Spatial AI Challenge 2024
Fostering FAIR Data and Open Science Practices Using the I-GUIDE Platform
Home Selected Winners Open Challenge Problems FAQ Accepted Abstracts
Welcome to the Spatial AI Challenge 2024
The Spatial AI Challenge 2024 is an international initiative at the crossroads of geospatial science and artificial intelligence. Hosted on the cutting-edge I-GUIDE Platform, this challenge brings together researchers, data scientists, AI enthusiasts, and geospatial professionals to develop innovative, responsible, and reproducible solutions for some of today’s most pressing sustainability challenges.
Winners of the Spatial AI Challenge 2024
We are proud to recognize the top teams whose submissions stood out for their innovation, technical rigor, and commitment to open science. After a competitive review process, the following projects have been selected as this year’s winners:
🥇 Winner
GeoMapCLIP
Team: Purdue RCAC
Team members: Jungha Woo, Elham Jebalbarezi Sarbijan — Purdue Research Computing
Summary:
This project developed GeoMapCLIP, a fine-tuned extension of GeoCLIP, to automatically localize unknown geospatial images using visual cues. Judges praised its well-structured, reproducible notebook, engaging storytelling, and generalizability across domains like agriculture, hydrology, and historical mapping.
🥈 Second Place
Wildfire Threat Detection for Transportation Infrastructure using U-Net for Semantic Segmentation
Team: Center for Naval Analyses (CNA)
Team members: John Crissman, Joel Diaz, Angie De Groot, Jeremiah Huggins, Lars Hanson, Steven Habicht, Matthew Prebble, Shaun Williams, Carey Whitehair-Conde, Rebekah Yang — Center for Naval Analyses (CNA)
Summary:
This submission presented a compelling approach to identifying wildfire risk for critical infrastructure using deep learning and open data. It leveraged a U-Net-based pipeline and performed well in data readiness, FAIR practices, and applicability to real-world emergency response planning.
🥉 Third Place
Pipeline for Roof Material and Geometry Classification
Team: Roof Mapper
Team members: Julian Huang, Yue Lin — University of Chicago; Alex Monito Nhancololo — University of São Paulo
Summary:
Focusing on disaster resilience, this project used satellite imagery and OSM data to classify roof materials across urban environments. Reviewers appreciated its real-world applicability, integration of open tools, and clear documentation.
What Is the Spatial AI Challenge?
At its core, the Spatial AI Challenge is about leveraging the power of spatial data and AI to address real-world issues. Participants will use the advanced cyberinfrastructure of the I-GUIDE Platform to explore new datasets, build state-of-the-art spatial AI models, and create applications that can make a tangible difference in areas like environmental management, urban planning, disaster response, and more.
This challenge is not just about competition—it’s about collaboration, open science, and the responsible use of technology. By adhering to FAIR data principles (Findable, Accessible, Interoperable, and Reusable) and embracing reproducible research practices, every contribution made during the challenge will help pave the way for a more connected, data-driven approach to solving global issues.
Why Participate?
- Innovate with Impact: Use advanced AI and geospatial analytics to develop solutions that directly address sustainability and societal challenges.
- Collaborate and Learn: Engage with a global community of experts and peers through interactive sessions, webinars, and virtual office hours.
- Showcase Your Work: Submit your project using computationally reproducible Jupyter notebooks and gain recognition from a panel of esteemed experts.
- Earn Rewards: Compete for cash prizes, travel support to key I-GUIDE events, and the opportunity to publish your work in renowned venues.
- Contribute to Open Science: Help drive the movement for responsible, transparent, and impactful research.
How the Challenge Works
The Spatial AI Challenge 2024 is structured around two key competition tracks:
- Data, Models, and Their Applications: In this track, participants can submit innovative datasets, develop novel spatial AI models, or showcase applications that integrate data and technology in new ways.
- Open Problems: Here, you will tackle a set of curated open geospatial challenges. Whether you choose to solve multiple problems or focus on a single issue, your work should aim to create actionable, sustainable solutions.
Each submission is evaluated on its technical excellence, creativity, and adherence to FAIR data and open science principles, ensuring that all contributions are both groundbreaking and accessible to the broader community.
Explore the Accepted Abstracts
Visit our Accepted Abstracts page to review the projects that will be shaping the future of spatial AI and sustainability research. These abstracts represent a diverse range of disciplines and approaches, all united by a common goal—to harness the power of spatial data for real-world impact.
Have Questions?
We’re here to help! If you have any questions regarding the I-GUIDE events or general inquiries, feel free to reach out.
Contact Us: events@i-guide.io