insights on the existing transport infrastructure and facilitate more inclusive and accessible transport planning. about this event: https://2022.stateofthemap.org/sessions/CEMMTQ/

Lighting talks registered during the State of the Map conference. ## Offline Web Mapping Server UNVT Portable The United Nations Vector Tile Toolkit. _by Shogo Hirasawa, Taichi Furuhashi_ ## Liaising OpenStreetMap (OSM) Community and Research Community with the Policy Makers: Reducing the Data Gap in Disaster Management _by Airin Akter, Shraddha Sharma_ ## Unique Mappers Network: The OpenStreetMap Community NGO in Nigeria _by Victor N. Sunday, Nwinkua Dumdibabari_ ## NOAH (Nationwide Operational Assessment of Hazards) Website, revamped! _by Feye Andal_ about this event: https://2022.stateofthemap.org/sessions/NNKX8K/

Null Island is where the prime meridian meets the equator at (0,0) longitude and latitude. While Null Island is a fictitious, dimensionless, point object, its existence stimulates vigorous debate making it worthy of serious consideration. Many examples exists illustrating how Null Island impacts OSM discourse. Our study considers what the geographic oddity of Null Island means for OSM. The main contribution is a structured knowledge-based resource facilitating understanding of Null Island’s impact on OSM. This socio-technical and philosophical investigation of Null Island can become a catalyst for deeper discussions and debates in OSM around mapping practices. Null Island refers to the location where the prime meridian meets the equator at 0o longitude and 0o latitude. With coordinates (0, 0), it is the origin of the WGS84 geographic coordinate system. It has been argued that Null Island can be considered a real place that is a product of our digital age [1]. Null Island’s significance comes from the fact that it is erroneously associated with large amounts of geographic data that spans across geo-social media, location-based services and map databases. Even though Null Island is a fictitious, dimensionless, point object, its existence stimulates debate that elevates Null Island into a global issue worthy of serious consideration (a detailed description of associated issues is given in [1]). Members of the OpenStreetMap (OSM) project often interact with this location in various ways, and therefore understanding what Null Island means for OSM is relevant. We can find several examples of Null Island affecting OSM, such as a recent debate that arose in the talk mailing list in January 2022 with the title “Was the deletion of Null Island reasonable?” [2], where contributors argued for or against the deletion of Null Island. In addition, a web search for the term “Null Island” on the openstreetmap.org domain [3] reveals that Null Island was mentioned across the entire OSM ecosystem, including mailing lists, forums, user diaries, notes, features, changesets, wiki pages, help articles, blogs and even the Ruby on Rails codebase of the OSM website uses Null Island for testing (https://tinyurl.com/OSM-Ruby-Null). These suggest that Null Island already has a widespread reach within the OSM project. The purpose of this study is to consider both qualitatively and quantitatively what the geographic oddity of Null Island means for OSM. No research works exist which tackle this issue in depth. Previous studies mentioning Null Island do so in a simplistic way and use the term to refer to the (0, 0) location (see e.g. [4]–[6]). Only a few studies recognize it as a special location and unique phenomenon ([7], [8]), and to our knowledge, only one study tackles the issue in depth [1]. In addition to contributing a robust academic study of Null Island, this work will produce a structured knowledge-based resource for the community to understand Null Island’s impact on OSM. Building on [1] we investigate the various ways Null Island is represented in the OSM project subsequently contributing an evidence-based narrative history on the evolution of Null Island. This includes the qualitative review of various OSM communications channels (e.g. mailing lists, discussion boards and wikis) for mentions and references to Null Island. We believe these channels help provide insights about how the OSM community contextualizes, describes and deals with Null Island. The history of special map features related to Null Island, such as node #1 (https://tinyurl.com/osm-first-node) and the node located at (0, 0) (https://tinyurl.com/OSM-Center) will also be reviewed to illustrate what actions the OSM community took in terms of adding and removing Null Island to the database. In addition to these qualitative approaches, we utilize the ohsome API [9] to extract and analyze map edits made on or near Null Island, which provides a quantitative way to assess the frequency of erroneous data added to OSM near (0, 0) as well as the semantics of such data. Interesting patterns have already emerged from the preliminary analysis of data. The most recent mailing list debate mentioned above [2] can be summarized as follows. 17 individuals contributed 45 e-mails to the discussion between January 3 and January 10, 2022. One of the (very few) rules of OSM is that data should be verifiable, meaning that others can visit the real location of a map object and see for themselves if the data is correct. This is also known as the “ground-truth rule” [10]. Null Island as a fictional place violates this rule, therefore a popular stand in the debate is that it should not be part of OSM. This was explicitly expressed by five individuals, including a member of the authoritative Data Working Group. A counter argument is that Null Island is fundamentally similar to localities and neighborhoods, that might not exist as political or physical entities, but are known only informally to a group of people inhabiting that area. In this sense, Null Island is a place that exists in the collective consciousness of people and the name refers to the same geographic area. This justifies tagging the (0, 0) location as place=locality and name=”Null Island” in OSM. This view was explicitly supported by seven members on the mailing list. The remaining five individuals that contributed to the discussion did not take a clear stand on whether to remove or keep Null Island, but have provided arguments both for and against the deletion of it. The full history of OSM data was extracted from the elementsFullHistory endpoint of the ohsome API [9] within the geographic bounding box defined by the southwest point of (-0.001, -0.001) and the northeast point of (0.001, 0.001) between January 1, 2012 and January 1, 2022. During this 10-year-long period, a feature was added, deleted or modified every three days on average within this bounding box, resulting in 1323 unique features (nodes, ways or relations). In addition, map Notes as well as GPS traces are also constantly being created, which makes Null Island and its surrounding a busy area in terms of OSM data activity. Null Island is a socio-technological concept that has only been sparsely present in the GIScience literature so far. Our novel approach highlights how a seemingly lighthearted topic like Null Island can generate serious debates that are technological, social and even philosophical in nature. OSM and Null Island have a long tradition together with sometimes heated mapping debates resurfacing from time to time with no apparent resolution in sight. While resolving these debates is entirely in the hands of the OSM community, our research contributes to the potential resolution of them in a meaningful way by providing a factual, detailed, and accurate account of Null Island in OSM. Furthermore, while Null Island is potentially the most prominent example of a fictional place affecting maps and mapping practices, other examples also exist. For example, the most remote location on Earth, Point Nemo (which is the point in the ocean that is farthest from land) [11] is also present in OSM (https://tinyurl.com/OSM-PointNemo). Our OSM specific investigations together with a more general introduction of Null Island from both technological and social perspectives presented in [1] will help demystify the abstract concept of a fictional place that is present in real databases. Increased understanding will potentially help OSM members resolve mapping debates about “real fictional places”. Discussion around Null Island and other fictional places is unlikely to end with this work. Our work will contribute in a technical, socio-technical and philosophical way to the Null Island story in OSM with the potential to become a catalyst for further discussions related to wider debates in OSM around mapping practices. about this event: https://2022.stateofthemap.org/sessions/LTA77E/

Sparked by concerns about OpenStreetMap's role in how the public accesses and recreates on protected lands, OpenStreetMap US volunteers, navigation app developers, national agencies and public land managers formed the OpenStreetMap US Trails Working Group in 2021. Bringing together a diversity of perspectives on trail mapping practices, trail safety, and protecting the environment, this group is working to address on-the-ground challenges, tagging schemes, authoritative data, and other topics related to mapping trails in OSM. Learn how this group is collaboratively developing solutions for responsible trail mapping in OpenStreetMap. about this event: https://2022.stateofthemap.org/sessions/CUV9H7/

Urban green spaces serve people for active and passive recreation. On the basis of OpenStreetMap data, suitable green spaces are to be derived in order to incorporate them as recreation destinations in a location-based service (the “meinGruen” app) as polygons. The modelling approach focuses on activity-related barriers in the context of urban green, transitions between different land use classes, and public accessibility. The case study was implemented for the city of Dresden in Germany. In addition to important ecosystem services such as clean air or local climate regulation, green spaces provide peace and recreation, contributing to a good quality of life for the population. In high-density urban areas, publicly accessible green spaces are used for a variety of recreational activities, which has become even more important, not least because of the COVID-19 pandemic [1]–[4]. In this context, the research project "Information and Navigation on Urban Green Spaces in Cities - meinGruen" examined publicly accessible green spaces with regard to a variety of criteria in order to assess their suitability for the pursuit of leisure activities, such as going for a walk or playing soccer [3], [5], [6]. The aim of this study is to derive a suitable polygon dataset to describe the spatial distribution of publicly accessible urban green spaces. The presented approach favors the use of OpenStreetMap data and intrinsic knowledge. Advantages of the use of OpenStreetMap data are the global availability, the often high completeness in urban areas as well as the unified open data license ODbL 1.0. In this way, problems with data availability and heterogeneity due to different responsible authorities can be avoided. Ludwig et al. [7] describe an approach to mapping public green spaces based on OpenStreetMap and Sentinel-2 satellite imagery in which barriers and land use changes are considered based on a priori (expert knowledge) assumptions for polygon generation. In the approach presented here, spatial delimitation is to be refined by describing barriers by probability values. The term "barrier" is first analyzed in an interdisciplinary way in order to then work out its meaning for the spatial delimitation of a green space. Here, barriers describe the action space of a recreational activity. While there are a number of object types (such as walls, fences, rivers, roads or railroad lines) can be assumed to be barriers with certainty, there are others (such as paths or the change of land use) for which knowledge is still lacking. The study area includes the city of Dresden in Germany, plus a buffer of five kilometers. OpenStreetMap represents the main data source. For training and validation, official cadastral data (ALKIS) as well as a dataset on cadastral parcels owned by the city of Dresden were used. The methodology consists of six steps: First, according to defined rules, types of barriers were extracted from OpenStreetMap data. Second, we derived a land use layer without overlaps and holes from OpenStreetMap. Here, two options were compared regarding different target schemes for land use classification. Third, a mapping in terms of a “ground-truth“ in selected areas in Dresden followed in order to be able to evaluate the existence of a barrier on site for the extracted paths and changes of land use. Fourth, generic probabilities for the existence of a barrier were determined based on path type or land use change type. Fifth, a polygon mesh was created by applying thresholds to the determined barrier probabilities. Sixth, the generated polygons were enriched with attributes on the number of green space-related POI, such as benches, trash cans, or trees. Models for "greenness" and "accessibility" are thereby trained. For the technical implementation mainly Docker, PostgreSQL/ PostGIS, Python (Geopandas, Scikit-Learn) and Jupyter Notebook were used. Data import was performed by osm2pgsql and ogr2ogr. For mapping we used the app QField. Land use layers were successfully generated for the study area using a residual class. The results indicated that the land use classification according to the area scheme of the IOER-Monitor (option B) has a higher thematic accuracy with a maximum of 33 classes (433 original OSM tags were assigned) than the option A based on a classification according to osmlanduse.org/ Schultz et al. (up to 13 classes, based on 61 OSM tags) [8], [9]. The classes of arable land (A: 28.40% / B: 28.06% share of area) as well as forest (A: 21.81% / B: 23.33%) are dominant in both variants. While the residual class takes up 6.29% of the area in option A, it is only 4.88% in option B. For the “ground-truth”, a total of approximately 82.3 km of paths (with 408 line objects) and approximately 64.2 km of land use changes (1720 line objects) were evaluated for the presence of a barrier in two selected areas in Dresden. The land use changes are based on variant B. Data were collected on 61 different land use transitions and four different trail types. While bike lanes can be safely assumed to be a barrier, the "track" (96.8%), "footway" (92.7%), and "path" (86.0%) trail types have a slightly lower barrier probability. Among land use transitions, the forest-meadow (12.6%), meadow-sports facility (22.8%), meadow-park (24.6%), and forest-grassland (26.7%) transitions have the lowest barrier probabilities. Together with the barriers assumed to be safe at the beginning, a line pool is formed, from which different polygon meshes are generated based on different intervals for the barrier probability (p ≥ 0%; p ≥ 20%; p ≥ 40%; p ≥ 60%; p ≥ 80%; p = 100%). The lower the probability threshold, the higher the number of polygons created (whose area decreases). For the "accessibility" model, the number of benches, trash cans, public toilets and public internet were considered per polygon. The logistic regression achieved 76.7% accuracy here, similar to a Support Vector Classifier (SVC). The "greenness" model is based on number of benches, picnic tables, trees, and trash cans per polygon. The accuracy is about 92.3% (for logistic regression and also Support Vector Classifier). This work successfully demonstrates a new approach to derive publicly accessible green spaces based on OpenStreetMap data considering different qualities of barriers in contact of green spaces. Based on the examined barrier probability of path types and land use transitions, more realistic spatial delineations of green spaces were made possible. The chosen approach is globally applicable due to the use of OpenStreetMap. In each case, locally prevailing climatic and cultural influences must be taken into account. The knowledge collected here can be applied in the Central European region. For other areas, a renewed “ground-truth” may have to be carried out on site. The schematic transformation of the land use into the area scheme of the IOER-Monitor leads to a reduction of classes compared to the original data. In addition to benefits in capturing barrier probability, it also simplifies comparability and transferability. Thus, other data could also be migrated into this scheme. The determined barrier probabilities correspond to the expectations. The polygon generation based on different barrier probabilities allows here a differentiated setting of the desired action space for the relevant leisure activities. The quality of the trained models is good, but can be improved. A variety of additional features can be calculated for each potential green space (polygon), such as path network density or density of path network intersections (see also Ludwig et al. [7]). Questions about the perception and use of green spaces can also be part of interdisciplinary research in the future. about this event: https://2022.stateofthemap.org/sessions/ASADTB/

What must be mapped to make routing for prams and wheelchairs practical? Three years ago, the local meet-up in Dortmund, Germany, started a campaign to make step-free routing available for the general public. The lessons learned mean that such routing is possible, but there is a lot missing to map - both in Dortmund and in all other parts of the world. Map the essential where fellow mappers are sparse. And codify the full ground truth where the passion allows it. I hope to encourage mappers for the quest to get their neighbourhood ready for wheelchairs, prams and all the other pedestrians! Routing for pedestrians is a much broader challenge than the well-known car routing. Cars all over the world are mostly uniform, but pedestrians vary widely in their capabilities. This means that a lot of details that a sportive person might not even notice can be literally a roadblocker for people with prams, for wheelchair users or simply lesser-abled people with not enough strength for a complete stairway. Becoming a father has been a good opportunity to check in practice what is and what is not feasible for a pedestrian with a pushed vehicle. It turns out that the first step is to get aware of the various kinds of obstacles that get in the way. Beside the obvious steps and kerbs, there are impassable surfaces, too narrow or too steep sections. Or simply sidewalks missing completely on the ground. As of now, OpenStreetMap data does not even suffice to figure out where one or both sidewalks actually exist. This puts into perspective the discussions about how to map best details of both detached ways and sidewalks. A couple of tagging approaches are compared to allow educated guesses which level of detail will allow for good results rather in weeks and months than in years or decades. I even dare to give suggestions what tagging practices we should additionally adopt to be able to map faster. The background of this talk is an initiative from the Dortmund meet-up: For the large event Kirchentag 2019, we mapped at least the city center sufficiently well for wheelchair mapping. The whole city with its 1500 km streets has turned out to be simply too much. Given that a city with a local meet-up is in a relatively good position to be mapped, it was no surprise that also elsewhere the data is simply not yet good enough for wheelchair routing. The hope is that simple suggestions what helps is getting more traction than a sophisticated mapping hierarchy. about this event: https://2022.stateofthemap.org/sessions/LKZYJ7/

The fitness of OSM for multi-label classification is proven. A workflow to enhance OSM-based multi-labels using machine learning is established. The results are provided to the OSM community via the HOT Tasking Manager. # Introduction Land-use and land-cover (LULC) information in OSM is a challenging topic. On the one hand, this information provides the background for all other data rendered on the central map and is used by applications like https://osmlanduse.org. On the other hand, this information has a difficult position within the OSM ecosystem. LULC information can be quite cumbersome or even difficult to map e.g. due to natural ambiguity. The growing tagging scheme provides a collection of sometimes ambiguous or overlapping tag definitions that are not fully compatible with any official LULC legend definition [1]. Furthermore, the data is highly shaped by national preferences and imports. This diversity of the LULC data in OSM is a fundamental principle of OSM that enabled the success of the project. Yet, this can create considerable usage barriers or at least caveats for data users unfamiliar with the projects' ecosystem. The remote sensing community for instance has started to use OSM LULC information as labels in their classification models. Frequently, OSM LULC data has thereby been taken at face value without critical reflection. And, while the quality and fitness for purpose of OSM data has been proven in many cases (e.g. [2,3]) these analyses have also unveiled quality variations e.g. between rural and urban regions. The quality of OSM therefore can be assumed to be generally high, but remains unknown for a specific use-case. The proposed work first assesses the impact of these challenges on a use-case of multi-label remote sensing (RS) image classification and then provides a machine learning (ML) based workflow to overcome and finally mitigate them. Multi-labels are a type of image classification where a satellite image is labeled with multiple containing LULC classes. In the presented study these labels are extracted from OSM and used to train the ML algorithm. # Methods and Results The fitness for purpose of OSM for multi-label RS image classification was tested on a Sentinel 2 scene with a resolution of 10m and four bands in south west Germany recorded in June 2021. The area was chosen for its estimated high completeness and low amount of imported data. OSM data was grouped by its tags into the four LULC classes 'forests', 'agricultural areas', 'build-up area' and 'water bodies'. 18 tags that could unequivocally be mapped to these classes were used and small elements below the image resolution or the classes minimal mapping unit were filtered. The chosen scene was then tiled into a 1.22 x 1.22 km grid of 8100 image patches. Zero to four labels were assigned to each patch, based on the OSM LULC elements therein. Evaluation was performed manually on 910 random patches, of which 80% had a correct OSM-based multi-label, thereby proving the assumed high completeness and quality in the region. The proposed workflow provides a method to enhance this OSM-based RS image multi-label classification and extend it to areas of lower OSM quality and completeness using ML (specifically deep learning (DL)). The main obstacle for ML and especially DL is the required amount of labeled training data. Volunteered geographic information (VGI) like OSM offers a potential solution to this challenge by providing an overabundance of LULC information that is suitable for this purpose if data quality is sufficiently high. The workflow uses the multi-label information extracted from OSM for training and then detects discrepancies between its predictions and OSM. Using this information and pinpointing the exact location of error within the patches provides valuable OSM data quality information. Apart from facilitating a fast quality estimation for large areas, the workflow can make its findings automatically available to the OSM community in a feedback loop using the HOT Tasking Manager framework. Thereby the valuable service by the OSM community of providing large amounts of free and generally high quality training data is recognised in the form of quality feedback including mapping hints to the OSM community. The five workflow stages are: 1) RS data collection and preprocessing, 2) OSM data collection and preprocessing, 3) LULC multi-label modeling, 4) OSM data issue flagging and 5) the community feedback loop. While each step is an atomic use case and application, the combination of all four steps creates a tool that is useful for the RS and the OSM community likewise. The tool is openly available at https://gitlab.gistools.geog.uni-heidelberg.de/giscience/ideal-vgi/osm-multitag under the GNU Affero General Public License v3 including example datasets. Manual input was kept as low as possible while enabling the 'human in the loop' to take full control over all input and output. The workflow extracts multi-label training data in stages 1) and 2) as described. Stage 3) then trains a DL model to predict multi-labels using solely RS imagery. For demonstration, the model was trained on the described Sentinel 2 scene in Germany. The models' performance was validated on the manually labeled 910 patches where it outperformed OSM in terms of multi-label accuracy by 7%. When additional errors were manually introduced to the training labels to simulate areas of lower OSM quality or completeness, the model maintained an overall prediction accuracy above the noisy training labels. Alternatively, in cases where OSM LULC multi-label accuracy is suspected to be low, pretrained models from comparable regions with higher OSM data quality can be applied, making the workflow widely applicable. Any patches where the models' multi-label prediction contradicts the OSM-based multi-label are then detected in stage 4). Multi-labels can be incorrect if either a label is missing (omission), meaning data is missing in OSM, or if a label is wrongly assigned, meaning OSM data is falsely mapped within the tile. The data error type and location within the patch is then extracted using explainable AI [4]. The final stage 5) uses these localised potential OSM data errors to create HOT Tasking Manager projects via the public API. These projects provide additional correction hints. Yet, no automatic editing takes place. The community is kept in full control of all mapping actions as a 'human in the loop'. # Discussion The high quality but diverse nature of OSM has been proven for the use-case of multi-label RS image classification. The proposed tool provides an automated OSM multi-label extraction, modeling and verification procedure including a return of results to the OSM community. A major challenge of the approach is the tiled view on the data. If OSM assigns correct multi-labels to a patch, more fine grained data issues will not be detected. Yet, this approach allows large scale data assessments, before the topic of more detailed data improvement is tackled. It also allows to run repeated OSM LULC quality and completeness estimations for large areas over time. Another major benefit is the usage of local OSM data for modeling, thus making regionalised models the standard procedure. This is required for OSM LULC information as regional data structures and communities exist, that need to be preserved. The model can lead to regional homogenisation and data cohesion within these regional communities. about this event: https://2022.stateofthemap.org/sessions/EKEZ7R/

Using intrinsic quality indicators we explore how network quality, in terms of its suitability for navigation, varies across areas with relatively high and low corporate editing in OpenSteetMap. Our work shows areas with relatively high rates of corporate editing exhibit not only an overall increase in data quality, but also increased rates at which quality improves. OSM (OSM) contributors have traditionally lacked explicit monetary incentives for contribution [1]. Since 2016, a handful of large corporations (including Apple, Facebook, Microsoft, and Uber) have increasingly contributed data to OSM. Corporate editors (CEs) represent a distinct community as their editors are compensated and thus their contributions cannot be labeled as ‘volunteered’. Additionally, corporations employ large editing teams and new state-of-the-art editing techniques aided by artificial intelligence, making them capable of editing large swaths of information in relatively short time [2]. Corporate teams are often led by long-time OSM community members themselves, emphasizing the multifaceted nature of a rapidly growing open mapping platform [3]. While there has been some contention about the quality of edits done by CEs, corporations argue their contributions improve existing data [7]. Our study provides a preliminary quantitative evaluation of data quality impacts of corporate edits on OSM. We assess intrinsic data quality across five regions that have high levels of corporate contributions: Dallas-Ft. Worth, Egypt, Jamaica, Thailand, and Singapore. The quality of these regions is compared to that of Denmark, a region which has witnessed relatively less corporate interest, yet possesses a well-mapped OSM presence due to a well developed local mapping community [4]. These evaluations were performed using measures of intrinsic map quality. While the most straightforward evaluation methods involve comparing against extrinsic sources, such as either ground reference information or authoritative data sources; lack of data availability, licensing terms, and costs often render this comparison untenable [5,6,7]. A transferrable, data driven way of assessing quality remains using Intrinsic Quality Indicators (IQIs), a sub-field of OSM analysis which provides a variety of approaches for evaluating intrinsic OSM data quality. We chose to focus on IQIs that apply to networks, and to evaluate IQIs for land-based transportation networks within OSM. We analyzed networks for our specified locations for every other year between 2014 and 2022. OSM editing archives were processed using R to extract maps of the relative activity of corporate editors [8]. Our list of corporate editors was sourced from OSM’s publicly available list of corporate editors accounts. We extracted entire networks that represented the first day of each year of interest (2014, 2016, 2018, 2020, 2022) from OSM’s historical archives. For the purposes of this study, we extracted all networks where “OSM WAY = Highway”. We evaluated several IQIs for our areas of interest. We focused on completeness of network, both in terms growth over time and in terms of its navigability. We operationalized “completeness for navigability” as an intrinsic measurement by exploring the percentages of networks that possessed attributes necessary for GPS navigation – street names and speed limits. Navigability was assessed and compared across time points using Origin-Destination matrices. By creating a regular matrix across the area and calculating the ratio between a direct route between points, and a route navigated within our network, we calculated a ratio that can be compared across time to evaluate the changing efficiency of the navigable network. Additionally, when building routing networks, we discovered an additional IQI : the presence and qualities of topological islands within our network. That is, areas which are disconnected from the main network due to mapping errors or incompleteness. After mapping these metrics, we analyzed how they correlate with each other and how they change over time. Overall, IQI trends for the road network reveal consistent patterns across all measures and locations. There is a trend towards increasing data quality in terms of gradual increase of network length, completeness in terms of attributes (name, speed limit, and pedestrian access), the increasing efficiency of ODM routing ratios, and the increasing amount of places that have “navigable” attributes. Importantly, we found differences between our control location (Denmark) and our other areas of interest. The primary difference of note is not with regards to the quality of the data, but with respect to the rate at which data quality improves: Denmark’s rate of quality improvement is slower than other locations. The faster rate of quality improvement in the test areas highlights that the data creation and editing activity by corporate editors and other organized editors in these locations are helping narrow gaps in data quality. While this presentation highlights the trends of data quality increase, it does not tease apart the quality assessment of contributions by corporate teams versus other mapping groups. As a crowdsourcing platform, data in OSM is co-produced by repeated editing of data objects by different members of the community [9]. The appearance of CEs in OSM represents the arrival of another community of ‘produsers’ in the OSM ecosystem, and thus a new evolution in its overall trajectory [2,10,11]. Consequently, there is significant interaction between CEs and non-CEs in data co-production in OSM, further reinforcing the idea that OSM is a ‘community of communities’ [11]. Each location has their own patterns regarding editing communities, what they edit, and the sociopolitical and economic ground truth in the real world. Each of these factors impact the data, and may make comparing editing patterns difficult, especially given the diversity of motivations both with CE communities and within other OSM communities. Hence, we do not try to pry apart the differences in trends between individual countries. Instead, we focus on the overall trend between our test and control locations. With these caveats, we find that the quality of the network has increased in these areas across all tracked metrics at a faster rate than it has in areas with low rates of corporate edits, indicating that corporate editing may have a positive effect on the overall quality of the map. about this event: https://2022.stateofthemap.org/sessions/EZPVPB/

This pilot project is connected to a larger initiative to open-source the assisted mapping platform for Humanitarian OpenStreetMap (HOTOSM) based on Very High Resolution (VHR) drone imagery. The study test and evaluate multiple U-Net based architectures on building segmentation of Refugee Camps in East Africa. Introduction Refugee camps and informal settlements provide accomodation to some of the most vulnerable population, the majority of which are located in Sub- Saharan East Africa (UNHCR, 2016). Many of these settlements often lack up-to-date maps of which we take for granted in developed settlements. Hav- ing up-to-date maps are important for assisting administration tasks such as population estimates and infrastructure development in data impoverished environments, and thereby encourages economic productivity (Herfort et al., 2021). The data inequality between the developed and developing countries are often resulted from a lack of commercial interest, especially with the recent trend of corporate OSM mappers (Anderson et al., 2019, Veselovsky et al., 2021). Such disparity can be reduced using assisted mapping tech- nology. To extract geospatial and imagery characteristics of dense urban enviornments, a combination of VHR satellite imagery and Machine Learn- ing (ML) are commonly used (Taubenböck et al., 2018). Classical ML based methods that exploit the textual (e.g. GLCM), spectral, and morphological characteristics of VHR imagery are based on the principles of Computer Vision (CV). Although many have shown promising results in satellite VHR (1m to 5m resolution) scenarios such as differentiating slum and non-slum (Kuffer et al., 2016 & Wurm et al., 2021), in VHR drone imagery (5cm to 10cm resolution) however, results might suffer from noise caused by environ- ment and drone-based specific problems such as motion artefacts and litter. Recent advances in CV based Deep Learning might be able to address these issues (Chen et al., 2021 & Carrivick et al., 2016). Purpose of the study The study is connected to a larger initiative to open-source the assisted mapping platform in the current Humanitarian OpenStreetMap (HOTOSM) ecosystem. This study is a pilot-project to investigate the capabilities of applying semantic segmentation using community open-sourced VHR drone imagery collected by the partner organisation OpenAerialMap. The study aims to rigourosly assess the various components and inputs that would contribute to the ML based mapping system, and to produce a detailed evaluation on class-based accuracy assessment (Congalton & Green, 2019). This pilot study focuses on 2 camps in East Africa, where data availability and the geography of the camps are within a similar savannah ecosystem. This enables highly-detailed method testing and analysis of transferability of the results between the two camps. Data and Methodology The first camp is located in Dzaleka, Dowa, Malawi, which is sub-divided into the Dzaleka North and Dzaleka main camp. The Dzaleka camps are home to around 40,000 refugees mainly coming from the African Great Lakes region. The Dzaleka North camp is characterised by a newer, spatially well- planned metal-sheeted roofs, while the southern main camp is characterised by complex, dense mud-walled building with stone-lined thatched-roofs (UN- HCR, 2014). The second camp, the Kalobeyei settlement is part of the sub- camp of Kakuma, located in the rural county of Turkana, North-West Kenya. The Kalobeyei settlement was home to approximately 34,849 refugees as of 2019. This camp is significantly more spacious and is characterised by spa- tiall well-planned metal-sheeted roofs (UNHCR & DANIDA, 2019). VHR drone imageries were provided for both camps and vector labels produced by HOTOSM volunteers were provided for the Dzaleka and Dzaleka North camp. Since CV based Deep Learning is very dependent on the quality of the labelled referenced data, especially when performing pixel-based semantic segmentaion, it is of crucial importance that care is taken when producing highly accurate labels that ensure sucessful training (Ng A., 2018). A large quanitiy of available labels did not have such a task in mind, imperfection in labelling around existing drone artefacts could cause the trained model to misclassify such pixels. In order to train a model which performs well on drone imagery, the motion artefact will be a signficant feature for the model to learn.he combination of data availability have allow a unique set of research questions concerning the input data quality and experiment setup to surface. Therefore, to test out U-Net and a few variation of the U-Net performance, an additional set of label data was created in order to supple- ment the imperfection in the labelled data of the Dzaleka camps. Initially, the models will be trained on the pixel-perfect and less complex Kalobeyei dataset, this will be then be followed by introducing the Dzaleka datasets of higher complexity. A comparison of baseline performance between the U- Net variations (Ronneberger et al., 2015) and the Open-Cities-AI-Challenge (OCC) winning model is conducted. The baseline experiement aims to keep the hyperparameters (e.g. optimiser, learning rate, weight decay etc.) con- stant to obtain an objective view of the architectual responses on the same dataset setup. This will provide a clear picture of the feasibility and how to take this project further, so that further resources could be justified to scale future experiments. Findings and Discussion Initial baseline experiments on the Kalobeyei dataset and Kalobeyei with the Dzaleka(s) seem to suggest limited transferability from the OCC model. This suggests that the OCC model is perhaps over-generalised to the compe- tition test dataset. Despite achieving very high confidence on metal-sheeted rooftops, it does not detect any of the more complicated thatched roofs com- mon in the Dzaleka camp. The OCC model also struggle with some of the more obscure drone motion artefacts occuring at the edge of the imagery in the Kalobeyei camp. Meanwhile, the Precision and Recall statis- tics favour other variations or further transfer training on the OCC model, and the EfficientNet B1 header U-Net pretrained with ImageNet weights. However validation loss suggests there might be little room for improvement in the further transfer training of the OCC model. Precision and Recall have both reached above 0.7 in most experiments, which outline the general capability of the strategies used. However there are still significant variations among different architectures and setups. The next step is to perform systemic fine-tuning to increase the confidence level of the appropriate architectures. about this event: https://2022.stateofthemap.org/sessions/FRJXCQ/