We are happy to announce 7 pre-conference workshops, which will be held on June 3 afternoon, starting from 13:30.
Note: We are happy to announce one more workshop “W7 (Hands-on and Discussion) – Advancing Feature Extraction and Map Quality Analytics through GeoAI and HERE Mapmaking“, which will be organised by Johannes Lauer from HERE Technologies.
W1 – State of the Art and Outlook of Street-Level Imagery for Urban Science
Organisers: Filip Biljecki, Xiaobing Wei, Pengyuan Liu, Weiming Huang, Fangli Guan, Stephen Law
Call for submissions: Please submit by e-mail to Xiaobing.Wei@Ugent.be before 30 March 2026. More details can be found in the PDF.
Goals and scope: Street-level imagery has become an important data source in urban research, offering a rich ground-level view of urban environments and capturing the visual and experiential qualities of streets and neighbourhoods on a scale. Recent advances in computer vision and geospatial artificial intelligence (GeoAI) have enabled the transformation of street-level imagery into structured and interpretable urban features, supporting applications across human perception, urban planning, transport, equity, public health, and related fields. This workshop explores how street-level imagery can advance human-centred understanding and fine-grained modelling of cities, supporting applications such as Image-based 3D Modelling, foundation models for analysing streetscapes, vision-language urban agents, human-in-the-loop urban digital twins, cross-view learning, generative place reasoning, and decision-making in the urban context.
W2 – GeoAI for Disaster Response: From research to practice
Organisers: Sam Leroux and Lisa Landuyt
Workshop website: https://geoai4dr.ugent.be/
Call for presentations: You are invited to submit an extended abstract (600-1200 words) by 20 April 2026, following the guidelines at the workshop website.
Goals and scope: By integrating geospatial data with artificial intelligence, GeoAI can enhance situational awareness, decision-making, and response efficiency in disaster contexts. This workshop explores recent advances in machine learning, remote sensing, and spatial analytics for hazard monitoring, damage assessment, and emergency response during man-made and natural disasters. By bringing together researchers, practitioners, and policymakers, the workshop aims to highlight emerging methods, practical applications, and open challenges in leveraging GeoAI for timely, scalable, and resilient disaster response.
W3 – Building Open Labs and Learning Activities for “Introduction to Geospatial AI”
Organisers: Meiliu Wu, Siqin Wang, Song Gao, …
Short description: A highly interactive, education-focused workshop organised as a “Teaching Exchange + Lab Gallery Walk + Assessment Clinic.” The core activity is a Lab Gallery Walk, where participants rotate across five themed tables (each hosted by a facilitator) to review and refine reusable lab blueprints, datasets, and teaching notes. Together, we will assemble an open introductory GeoAI teaching pack, including a reading list, a lab catalogue, assessment templates, and a responsible GeoAI checklist for classroom use.
Goals and scope: 1) To identify a shared “minimum viable” curriculum for GeoAI education that bridges AI and geographic thinking (e.g., scale, spatial dependence and heterogeneity, and uncertainty); 2) To share and refine ready-to-teach labs/activities based on multimodal data (e.g., imagery, vectors, and text) and deployment realities (e.g., data standards and reproducibility). 3) To co-create practical assessment and integrity guidance for GeoAI teaching (e.g., rubrics, spatial generalisation tests, and ethical use of generative AI); 4) To publish an open, community-maintained teaching resource pack after the workshop (e.g., a GitHub repository with a Zotero reading list and editable templates).
W4 – The dark side of GeoAI: mapping the risks and dangers of applying AI in the geo domain as well as potential countermeasures
Organisers: Chris Kray, Sigrid Kannengießer, Eftychia Koukouraki, Milad Malekzadeh, Rainer Mühlhoff
Workshop website: https://www.uni-muenster.de/Geoinformatics/en/sitcom/darkside.html
Call for participation: People interested in participating are asked to submit a simple text file containing a one-paragraph statement of interest (why they want to participate, what aspects they are particularly interested in, what expertise they bring to the discussion) and a suggested topic for the small group discussion in
the first half of the workshop. Send your submission to scala24@uni-muenster.de with the subject line “Statement of interest: DarkSide Workshop” by March 13.
Goals and scope: The goal of the workshop is to map out the risks and dangers of AI in the geo-domain, and to identify and map potential countermeasures for these risks. We plan to focus specifically on generative AI and machine learning as these AI technologies are already used in the geosciences and rapidly developing. Risks include, for example, using generated geodata for training, inferences or decisions, biases and hallucinations as well as the potential harms of geomedia that contain mis- and disinformation. Taking generated results at face value and ecological consequences of AI use constitute further risks.
We plan to create and discuss an initial set of risks, dangers and countermeasures at the workshop, working as a group to select a consistent subset to investigate more thoroughly at the workshop. Based on this initial investigation, we then plan to develop this into a more robust framework and (potentially) a journal paper afterwards. We consider this an essential endeavour for GeoAI that will help researchers and practitioners to develop useful and robust GeoAI tools and to responsibly create and use them.
W5 (Hands-on) – Operational GeoAI for Scalable Satellite Analytics with Foundation Models on EuroHPC/MeluXina supercomputer
Organiser: Eun-Kyeong Kim
Goals and scope: Recent progress in geospatial foundation models has accelerated Earth Observation (EO) research, yet many GeoAI methods are still evaluated at limited scale, obscuring issues of generalization, reproducibility, and operational readiness. This workshop provides a research-oriented but operationally realistic introduction to scalable GeoAI workflows, combining open EO foundation models with supercomputing resources such as EuroHPC/MeluXina.
Focusing on large-scale land cover segmentation from Sentinel-2 imagery, the workshop adopts a hybrid format. Participants first run lightweight inference of EO foundation models on their own laptops using small, pre-packaged data samples to build model intuition. The workshop then transitions to an instructor-led live demonstration on MeluXina, showing how the same workflows work at scale.
Target audience: The workshop targets GeoAI and remote sensing researchers, applied AI researchers, and PhD students. It is also relevant to R&D engineers from the space sector seeking research-aligned, scalable methodologies. No prior HPC experience is required; basic Python and deep-learning familiarity is assumed.
W6 (Hands-on) – Building GeoAgents: A Hands-on Guide to Agent Protocols
Organiser: Shoaib Burq
Goals and scope: As agentic AI systems become increasingly central to geospatial applications, understanding emergingprotocol standards is crucial for building interoperable, reliable, and reproducible GeoAI systems. Thisworkshop advances knowledge and understanding of protocol-driven GeoAI architectures throughhands-on exploration of key protocols: Agent-to-Agent (A2A) for backend agent orchestration andinter-agent communication, Model Context Protocol (MCP) for persistent state management andreproducibility, and Agent-Governed UI (AG-UI) for frontend UI/UX and safe human-in-the-loopworkflows.
Target audience: Geospatial scientists and researchers working with location data; AI/ML researchers and practitioners interested in geospatial applications; Software developers building location-aware applications; Data scientists working with Earth observation or spatial datasets; Policy makers and domain experts (urban planning, climate science, transportation); Students and early-career researchers in GeoAI; Industry professionals in logistics, real estate, environmental monitoring
W7 (Hands-on and Discussion) – Advancing Feature Extraction and Map Quality Analytics through GeoAI and HERE Mapmaking
Organiser: Johannes Lauer (HERE Technologies)
Goals and scope:
In this workshop, we will explore the HERE Mapmaking ecosystem and demonstrate how it enables AI‑supported map data creation and maintenance.
Participants will see how to access, inspect, and interact with map data directly within the platform.
Using a navigation‑ and road‑network‑focused dataset, we will jointly brainstorm strategies for keeping the digital representation of the real world fresh, complete, and high‑quality.
This includes discussing how GeoAI can support feature extraction, change detection, and continuous data quality improvement.
We will also showcase practical examples of how to connect to the HERE Mapmaking ecosystem using tools many participants already work with – such as Python, QGIS, and other familiar geospatial workflows – highlighting how these integrations can accelerate analysis, automation, and experimentation.
The workshop will provide a collaborative environment, where the attendees can combine their perspectives and experiences in GeoAI, maps, data analytics with tools and data to develop and create improved and/or new approaches for enhanced maps.