Welcome to the latest FOSS4G:UK

Personal & professional perspectives on the industry

CRAG (Collection and Reporting of Associated Geodata) is a QGIS plugin for geological mapping and field data capture developed by the British Geological Survey (BGS).

The plugin performs three main functions:
1) sets up any QGIS project for geological mapping work, complete with configured styles, forms and a GeoPackage for storage
2) provides tools for efficient data entry/editing and for photo imports
3) generates reports in HTML and PDF formats

CRAG projects use advanced features of QGIS forms, such as Relation References and Expressions, and of GeoPackage databases, such as foreign-key references and views, to create a more expressive data model than the flat table structures normally associated with GIS data.
This ensures that field data can be easily incorporated into BGS internal databases.

Mergin Maps provide an open-source system for synchronisation of QGIS project data between multiple users and for data collection on mobile devices such as phones and tablets.
By using QGIS-native tools and data structures, CRAG works seamlessly with the Mergin Maps system so that geologists can work collaboratively in the field and the office.
BGS worked with Mergin Maps developers, and contributed code, to enable the Mergin software to handle the database constraints that they require.

BGS geologists having been using CRAG in production since early 2025.
The public release in autumn 2026 aims to to simplify collaboration and data sharing with partners in the public, private and academic sectors in the UK and beyond.

The UK's Earth Observation Data Hub is now operational and ready to use with new tools for data explorers, analysts and developers. Join this workshop to get started using EO data from the hub by following along a practical example.

No experience of EO data or coding required! This hands-on workshop will get you started using EO Data Hub whatever your level of experience with Earth Observation data.

We will explore new features of the Data Hub that make data access and analysis easy, using a combination of open data sources.

Come and join in by following along the example workflows, then explore where you are most comfortable:

Explore, compare and export images in the hub's online Resource Catalogue
Play with the QGIS plugin
Work with the hub's Jupyter notebook service using ready-made notebooks to develop more customisation

By the end of the workshop you will be able to use the EO Data Hub for analysing landscape change using one or more methods that you will be able to take away and use in your day to day workflows for providing data driven evidence.

Across the world, millions of people remain “missing” from the maps that shape decision-making, emergency response, and resource allocation. This workshop introduces participants to the Missing Maps initiative and explores how open, collaborative mapping can help address these data gaps and support vulnerable communities.
Through a hands-on session, attendees will learn the fundamentals of mapping using OpenStreetMap (OSM). While a range of features can be mapped, including buildings, roads, and other key infrastructure, the workshop will focus on building digitisation, as this is typically the most accessible entry point for beginners.
Designed for both beginners and those with GIS experience, the session provides a practical opportunity to contribute meaningful data while gaining insight into the broader role of geospatial information in global development. Participants will leave with the skills and confidence to continue mapping independently.

IMPORTANT
Bring own laptop and a mouse (touchpad will make it much harder, so it isn't ideal).

Please see the resources to create your account before joining the session, so we have more time to focus on mapping.

Anyone who has tried to get their hands on geolocated address data in the UK will recognise the challenge. Navigating Ordnance Survey's catalogue is harder than it should be and even getting the right license can be difficult. At Rebalance Earth we use geolocated address data for identifying properties owned by pension funds and asset owners that are at risk of flooding so that we can propose nature-based flood management solutions to mitigate that risk.

As part of our flood risk reporting process, asset owners often send us a list of property addresses associated with individual commercial buildings, retail parks with several buildings, hospitals, or entire industrial parks. But for flood reporting we need polygons, and even the OS AddressBase data only gives us addresses as points. So, we had to find a better way of turning raw address strings into accurate, georeferenced polygons.

To do this, we built the Locator App: a CLI tool that stitches together several open-source geospatial datasets and uses a tiered address matching system to find or create a georeferenced polygon for every address in a list. The app works best for geolocating large areas with lots of buildings like retail parks, industrial estates, or commercial properties, rather than residential addresses.

This talk is about how we built the Locator App and why it matters beyond our own work. I want to share what tools we used, including Python, GDAL, pg_tgrm for fuzzy name matching, Claude for building the interface, and the open data we used such as OS Overture Places and OSM Land Use. I want attendees to leave inspired and thinking about how they can adopt an open-source approach to solving problems and break their own dependence on proprietary data.

Within this presentation Astun will be highlighting how Place and Property based open data can gain Intelligence and in turn what it takes to be able to visualise the data across the nation.

There are a variety of datasets (totalling in the millions) from Central Government and agencies such as Ordnance Survey, Office of National Statistics. Ministry of Housing, Communities and Local Government (MHCLG) and others.

The last two years have seen a period of unprecedented change for the GeoServer project with an intensive period of updates and developments. Following a successful fund raising campaign GeoServer 3 marks the completion of a long-planned modernization effort aimed at keeping the project aligned with the current Java ecosystem, while preserving the stability and backwards compatibility that users rely on. Now that GeoServer 3 has been available for a few months, this talk provides a final update on the work and its outcomes.

We’ll review the initial drivers behind the transition, starting from the upgrade of core dependencies, how
that cascaded to more updates, resulting in a coordinated effort across multiple projects. From there, we’ll outline the main phases of the work: how the upgrade was organized, funded, managed new needs, and ultimately delivered.

GeoServer 3 introduces a modernized foundation while maintaining the familiar concepts and workflows that users rely on. This session offers a guided tour of the new release, focusing on what has changed in practice and how to approach the transition from GeoServer 2.x.

We’ll walk through the upgrade process, highlighting prerequisites such as updated Java and servlet container requirements, and what to expect when migrating existing installations. The session will also showcase the refreshed user interface, discussing its evolution and the improvements it brings in terms of usability and productivity.

In addition, we’ll explore the updated module structure, what is included in the core distribution, and how extensions are organized in GeoServer 3.

Large AEC (Architecture, Engineering and Construction) companies have the challenge of managing GIS software access for thousands of software users across multiple countries and business units, each with their own business priorities. Unregulated software growth can result in excessive costs, cyber risks and put up barriers to effective global collaboration. Over the last 2 years Arcadis has been on the journey to establish robust governance for QGIS use across the company, to reduce risks and improve productivity. This work has also prepared the way for QGIS to be scaled more extensively across the company to mitigate vendor lock in and reduce undue friction and costs in project delivery.
This presentation will highlight the software governance challenges that AEC companies face in relation to open-source GIS and how these have been addressed at Arcadis as part of a long-term plan to scale QGIS. The presentation will also cover practical topics such as stakeholder engagement and the impact of AI on effective corporate open-source GIS software management.

Offshore wind is a key player in our fight against climate change, but can it also deliver wins for the environment? Here I explore how the use of multi-criteria spatial heatmap analysis in GIS can support the development of offshore Artificial Nesting Structures (ANS) as a strategic compensatory measure for auk species, to aid delivery of offshore wind projects. By combining ecological, engineering and consent risk datasets, the methodology provides a practical framework for site selection that can also be applied across a range of planning and design challenges. The outcome of the study identified key areas where an offshore ANS for auk species could feasibly be constructed, alongside providing further evidence to bridge known data gaps.

From midnight on the 1st of January 2028, all drones in the UK weighing over 249g will have to broadcast their location, the location of the remote pilot, and the unique ID of the drone's owner. This information won't just be available to the police, but to anyone with a mobile phone and a free app. Why is this becoming a requirement? What are the technologies behind this? And what could possibly go wrong?

Abstract

Historical maps are valuable resources for studying long-term landscape, settlement and environmental change, yet transforming scanned sheets into analysis-ready GIS data remains challenging. Using materials from the MAHSA project (Mapping Archaeological Heritage in South Asia), participants will work with digitised Survey of India (SoI) map sheets (1905-1950s) using hands-on exercises. MAHSA is an ongoing collaborative documentation project at the University of Cambridge, funded by Arcadia. With more maps being made available online, navigating and applying historical maps in geospatial research and documentation can be highly complex. This workshop provides a practical end-to-end, open-source workflow to access, transform, visualise, and organise historical maps using some of the latest geospatial technologies, such as IIIF, STAC, Cloud Optimised GeoTIFFs (COGs), Arches, and QGIS.

The workshop is structured around three short hands-on exercises. 1) Participants will georeference a digitised map sheet from the MAHSA Arches database, an open-source geospatial database platform using the related IIIF URL from library repositories. 2) Participants will subsequently explore a STAC catalogue of publicly released MAHSA maps, filtering by map date and extent before loading selected Cloud Optimised GeoTIFFs (COGs). 3) Participants then learn how to remove sheet margins using the MAHSA SoI Map Clipper (an internally developed open-source QGIS plugin).

The workshop concludes with a discussion of broader applications for historical map workflows and applications. Within MAHSA and beyond, historical maps support applications such as automated feature extraction, long-term landscape and settlement analysis, and many other forms of spatial analysis. In the discussion, we invite participants to provide feedback on how the presented tools might support different research and public-facing uses. As well as reflect on the challenges of IIIF and STAC to build sustainable infrastructure for the use of historical maps that can serve researchers, organisations, commercial users and the wider public.

Prerequisites

A laptop with a browser, an internet connection, and QGIS 3.44 (LTR) installed. Beginner-friendly, basic QGIS familiarity is assumed. No prior experience with IIIF, Arches or STAC is required.

The Atlantic Forest extends from northeastern Brazil along the Atlantic coast and inland into northeastern Argentina and eastern Paraguay, encompassing 15 ecoregions that include tropical and subtropical rain forests, coastal forests, mangroves, and Araucaria moist forests. Ranking among the world’s top biodiversity hotspots, the forest also plays a vital role in climate regulation. Once spanning more than 350 million acres, the Atlantic Forest has lost ~88% of its original vegetation to human activity, leaving it one of the most endangered biomes on the planet. Although much of this deforestation occurred over the last century, the process began with the colonisation of the region by the Spanish and Portuguese during the 16th–19th centuries, with early colonial settlement patterns, commodity frontiers and and key transport routes influencing the spatial distribution of deforestation in the region.

Recent studies have begun to recognise contemporary Indigenous territories as critical spaces for forest conservation, but less attention has been paid to the historical impacts of colonial-era Indigenous resistance on present-day deforestation patterns. To address this gap, we first generated forest cover maps for the Atlantic Forest from declassified CIA satellite imagery (c. 1965) and collated a database of geolocated sites of historically documented Indigenous resistance (c. 1500-1800) events from national archives. Next, we calculated the impact of Indigenous resistance on probable colonial routes through the forest using a least cost path based approach that incorporated topography, river direction and flow speed, the presence of rapids and waterfalls, and the cost of portage. Finally, distances from both sites of Indigenous resistance and disrupted colonial routes were incorporated into a an explanatory spatial Random Forest model as predictors, alongside other key anthropogenic and environmental covariates. Model evaluation revealed that Indigenous resistance was the top-ranked predictor of forest survival in the region and a counterfactual simulation predicted that resistance reduced forest cover loss by 20 to 22% between the colonial period and the 1960s. These results highlight the historic importance of Indigenous resistance and indicate that researchers and policy makers should allocate more attention to the conservation effects of historical Indigenous actions.

This interactive session will introduce you to eoAPI, a powerful cloud-native framework that simplifies access to Earth Observation data. By the end of this workshop, you'll understand how to use eoAPI to catalog, discover, visualize, and analyze geospatial data efficiently.

Workshop Objectives:

• Explore how eoAPI can fit into your geospatial workflows
• Learn about STAC (SpatioTemporal Asset Catalog) and its role in Earth observation
• Explore the key components of eoAPI and how they work together
• Gain hands-on experience working with metadata, raster, and vector services

Link to the eoAPI project: https://eoapi.dev/

OpenStreetMap data is used in many of our projects, and many of us have experience with contributing as volunteer mappers too. But how does this work behind the scenes? How is the data stored, how do the APIs work, and who writes the core software that powers OpenStreetMap?

We'll look at the database used to store the data, the software that runs the website and powers the editing API, and how the data is distributed. We'll learn why OpenStreetMap doesn't use PostGIS, how you can receive all the OpenStreetMap edits every minute, and who the volunteers maintaining all of this actually are.

Andy is a volunteer maintainer of the openstreetmap-website codebase, and he was previously a member of the OSMF Operations Working Group. Andy is also the founder of Thunderforest, which provides detailed basemaps built on OpenStreetMap data for web and app developers.

Power BI is one of the most widely deployed business intelligence platforms in UK public and private sector organisations, yet its built-in and freely-available map visuals offer poor support for the British National Grid, UK-specific address locators, or ONS administrative boundaries - the everyday currency of UK geospatial analysis. The OS Power BI Map Visual was built to close that gap.

Under the hood, the visual is a TypeScript application built almost entirely on open-source geospatial libraries: Leaflet.js for rendering, Turf.js for spatial operations, proj4 for coordinate transformations (including OSTN15 datum shift between ETRS89 and OSGB36), and shpjs and topojson-client for client-side geometry import and parsing. It geocodes tabular data using UK-native identifiers - postcodes, UPRNs, Eastings/Northings, and GSS statistical codes - the last of these automatically resolving to the corresponding ONS boundary (LSOAs, MSOAs, wards, local authorities, and many more) by querying the ONS Geoportal directly, with no boundary file management required from the user.

The visual supports choropleth and point maps, custom geometry upload (GeoJSON, WKT, TopoJSON, shapefile), and two-way cross-filtering with other Power BI visuals. All of this runs inside Power BI's sandboxed custom visual API - a constrained environment that creates its own engineering challenges.

Ordnance Survey has released the visual as open source under an MIT licence. In this talk we will show what it does, sketch the FOSS architecture that makes it work (as well as the places where non-FOSS tools are used), discuss design decisons we made, present a live demo, and invite the community to use, fork, and contribute.

The IFRC GIS Training Platform is a collaborative and open source platform developed by Red Cross societies in collaboration with the Heidelberg Institute for Geoinformation Technology (HeiGIT). The platform provides a modular and hands-on GIS training curriculum using QGIS and specifically designed to match and support humanitarian use cases. The training programme enables learners with basic information management or data skills—but little or no prior GIS experience—to start creating, using, and analysing spatial data effectively.

The platform includes interactive tutorials, real-world data exercises and practical examples that resonate with humanitarian professionals. The content is designed so that anyone can use the materials to run their own trainings without relying on a central team. This model has already been adopted by several National Red Cross and Red Crescent Societies, resulting in a growing network of GIS-trained humanitarian practitioners who are also empowered to pass on their knowledge to others.

This talk will outline the project’s aims and progress so far, and will be of interest to anyone involved in GIS training and skills development, or in humanitarian GIS workflows.

Applying deep learning to large-scale geospatial imagery presents a fundamental trade-off: high-resolution datasets are simply too large to analyse in a single pass. While downsampling preserves broader spatial context, it sacrifices fine-grained detail. Conversely, slicing imagery into discrete tile grids introduces arbitrary boundaries that bisect objects and isolate features from their surroundings. Consequently, downstream detection and segmentation models receive truncated context, compromising inference quality. While overlapping tiles offer a partial mitigation, they merely shift the burden to complex post-processing required to resolve conflicting predictions.

In this talk, we present a novel solution to bridge tile boundaries using the latest vision foundation models like DINOv3. DINOv3 processes raw imagery into uniform token grids, where each patch embedding captures rich, high-dimensional spatial context without relying on human annotations. By comparing these embeddings, semantically and spatially coherent patches can be clustered across an image. We demonstrate how this unsupervised clustering capability extends beyond individual tile boundaries to encompass arbitrarily large regions. This approach reduces reliance on tile overlaps and lossy downsampling, equipping downstream detection and segmentation models with the unbroken context needed to accurately delineate complex geographical features.

To highlight its practical value within geospatial workflows, we demonstrate our Pytorch pipeline and showcase its performance against a diverse range of satellite and aerial imagery datasets.

Ordnance Survey (OS) has captured aerial imagery of Great Britain for decades to support the maintenance of its national datasets. Once new imagery is captured, it is processed (Change Detection and Automated Feature Extraction), and then carefully archived and preserved (i.e. locked in a cupboard). The latest imagery is used to support existing OS products, but the older imagery is left alone. These image archives contain a visual record of how the UK has changed over time which could be very valuable!

This talk explores how this imagery archive is currently preserved within an open-source Fedora Commons repository, and how it could evolve into a discoverable geospatial asset using the OGC standard STAC.

This is the Annual General Meeting of the UK chapter of OSGeo.

For more than four decades, the geospatial community has been building the digital foundations of our world. Open source projects such as GDAL, PostGIS, QGIS, PDAL and OpenStreetMap have transformed how we capture, process, share and understand geographic information. They are the foundations on which an entire industry has been built.

But we have reached a new frontier.

We have become exceptionally good at representing the world. We can map it, model it, render it and query it. Yet representation is not understanding. A coordinate is not a place. A building footprint is not a building. A point cloud is not a physical environment.

A place has context, history, relationships, purpose and meaning.

As AI systems move from answering questions to taking actions in the physical world, the next generation of geospatial technology faces a fundamental challenge: how do we give machines a sense of place?

Foundation models have demonstrated remarkable capabilities, but intelligence about the physical world requires more than language. It requires spatial memory, continuous observation, simulation, context and reasoning. The future will not be defined only by larger models, geometries or attributes, but by better representations of the world: representations that capture context, relationships, change, history and meaning of its existence.

In this keynote, I will explore the transition from geospatial data to spatial intelligence. Drawing from lessons building a multi-million-pound geospatial technology company on the shoulders of open source, I will challenge this awesome community:

What is the equivalent of GDAL, PostGIS or OpenStreetMap for spatial reasoning?

How do we move from mapping the world to understanding it?

What role should the open geospatial community play when the next generation of GIS users may not be human?

The next era of geospatial is not about creating more maps or more AI. It is about creating systems that understand places.

The question is: who will build the open foundations for machines to understand our world?

AI is changing how software is created, shared, maintained and used, and potentially what we mean by “open” itself. For geospatial FOSS, the implications extend well beyond AI-generated code. This talk looks into some of the less-recognised medium and longer term implications and challenges that AI introduce. We discuss how AI could reshape reproducibility and provenance, lower barriers to software development, disrupt traditional pathways for learning and entering the field, and transform how open-source communities collaborate and contribute. At the same time, it creates opportunities to make geospatial software more accessible, adaptive and powerful. This talk explores both sides of this transition and asks a broader question: in a world where machines can increasingly write the code, what should remain distinctly human about open source?

Leeds is a friendly, diverse, international city.

But getting here can be awkward.

If you were going to build a global high-speed transport network to serve the needs of the students and workers of the People's Republic of West Yorkshire, what would it look like?

The Sentinel programme changed what many of us expect from satellite data: open access, regular coverage, and observations we can actually build services around. The next Copernicus Expansion missions take that idea further, adding measurements we do not currently get from the core Sentinel fleet: hyperspectral land imaging from CHIME, L-band radar from ROSE-L, anthropogenic greenhouse-gas monitoring from CO2M, high-resolution land surface temperature from LSTM, microwave radiometry for ocean and polar monitoring from CIMR, and polar ice and snow altimetry from CRISTAL.
Expansion missions fill important gaps: richer spectral information for soils, crops, minerals and habitats; longer-wavelength radar for vegetation structure, soil moisture and ground motion beneath forest cover; field-scale thermal data for water stress and evapotranspiration; direct observations of CO₂ and methane plumes; improved sea-ice, ocean and salinity products; and continuity of polar ice thickness, snow-depth and sea-level evidence.
For the UK, these missions could strengthen evidence for food security, water management, nature recovery, flood and coastal resilience, net zero monitoring, climate adaptation, maritime safety and Arctic policy. For the FOSS4G community, the message is simple: a new wave of open Earth observation data is coming, and it will be far richer, more varied and more policy-relevant than anything available today. Be ready. These missions are a huge opportunity to turn open satellite data into better services, better decisions and wider public value.

One of the Open Source advantages is the ease of combining tools that perhaps initially were quite separate. In this workshop you will learn the various ways that QGIS and R can work together to apply statistical methods to your GIS or add GIS methods to your statistics.

After an introduction into how these two systems can integrate on different levels, we'll demonstrate the packages and plugins that allow users to quickly work across the two applications. This includes running QGIS processing algorithms from R, and calling R code from QGIS, making the workshop useful for people who consider themselves primarily either QGIS users or R users.

A number of practical exercises will be prepared for attendees to work through with help and guidance.

For those who don't know any R, or even have never used it, exercises will include complete R code files which we'll show how to add to QGIS to enhance its processing capabilities.

For R coders, exercises will also cover the R qgisprocess package enabling QGIS algorithms to run from R, and also the R Console plugin which runs R within a QGIS window.

Attendees are encouraged to bring their own data and problems which we can either work through as a class or one-to-one with the workshop team.

Installing the latest version 3 of QGIS and a recent version 4.xx of R before the workshop will save time. QGIS 4 is not well supported at this time.

There has been a lot of progress in the capabilities of the 3D map visualization within QGIS. With support for various 3D data formats, QGIS 3D is moving far beyond the simple viewer when it was first released in QGIS 3.0 in 2018.

This talk summarizes the work that has been done in 2025 and 2026 to make QGIS 3D faster, more powerful and easier to use. We will discuss support for ESRI’s scene layers (I3S / SLPK), improvements to 3D Tiles support, significant performance improvements for handling of vector layers and much more!

In 2025, Lutra Consulting and North Road ran a successful crowdfunding campaign to push QGIS into the world of geospatial digital twins. The goal was to transform QGIS 3D from a capable but limited viewer into a credible open-source platform for urban planning, infrastructure management and environmental simulation. With the funded work now delivered across QGIS 4.0 and 4.2, this session reports back on what was built, what it looks like in practice, and what it means for anyone working with 3D geospatial data.

What can the history of computer chess teach us about the future of geospatial development? As AI fundamentally alters how we develop GIS software, write spatial algorithms, process imagery, and automate workflows, our community is currently struggling with existential and practical questions that the chess world has been navigating for decades. By looking at how chess players adapted to, survived, and ultimately thrived alongside super-intelligent machines, this session discusses options for FOSS4G professionals looking to navigate the new rules of the GIS game in the age of AI.

The history of mobile libraries in the UK goes back to the 1920s. Although we don't have official counts of how many mobile library stops there are, there are likely to be between 15,000 and 20,000 in the UK, largely serving rural communities without access to other library services. These provide essential access to reading material and more.

This talk will look at the history of mobile library routes and stop data, from the books still in use today that detail using pins and string on a large OS map. It will then describe creating a standardised open data schema for mobile library stops, routes, and schedules, and geocoding existing mobile library data, often only available in PDF files with stop names like 'Opposite the willow tree'. Finally the talk will demo presenting that data using open source tools like PostgreSQL and OpenRouteService and web mapping software to use that data in modern applications, and explore the future potential for widescale open mobile library data.

The Environment Agency provides openly-licensed LiDAR-derived DEM data covering the whole of England, but it is difficult to access this data. The web interface only allows downloads of small areas, manual bulk downloads produce hundreds of small TIFF files and the web map services have limited visualisation options. At Rebalance Earth, we needed easy access to large volumes of DEM data to analyse flood risk and design nature-based restorations to reduce flooding.

To solve this problem, we have converted the whole dataset to Cloud Optimized GeoTIFF (COG) and produced a STAC catalog linking to all these COG files. As the Open Goverment License allows redistributing the data, we’ve been able to make the new files and STAC catalog available online, free for anyone to use.

In this talk I’ll explain the benefits of cloud native geospatial data, and how we did the conversion (including some useful tips for creating COG files). I’ll then show some demos of how this data can be used – for example to easily produce mosaics of DEM data across arbitrary areas, to visualise in custom ways on a web map (tools like titiler and titiler-pgstac) and more. I’ll finish by explaining how you can connect to this dataset and access it yourself.

There is a need to create virtual worlds, for a number of different reasons, but for military synthetic training between distributed systems and platforms – it becomes essential. These virtual worlds consist of layers of geospatial data, building a rich immersive environment for the trainees. When consistency between simulators lacks, and correlation error creeps in, the result is HOVER TANKS!

Terrain correlation for distributed simulation is a known, and often occurring problem for participants in a distributed training event. Numerous approaches to ‘standardise’ the standard have been considered, including using open standards. Even with open standards, there still can be an issue of interoperability between different simulators.

For my MSc in Geographical Information Systems, I used FOSS4G tools to investigate the problem – understanding why simulators that employ open standards are not necessarily fully interoperable. The result of the analysis led me to understand the importance of co-ordinate systems, pixel resolution, tiling specifications and common data source – and more importantly the incredible power and utility of FOSS4G, and how organisations are missing a trick on superb software.

Yes, this may be Thesis presentation... but I've graduated now and I promise to make it fun. Plus... I had you at hover tanks right?

I emphasised the use of FOSS4G in my thesis research as a way for me to better understand their capabilities, their usage and to better advocate for their use in organisations where possible. I wanted to be able to argue for the use of FOSS4G when people just default to the common enterprise solution, and others argue them being expensive or cost prohibitive to do something for small scale problems / investigation. The software is out there, the knowledge to use it is available or trainable; and if the problem isn't solved in the open source space, it either isn't worth solving or you haven't started coding yet!

An important step in FOSS development is checking what your neighbours are doing. For generic geo-software, we usually do that and share what we have (or know). But for our Spatial Data Infrastructures, things tend to get more complicated. However, there is a lot to be gained by just getting some first hand experience in what the colleagues in other countries are working on (this is one of the reasons I love the FOSS4G events, and have been visiting both the international and local ones in other countries before).

In this presentation, I will share some of the Dutch government (both regional and national) projects that contribute greatly to our national SDI, and might be usable in the UK context as well. One of those is undoubtedly the PDOK (Publieke Dienstverlening Op de Kaart or Public Services On the Map) initiative which is closely related to (but not exactly the same as) our national data portal, and the integration that makes it easily accessible in QGIS (becoming by far the most downloaded QGIS plugin in the NL). Others include the combined 3D-initiatives that are now taking place.

The European Space Agency's "APEX" project provides a set of project tools and services designed to help Earth Observation community disseminate the outputs of EO based research and development work. One of these tools is the Geospatial Explorer - a cloud native interactive web map dashboard which practitioners can use to build their own solution using entirely no code environment.

With an Earth Observation focus, delegates will use the Geospatial Explorer Configuration Builder to create their own web map application. Lead by Sparkgeo (developers of the "GE") and accompanied by comprehensive workshop notes, attendees will:

• Gain a comprehensive insight into the capabilities of the Geospatial Explorer
• Build their very own bespoke configuration, featuring cool data visualisations, interactive charts and graphs, time series animations, data filters and constraints, interactive "storymaps" and more
• Work with and learn about cloud native geospatial data and services including a fascinating array of earth observation data and derived products
• Work with and learn about spatio-temporal asset catalogues - the metadata standard for EO
• Witness on-the-fly EO algorithm execution from the APEX algorithm catalogue
• Have fun

Prerequisites!

Here is the best bit. The workshop requires:
* NO earth observation experience
* NO software development experience
* NO pre-installed tools or environments on your machine
* NO detailed technical knowledge

The ONLY requirements are:
* A laptop and WiFi
* Enthusiasm
* Oh, and maybe some coffee

Research into physical, ecological and socio-economic changes that have occurred across landscapes relies on multi-disciplinary skills and information from a variety of sources including local knowledge. Although landscape history is an academic field backed by research groups and professorships in several universities, much of the action takes place in voluntary groups, composed of citizen scientists and local historians operating within more-or-less formalised communities of practice. Example local history and naturalist groups include the now-disbanded Plinian Society, founded in 1823 in Edinburgh and whose membership included Charles Darwin, and the Woolhope Naturalists' Field Club founded in Herefordshire in 1851, which still runs to this day.

A wide variety of established methods are available to landscape historians, ranging from desk-based collection, translation and synthesis of information from historic archives to the field-based observation of biodiversity and methods requiring scientific equipment including dendrochronology, ground-penetrating radar, and radiocarbon dating. The use of geographic datasets and methods such as remote sensing, georectification and digitisation of historic maps, and communication of results through interactive maps are relatively new developments that offer great potential to empower many current and potential landscape historians of every stripe.

However, the knowledge and skills needed to collect, work with, and gain insights from geographic datasets for landscape history are few and far between, with expertise concentrated among a few people working at the intersection between landscape history and the broadly defined field of geocomputation. David Lovelace (1948–2026) worked at this intersection and throughout his 30+-year career he collected more than 2,000 Gigabytes worth of data on the landscape history of Herefordshire. An advocate of empowering people through technology, he was a strong proponent of free and open source software for geographic applications (FOSS4G).

This talk will discuss the geographic research methods that David developed, with a focus on open tools he used and with reference to the nascent David Lovelace Archive (hosted at bosci.net), which provides an example of how new technologies can help get research out of dusty desks and into the wider world for the greater good.

Mergin Maps is a powerful field data collection app that lets you work collaboratively and take your QGIS projects into the field on your mobile device. But did you know that Mergin Maps is also making it easy to publish your QGIS projects on the web and share them with anyone? We've recently added new features that let users share their projects with others easily with a link to a web map or embed your project on a website with a generated code snippet you can copy and paste into the HTML of your website.

In this workshop, users will learn how easy it is to create a project in QGIS and upload it to the Mergin Maps Cloud. Then we'll collect some data, publish it and view it on a web page, all in under one hour!

Anyone can speak about anything - times for speaker slots dependant on number of topics to get through

Sometimes, our customers needed a different background for their maps. Sometimes, they didn't want any label on them. Sometimes, they really wanted those lines to be ticker, while those other ones to be red. Most of the time, they didn't know how to code to do implement those changes themselves.

That's when we, at Ordnance Survey, started working on the Map Style Editor. A no-code webapp to edit stylesheets live while interacting with the map, to then be able to export and/or share them.

Then, the Roadmap changed and Map Style Editor was shelved.

However, the idea was out there and I thought it was too good to simply let that be forgotten.
Enters Map Fashionista, my free open source take on it. It follows similar principle, albeit using OpenStreetMap data.

This talk is going to be about the journey I went through, from the inception of Map Style Editor to the rise of Map Fashionista.

Martin Eglin¹, Philip Midgley¹, David Tarrant² and Simon James²

¹JBA Consulting
²Natural Resources Wales (NRW)

NRW maintain a suite of flood risk datasets that support decision-making, investment planning and resilience across Wales. Two important datasets are flood defence lines, which denote the location and key attributes of assets on the ground, and defended areas, representing the geographical areas protected by those assets. Although closely related, these datasets are maintained independently, creating spatial inconsistencies that can affect downstream analysis.

Issues became apparent when calculating the costs associated with defences within national investment planning tools. Maintenance and operational costs are reported at the defended area level and individual defence lines frequently span multiple defended areas. Cost allocation therefore relies on spatial intersection analysis and unit-cost calculations. Where defended area boundaries are not correctly aligned with defence lines, costs can be under-reported, affecting economic assessments and future investment decisions.

To address this challenge, JBA developed an automated GIS workflow using Python. The workflow identifies defence lines that are not correctly connected to defended area polygons using near-table analysis and distance-based calculations. Polygon vertices requiring adjustment are identified automatically, and translation vectors are generated between those vertices and their nearest defence line. These vectors are then used to reshape defended area boundaries, improving alignment while maintaining user control through configurable tolerances and filtering options.

The workflow is executed from the command line and has significantly reduced the manual effort required to review and correct defended area boundaries. Quality assurance is supported through the creation of intermediary outputs that visualise every vertex translation, allowing analysts to review and validate automated changes before acceptance.

This presentation will describe the challenge, spatial analysis techniques used and lessons learned. It will also explore the balance between automation and manual intervention when maintaining large, interconnected geospatial datasets. Whilst some complex line-polygon interactions still require manual editing, the approach has substantially improved the data quality and reporting accuracy that underpins key investment decisions.

The European Space Agency (ESA) is developing a new unified Zarr-based storage and protocol format for Sentinel (1, 2, and 3) mission products under the Earth Observation Processing Framework (EOPF) initiative. EOPF Zarr enables scalable, cloud-native access to Earth Observation data and represents a significant shift in how EO products are distributed by ESA.

This talk introduces EOPF Zarr: its design, current status, and practical capabilities. We'll also cover the growing ecosystem of open-source tools being built around it such as plugins for other services (e.g. GDAL, Xarray, QGIS. OpenLayers, ...) to interactive browser-based exploration of Sentinel imagery with no downloads or preprocessing required.

Whether you're new to cloud-based EO workflows or want to learn about Zarr, this session will give you a clear picture of where EOPF stands today and how to get started with it.

At Ordnance Survey we create a lot of varied geographic data using many different tools. For a project on automatically generating lines representing the high and low tides we compared the relative effectiveness of one open source tool, GDAL, against a couple of non-open source ones. Apparently, GDAL is pretty good! (with some fixable caveats).

In this lightening talk I'll be demonstrating the use case for GDAL in the admittedly narrow context of generating tide lines, comparing some outputs with paid-for solutions, and generally considering the strengths of open source tools.

GIS is amazing, but there are many people who would like to create a map, who don't know GIS. gitRmap is an easy way of creating a simple map. It is designed for people who know a little bit of Git, but don't know any R or GIS.

International treaty participation shapes the geopolitical landscape in ways that are rarely visible to analysts and researchers. TreatyGraph is an open-source, serverless geospatial visualization platform designed to surface these relationships, built entirely from free and open components across every layer of the stack.
A Python data pipeline harvests treaty membership data from Wikidata via SPARQL, integrates bilateral trade statistics from public databases, and cross-references the Correlates of War dataset. This produces a Graphology JSON graph representing approximately 200 countries, 500+ international treaties, and thousands of membership edges — entirely from open, citable data sources with no proprietary dependencies.
The graph is consumed by a browser-based frontend that requires no backend infrastructure. MapLibre GL provides the cartographic base. Deck.gl renders WebGL arc layers connecting co-signatory countries, with colour and thickness encoding trade balance, treaty count, or treaty category. D3.js drives animated donut charts, hierarchical edge bundling, and comparative bar charts that update dynamically as the user navigates. Sigma.js, built on the Graphology graph library, powers a force-directed network view enabling bilateral treaty comparison between any two countries.
The complete pipeline — from SPARQL query to rendered map — relies exclusively on open-source libraries. The output is a single static HTML file paired with a ~5 MB JSON graph, deployable on GitHub Pages, object storage, or any static host.
This talk presents the full architecture: how SPARQL queries are structured to extract treaty participation from Wikidata, how the Python graph builder normalizes and enriches the data, and how the visualization layers are composed to support two analytical modes — a country-centric co-membership view and a bilateral shared-treaty comparison.
The underlying pattern — Python preprocessing into a static graph consumed by a performant browser frontend — is broadly transferable to any domain where open geospatial data can be pre-aggregated. TreatyGraph demonstrates the analytical depth achievable within the open-source Python geospatial ecosystem, without a single proprietary dependency or a persistent server.

QGIS provides a powerful open source environment for managing, editing, and visualizing geospatial data. However, collaboration around QGIS projects often remains fragmented. Production teams need centralized data storage, reliable synchronization between desktop and web environments, controlled multi-user editing, conflict-resolution tools, and a clear record of changes.

This presentation introduces an open source workflow for collaborative geospatial data management based on QGIS, NextGIS Web, and the NextGIS Connect plugin.

NextGIS Web acts as the server-side platform for storing, managing, and publishing geospatial resources. It provides granular access permissions and built-in version control for vector data.

NextGIS Connect links the desktop and server environments. The plugin allows users to publish QGIS projects as web maps, open existing web maps as QGIS projects, edit server-hosted data from QGIS, synchronize changes, and resolve concurrent-editing conflicts interactively.

Recent developments extend this workflow to feature attachments, including photographs, documents, and other files associated with spatial features. Attachments can be managed consistently through both QGIS and the web interface, while their additions, modifications, and deletions are recorded alongside geometry and attribute changes.

The talk will demonstrate practical multi-user scenarios and explain how synchronization, permissions, conflict resolution, and version history work across the stack. It will also show how organizations can evolve from isolated local QGIS projects toward a shared, auditable, and fully open source collaborative GIS environment.

This talk will demonstrate how to give AI tools access to your geodatabase, so colleagues can run geospatial analyses on your data using tools like Claude and ChatGPT. I work for Rebalance Earth, which requires large amounts of data to make decisions about investing in nature-based projects. These data are stored in a large PostGIS database. Frequently, less technical colleagues want to be able to run their own analyses using AI. However, we want to make sure that this is done in a safe and secure manner.

We initially developed a custom chatbot, however as the use of AI increased in the company it made sense for the app to migrate to an MCP server that any agent could access. I will start off with covering how I developed the initial toolbox to access our PostGIS database, getting the LLM to understand the structure and schemas of our geospatial data – including dealing with co-ordinate systems sensibly. Then I will talk about the move to an MCP server and the challenges faced with this. Topics will include: keeping the whole system safe and secure, ensuring that enough guidance was given to allow AI tools to produce the correct answer, dealing with large volumes of data without filling the AI model context, deploying using Azure Functions, and the quirks of developing MCP servers locally. I will hopefully be able to include live demos of using the MCP server.

The MapTiler Plugin for QGIS is now over 6 years old and, with well over 500,000 downloads, is in the top 30 QGIS plugins. This talk reveals the plugin's history, functionality, and future.

I'll show the great cartography achievable by combining the power of open-source tools and data, and discuss the business benefits of creating open-source components.

Finally, the talk will cover MapTiler's commitment to open source through the MapTiler Open Foundation. Including the donations to OSGeo and the wider FOSS4G Community, and who we are looking to support in the future.

Maps are stories. But too often, critical places and entire communities are missing from the map and their stories are not told. At the Humanitarian OpenStreetMap Team (HOT), we envision an ecosystem of open mapping technology that enables everyone, and in particular vulnerable communities, to make open map data available in order to use in disaster response and humanitarian context.

In this talk, I would like to take you on a journey in introducing the full end-to-end open mapping workflow and the open source tools enabling that process - from the Drone Tasking Manager , Open Aerial Map to fAIr, Field Tasking Manager, ChatMap and uMap. I will share the stories where the end to end mapping workflow has been used in Indonesia, Sierra Leone and more recently in response to the earthquake in Venezuela.

We hope that you will leave this talk inspired and with an understanding on how YOU can interact with the open mapping tools and create your own story!

Maps are one of the least accessible parts of many public digital services. Even when a web map meets basic interface guidelines, the spatial information inside it is often unavailable to screen-reader users, keyboard-only users, people with low vision, cognitive accessibility needs, or users who simply cannot interpret dense cartography.

This talk asks a different question: what if the map is not the primary interface?

Using open source geospatial tools, we will explore patterns for making spatial information available through a voice enabled interface, GDS compliant UI, text, tables, keyboard navigation, structured summaries, audio-friendly descriptions and simplified map views.

The talk will show how accessibility can be treated as a geospatial problem, not just a front-end compliance task. Attendees will leave with practical patterns for building services where everyone can understand the where, not only those who can see or operate a conventional map. We will include a demo where one will be able to ask generic questions and hear the map provide the answer.

Postal address parsing has two well-worn approaches: hand-tuned rule engines that break down outside their home locale, and large language models too heavy to ship to a browser. Mailwoman is a compact neural sequence labeler fed by reference data (postal registries, gazetteers, postcode indexes) supplied to the model as soft evidence rather than hard overrides.

The result is a full geocoder. A shared ~40 MB int8 ONNX model (SentencePiece tokenization; BIO decoding with transition masking and Viterbi) covers every locale, and each additional locale ships as an overlay of a few hundred kilobytes. The same weights run under Node.js and in the browser over WebGPU or WASM, next to a SQLite-WASM gazetteer, so the entire parse-then-geocode path can execute client-side. For existing tooling, Mailwoman exposes drop-in Nominatim, Photon, and libpostal-compatible APIs.

This talk covers what we learned shipping it: how we ruled out the rules-engine, and how OpenStreetMap, Overture, Who's on First, and national open-address registers (BAN in France, LINZ in New Zealand, TIGER in the US) feed the training and resolution pipeline. We close it off with the recipe for the en-GB build, and how the same pipeline can be used to add any locale with a gazetteer and a labeled corpus.

Attendees will leave able to evaluate whether their own address pipeline could benefit from a neural parser, and with a concrete recipe for adding a new locale.

A trailer for Bristol 2027, why it's a big deal and how you can get involved.

Thank you to everyone, and let's go for 2028!