News
Categories
🔬 Research & Projects · 🎓 Teaching & Supervision · 🌍 Community & Leadership
🔬 Research & Projects
🎤 Two Research Presentations at IGARSS 2026
Aug 2026 @ Washington, D.C., USA
I presented two collaborative works at IGARSS 2026 in Washington, D.C.
As first author, I presented “Standards for AI-Enabled Earth Observation in Disaster Management: A Cycle-Based Landscape Analysis”, together with Nina Merkle and Monique Kuglitsch. We systematically examined existing standards across the disaster-management cycle and identified where AI-enabled Earth Observation still lacks common guidance and standards.
This work is part of the Global Initiative on Resilience to Natural Hazards through AI Solutions, where I lead the Roadmap Working Group. Our work connects AI-enabled Earth Observation, disaster-risk management, and international standardization, with the aim of developing a practical roadmap for future research and applications.
I also presented “Semi-Supervised Remote Sensing Segmentation via Foundation Model Distillation and Consistency Learning” as co-author, together with Pu Xiao, Xiaoyan Lu, Yuting Dong, Ji Zhao, and Yanfei Zhong. The work explores how knowledge from foundation models can improve remote-sensing segmentation when only limited labelled data are available.
Great discussions around both methodological AI research and the broader question of how we make AI-based Earth Observation reliable and useful in real-world applications.
🧪 TUM Data Innovation Lab Project Selected | Pop-up City
Jun 2026
Very happy that my project idea “Pop-up City: Layered Urban Asset Reconstruction from Single Images” was selected for the TUM Data Innovation Lab!
The project is hosted by the TUM Chair of Geoinformatics in collaboration with DLR and will start in October 2026. We will explore how urban scenes can be decomposed into meaningful layers, how occluded objects and surfaces can be reconstructed, and how these representations can support urban mapping and digital twins.
Looking forward to working closely with Thomas Fröch and the TUM Geoinformatics team, and to seeing what the student team will build from this idea! 🏙️🚀
🎤 First Results of Our VLM Building Research at EGU26
May 2026 @ Vienna, Austria
Great to present the first results from Ahmed’s Master’s research at EGU26: “Vision-Language Models for Structural Exposure Modeling from Street-Level Imagery”, together with Ahmed Abdelsalam, Xizhe Xue, Patrick Aravena Pelizari, and Christian Geiß.
We investigated whether vision-language models can infer engineering-relevant building attributes directly from street-level imagery. The preliminary results show promising potential for scalable and low-cost structural exposure modelling and also highlight where domain-specific adaptation is still needed.
It was great to share these early results with the community and receive useful feedback for the next stage of the work.
📝 Preprint | Governing Generative AI in Disaster Risk Management
Feb 2026
Our perspective “Governing Generative AI in Disaster Risk Management” is now available on EarthArXiv.
The work examines both the opportunities and governance challenges of using generative and agentic AI in disaster risk management. We argue that meaningful use of these technologies in high-stakes disaster settings requires human oversight, transparency, interdisciplinary collaboration, and clear governance from design to deployment.
A timely topic as generative AI increasingly moves from experimentation toward real-world applications in disaster management.
📝 TUM2TWIN Published
Feb 2026
Great to see “TUM2TWIN: Introducing the Large-Scale Multimodal Urban Digital Twin Benchmark Dataset” published in the ISPRS Journal of Photogrammetry and Remote Sensing! 🎉
This large interdisciplinary effort brings together multimodal urban data from a real urban environment, including point clouds, images, and 3D representations.
The dataset provides a rich benchmark for research ranging from 3D reconstruction and semantic understanding to multimodal data fusion and urban digital twins.
👉 TUM2TWIN project and publications
📝 New Publication | A Deep Dive into OpenStreetMap Research
Jan 2026
Happy to see our paper “A Deep Dive into OpenStreetMap Research Since its Inception (2008–2024): Contributors, Topics, and Future Trends” published in the International Journal of Geographical Information Science, together with Liqiu Meng, Andrés Camero, Stefan Auer, and Xiao Xiang Zhu.
We analysed nearly two decades of OpenStreetMap research to understand how the scientific community, collaboration networks, and research topics have evolved, while also comparing academic research with discussions from the broader OSM community.
All data and code are openly available so that the analysis can be reproduced and extended by the community.
🎤 GNI 2024 Paper Selected for Advanced Engineering Informatics
24 Oct 2024
Our extended abstract “Generating Façade Segmentation Datasets using Diffusion Models” was selected as one of the best submissions and presentations at the GNI Symposium and Expo and invited for submission as a full paper to Advanced Engineering Informatics. 🎉
Great work, Alejandro Rueda Segura!
📝 CAADRIA 2025 Paper Accepted
1 Oct 2024
Our paper “Generation of High Quality Annotated Building Façade Dataset using Pre-Trained Diffusion Models” was accepted at CAADRIA 2025.
Stay tuned for the full paper and dataset! 🎉
🎤 GNI 2024
10–12 Sep 2024
Alejandro Rueda Segura presented our work “Generating Façade Segmentation Datasets using Diffusion Models” at the GNI Symposium & Expo on Artificial Intelligence for the Built World. 👏
🎤 EuroCarto 2024
9–11 Sep 2024
I presented our work “Learning Building Floor Numbers from Crowdsourced Streetview Images” at EuroCarto 2024 in Vienna, Austria. 🏡🏢🌇
🎤 IGARSS 2024
7–12 Jul 2024
Yi Wang presented our work “Post-Earthquake SAR-Optical Dataset for Quick Damaged-Building Detection” at IGARSS 2024 in Athens, Greece. 👏
📝 New Publication | QuickQuakeBuildings
30 May 2024
Our paper “QuickQuakeBuildings: Post-earthquake SAR-Optical Dataset for Quick Damaged-building Detection” was published in IEEE Geoscience and Remote Sensing Letters.
🎤 ORNL Invited Talk
3 May 2024
I gave an invited webinar on “Building Information Extraction From Remote Sensing and Street-Level Imagery” for the Geospatial Science and Human Security Division at Oak Ridge National Laboratory (ORNL).
The talk brought together our work on extracting detailed building information from remote-sensing and street-level imagery and opened up interesting discussions on scalable urban and disaster applications.
🎓 Teaching & Supervision
🎓 Ahmed Successfully Defends His Master’s Thesis
Aug 2026
Congratulations to Ahmed Abdelsalam on successfully defending his Master’s thesis, “Multi-modal Building Characterization using Vision Language Models”! 🎉
It has been a pleasure supervising this work together with Dr. Patrick Aravena Pelizari and Prof. Christian Geiß. The thesis explored how vision-language models can combine street-level and satellite imagery to extract detailed building attributes at scale.
Many thanks to Prof. Debarghya Ghoshdastidar at TUM for hosting and supporting the thesis. We will continue developing this research further!
🧑🏫 IGARSS 2026 Tutorial | Responsible and Trustworthy AI for Earth Observation
9 Aug 2026 @ Washington, D.C., USA
A highlight of IGARSS 2026 was developing and delivering the half-day tutorial “Responsible and Trustworthy AI for Earth Observation: From Fairness to Reproducibility and Societal Impact.”
Together with Pedram Ghamisi, Raian Vargas Maretto, and Kasra Rafiezadeh Shahi, we discussed fairness and bias, explainability, uncertainty, reproducibility and standardization, and societal impact in AI for Earth Observation.
It was a very rewarding experience to bring these topics together and exchange ideas with the community on how we can make AI for Earth Observation more trustworthy and useful in practice. 🌍🤖
🎓 Joining Zhenjiang Li’s PhD Committee
Jun 2026
Happy to join the PhD thesis committee of Zhenjiang Li in the PhD Program in Epidemiology and Biostatistics of the Life Science Zurich Graduate School, together with Prof. Jan Fehr, Dr. Andrea Farnham, and Prof. Viktor von Wyl.
Zhenjiang’s PhD project, “Climate Change, Human Mobility, and Health in Mozambique: A Geospatial and Field-Based Study,” brings together climate hazards, human mobility, public health, field studies, and geospatial data.
This is a highly meaningful interdisciplinary topic, particularly for understanding how climate impacts translate into real consequences for vulnerable populations. I’m glad to contribute the Earth-observation and geospatial perspective, and I look forward to strong results and new collaborations across these fields.
🎓 Ahmed Starts His Master’s Thesis at DLR
Nov 2025
Happy to welcome Ahmed Abdelsalam from TUM to DLR for his Master’s thesis, “Multi-modal Building Characterization using Vision Language Models.”
Together with Dr. Patrick Aravena Pelizari and Prof. Christian Geiß, we are exploring how vision-language models can combine street-level and satellite imagery to characterize buildings at scale.
The goal is to extract structural and semantic building information that is difficult and expensive to collect manually and make it useful for large-scale exposure modelling and urban analysis. Looking forward to seeing where the project leads!
🌍 Community & Leadership
🧭 Chairing Two Sessions at IGARSS 2026
Jul 2026
Looking forward to IGARSS 2026 in Washington, D.C., where I will co-chair two sessions closely connected to my current work on responsible AI and standards for Earth Observation.
TU1.R8: Responsible AI for Earth Observation: Building Trustworthy, Transparent, and Ethical Pathways for the Future, together with Raian Vargas Maretto (ITC, University of Twente).
TU4.R8: Remote Sensing Standards for a Dynamic Earth, together with Hugo Carreno-Luengo (NASA & University of Michigan).
Both sessions address questions that are becoming increasingly important as Earth-observation AI moves from research toward operational use: trustworthiness, transparency, reproducibility, and standards.
🧭 EGU26 | Remote Sensing and Explainable AI for Disaster Monitoring
May 2026 @ Vienna, Austria
We had a great session at EGU26 on “Remote Sensing and Explainable AI for Hazard Assessment and Real-Time, Large-Scale Disaster Monitoring”.
The session was convened by Paraskevas Tsangaratos, together with Nina Merkle, Raffaele Albano, Wei Chen, Ioanna Ilia, and me as co-conveners.
The contributions covered rapid hazard and damage mapping, explainable AI, uncertainty, multimodal Earth Observation, and operational disaster monitoring. It was particularly interesting to see how rapidly these methods are moving toward real-world disaster applications and to discuss how they can become more transparent, robust, and operational.
🤝 Dr. Philipp Barthelme Visits DLR
May 2026
It was a pleasure to welcome Dr. Philipp Barthelme to DLR. He gave a very interesting talk, “Near Real-time Large Scale Urban Damage Monitoring in Conflict and Disaster Settings.”
We had a great discussion on large-scale urban damage monitoring and how InSAR time series, optical imagery, foundation-model representations, and other complementary data sources could be combined to make damage assessment more scalable and useful for humanitarian and disaster-response applications.
Lots of interesting connections to our own work and ideas for future collaboration!
👉 Related work presented at EGU26
🌐 Selected for BRAGFOST 2026
Mar 2026
Very happy to have been selected for the 16th Brazilian-German Frontiers of Science and Technology Symposium (BRAGFOST 2026), jointly organized by the Alexander von Humboldt Foundation and CAPES.
The symposium brings together researchers from Germany and Brazil across different disciplines, this year under the theme “AI for Good: Global Responsible Innovation and Business Transformation.”
I’m looking forward to exchanging ideas on trustworthy and socially impactful AI, meeting colleagues from different fields, and building new Brazilian-German collaborations. 🌎✨
🤝 Prof. Hongchao Fan Visits DLR
Feb 2026
It was a pleasure to welcome Prof. Hongchao Fan from NTNU to DLR after we met at the ISPRS–ISDE conference in Nairobi.
He gave a very interesting talk on his recent research, followed by a lively discussion on 3D city modelling, urban digital twins, crowdsourced geospatial data, and possible directions for collaboration.
There are many connections between his work and the urban research we are developing at DLR. It was a very inspiring visit, and I’m looking forward to continuing our discussions and developing new collaborations.
🧭 Next-Gen Disaster Management Workshop at ISPRS–ISDE 2025
Dec 2025 @ Nairobi, Kenya
A memorable few days in Nairobi, where Yu Li and I organized and chaired the workshop “Next-Gen Disaster Management: Multi-Modal Geospatial Data & AI for Resilient Communities” at ISPRS–ISDE 2025.
We brought together research on risk mapping, early warning, damage assessment, urban exposure, forced displacement, multimodal geospatial data, explainable AI, and responsible AI for disaster management.
The workshop generated many interesting discussions, new connections, and follow-up ideas beyond the event itself — exactly what we hoped for when we started organizing it. Many thanks to all speakers, participants, and colleagues who contributed!
🧭 iCARE 2025 | GeoAI and Digital Twins for Climate Resilience
Oct 2025 @ Munich, Germany
A great few days at the International Climate Resilience Conference (iCARE 2025) in Munich.
Together with Xinyue Ye, Lukas Lehnert, and Christian Geiß, I co-convened the session “Leveraging GeoAI and Digital Twins for Climate Resilience”.
We had very interesting discussions on how GeoAI, digital twins, environmental data, and socioeconomic information can be brought together to better understand climate risk and support adaptation and resilient communities. Great to see so many complementary perspectives coming together in one session!
🌐 New Role | Co-Lead of IEEE GRSS IADF WG-RAI
Sep 2025
I’m honored to serve as Co-Lead of the Responsible AI Working Group (WG-RAI) within the IEEE GRSS Image Analysis and Data Fusion (IADF) Technical Committee.
Together with Prof. Pedram Ghamisi, Dr. Kasra Rafiezadeh Shahi, and Dr. Raian Vargas Maretto, we aim to advance responsible AI practices for Earth Observation — from explainability and uncertainty to fairness, reproducibility, transparency, and societal impact.
Looking forward to working with the community to make responsible and trustworthy AI an integral part of Earth-observation research and applications. 🌏🤖
👉 IEEE GRSS IADF Technical Committee
🎤 EGU25
2 May 2025
I had a wonderful time at EGU25 in Vienna. I especially enjoyed our debate session and reconnecting with so many friends and colleagues. 💗
Looking forward to what’s ahead! 🌏🚀
🌐 Invited to EGU25 Debate Session
1 Apr 2025
Happy to share that I’ve been invited as a panelist in the debate session “Responsible AI for Natural Hazards Assessment and Management” at EGU 2025.
Looking forward to meaningful discussions on the responsible and impactful use of AI in this critical field.
🌐 New Position at DLR
1 Jan 2025
Happy to start a new position at the German Aerospace Center (DLR). 🔭🌏🚀
I’m looking forward to developing new research on Earth Observation, urban environments, and disaster applications, and to building new collaborations within DLR and beyond.
🧭 IGARSS 2025 Community-Contributed Session
7 Nov 2024
Happy to announce that our session “Datasets and Benchmarking for Rapid Disaster Response: Multimodal Approaches for Post-Disaster Damaged Building Detection” has been accepted for IGARSS 2025, taking place in Brisbane, Australia.
The session focuses on datasets, benchmarks, and multimodal approaches that can make post-disaster building-damage assessment faster, more comparable, and more useful in operational response.
Looking forward to seeing you in Brisbane! 🌏✨
🧭 Mapathon at TU Munich
22 Jul 2024
YouthMappers at TU Munich, in collaboration with Geoawesome, Deutsches Rotes Kreuz (German Red Cross), and Runder Tisch GIS e.V., organized a successful Mapathon event.
The event brought together students and partners to contribute geospatial data for humanitarian mapping while learning how collaborative mapping can support real-world applications.
👉 LinkedIn Post 🌏
🤝 GFZ Research Stay
May–Jun 2024
I visited the GFZ German Research Centre for Geosciences, working with the group Radar and Optical Remote Sensing for Analysis of Geohazards. 🍀🌟
The stay provided a great opportunity to exchange ideas on remote sensing for geohazards and explore connections between SAR, optical Earth Observation, and disaster research.