The Institute for Geospatial Understanding through an Integrative Discovery Environment (I-GUIDE, https://i-guide.io/) is supported by the National Science Foundation with the vision to enable digital discovery and innovation through harnessing the geospatial data revolution and the mission to advance convergence and geospatial sciences for holistic sustainability solutions.

The I-GUIDE Platform (https://platform.i-guide.io/) provides an open science and collaborative environment for geospatial data-intensive convergence research and education focused on sustainability and resilience challenges and enabled by AI and advanced cyberGIS & cyberinfrastructure. The I-GUIDE Platform has been developed and supported by I-GUIDE and its multi-institutional team and various partners. The Platform promotes FAIR (Findable, Accessible, Interoperable, and Reusable) data, and responsible & open science practices.

Spatial Artificial Intelligence (Spatial AI) refers to the integration of AI technologies with spatial data and models to understand space and the environment and make related decisions. It involves advancing and using machine learning, computer vision, and other AI techniques to process, analyze, and synthesize data that has a geographic or spatial component—such as cartographic maps, satellite images, and location-based data.

The Spatial AI Challenge is a community engagement event designed to bring together researchers, data scientists, AI enthusiasts, and professionals to advance AI-ready spatial data and machine learning & deep learning models while promoting FAIR (Findable, Accessible, Interoperable, and Reusable) data and responsible & open science practices. Hosted on the I-GUIDE Platform, the challenge focuses on advancing spatial AI to solve complex geospatial problems and pressing sustainability challenges.

Participants will showcase innovative approaches and contribute to open, responsible spatial AI solutions that benefit the broader research community. The challenge offers hands-on experience with the I-GUIDE Platform and emphasizes FAIR data principles, ensuring that data and solutions are accessible, trustworthy, and reusable. Participants can register individually or as part of a team and compete for awards and recognition.

Each submission is highly encouraged to holistically advance data, models, and applications; or solving one or multiple open problems. Submissions will be evaluated on computational reproducibility, innovation, technical excellence, FAIR data principles, and societal impact. The event starts on October 15, 2024, with the final competition deadline of April 15, 2025.

Participating in the Spatial AI Challenge offers an array of benefits for researchers, data scientists, and AI enthusiasts, including:

• Hands-on Experience with Cutting-edge Technology: Gain direct access to the I-GUIDE Platform and its backend high-performance computing resources. Participants will have the opportunity to explore state-of-the-art AI, rich spatial data, and cutting-edge geospatial tools in a collaborative environment.
• Advance Responsible & Open Science: Showcase your work while contributing to the global movement of responsible & open science. Submissions are encouraged to follow FAIR (Findable, Accessible, Interoperable, and Reusable) data principles for achieving broader impacts.
• Recognition and Awards: Compete for prizes and earn recognition for innovation in spatial AI, data science, and responsible & open science. Stand out for your contributions to solving real-world geospatial problems and pressing sustainability challenges.
• Contribute to Real-world Problem Solving: The challenge emphasizes the potential for real-world impact, with a focus on tackling pressing sustainability challenges.
• Skill Building and Learning Opportunities: Enhance your skills in spatial AI and data science. The challenge will also offer webinars, virtual consulting office hours, and resources to guide you along the way.

By participating, you’ll not only contribute to the advancement of spatial AI and data science, but also benefit from professional development, networking, and the chance to make a lasting impact on the future of responsible & open science, geospatial discovery and innovation, and sustainability research and education.

You need to submit an abstract first. If your abstract is accepted, you will be invited to participate in the challenge.

Each submission must specify one of the following two tracks:

• Integration of Data, Model, and Application: Focus on AI-ready spatial datasets and geospatial data fusion tasks, explainable and generalizable spatial AI models, or real-world applications that address pressing sustainability challenges.
• Solving Open Problems: Focus on novel solutions to challenging problems that have real-world impacts.

• Challenge Launch Date: Oct 15, 2024
• Abstract Submission Deadline: Dec 1, 2024
• Abstract Acceptance Notification Date: Jan 3, 2025
• Submission Deadline: Mar 31, 2025
• Open Competition Deadline: Apr 15, 2025
• Announcement of Winners: May 15, 2025

Participants can submit datasets, models, and applications across various domains like environmental analysis, social equity, or sustainability. Submissions should align with open science practices, such as following FAIR (Findable, Accessible, Interoperable, and Reusable) data principles, and responsible data science guidelines. For example, you could submit a new dataset on climate change impacts on urban planning, which should be reusable for spatial AI tasks such as for predicting social disparities in access to public services.

All submissions must include well-documented Jupyter notebooks explaining the methodology used, which will be made available on the I-GUIDE Platform.

No, there is no fee to participate in the Spatial AI Challenge. Participation is free and open to everyone interested in advancing spatial AI research, FAIR data principles, and open science practices.

The challenge is open to everyone. Researchers, data scientists, AI enthusiasts, educators, students, and industry professionals from all backgrounds are encouraged to participate. Whether you're part of an academic institution, private industry, or working independently, you're welcome to participate.

There is no strict limit on the number of people that can be part of each team. Teams can consist of as many members as needed to effectively collaborate and participate in the challenge. Whether you choose to work individually or within a group, all submissions will be evaluated equally based on the criteria provided.

Yes, you can submit both previously published and unpublished data.

Large datasets and models may require additional coordination with the I-GUIDE Platform team. Please contact spatial-ai-challenge@i-guide.io and explain your situation and what you plan to submit.

Proprietary and private information is generally not allowed, thus authors must verify the licensing and privacy of their submissions before proceeding with the challenge. Submissions which are found in violation will be removed from consideration.

Submissions must be made through the I-GUIDE Platform. Detailed submission guidelines will be provided upon acceptance of abstracts.

Each submission must include:

• A dataset, model, or application based on one of the three competition tracks.
• A well-documented notebook that explains the methodology, datasets, models, and steps taken.
• A clear description of how FAIR data principles, and responsible & open science practices are followed.

Each participant or team is allowed one submission per competition track. You can submit up to three entries if you wish to participate in more than one track.

Yes, a Jupyter Notebook is required. The notebook will serve as a guide to how your submission meets evaluation criteria. It's encouraged to use clear explanations and storytelling to describe your workflows. This will not only help showcase how your work aligns with the criteria, but also make it easier for others to understand and access your data and workflows. A well-organized notebook improves accessibility and ensures that your data and algorithms are easy to follow for the wider community.

You can connect with other participants once submissions are made publicly available, two weeks before the final submission date. During this period, teams are encouraged to collaborate, share insights, and improve their work by reaching out to other teams. This open exchange of ideas helps foster collaboration and refine submissions ahead of the final deadline.

Submissions for the Spatial AI Challenge will be evaluated based on the following criteria:

• AI-ready Data (30%): Newly created datasets must be immediately usable for spatial AI models or applications, with metadata and/or APIs provided. Both new and existing datasets should be complete, well-structured, and include clear splits for training, validation, and testing. Ease of integration and standardization for AI practitioners are important.
• Adherence to FAIR Data Principles, and Responsible & Open Science Practice (20%): Contributions to open science through accessible, well-documented, and reusable data and models are critical. Evidence of attention paid to maximizing societal benefits and minimizing risks of harm in data acquisition and use, model development, and applications.
• Notebook Quality (20%): A well-structured and clearly documented notebook is crucial for computational reproducibility and ease of understanding. Submissions should explain data, models, and results clearly with adequate comments, documentation, and effective use of visualizations.
• Novelty (10%): We value the definition and exploration of novel or underexplored spatial AI problems. Creative and original problem identification that advances the field of spatial AI is highly encouraged.
• Impact and Applicability (10%): Submissions are encouraged to demonstrate real-world applicability, particularly in solving geospatial or sustainability challenges.
• Innovation and Creativity Based on Notebook and Storytelling (10%): Innovation in spatial AI techniques and storytelling demonstrated through Jupyter notebooks is expected to earn a high score, while creative approaches to problem-solving and effective presentation of findings will be evaluated highly.

All submissions will be evaluated by a committee of experts based on these criteria.

To ensure your submission complies with ethical standards, please consider the following guidelines:

• Data Privacy and Consent: Ensure that any data used in your submission respects privacy laws and regulations, such as the General Data Protection Regulation (GDPR). If your dataset includes personal or sensitive information, ensure that it is properly anonymized or that you have obtained necessary permissions or consent for its use.
• Responsible AI Use: The AI models and techniques you employ should be used responsibly, without causing harm to individuals, communities, or the environment. Avoid using biased or discriminatory data, and ensure that your models are designed and tested to avoid unintended consequences or ethical risks.
• Transparency and Computational Reproducibility: Be transparent in your methodology and data usage. Clearly document your processes, models, and data sources so that others can verify your work. Open sharing of both your data and code aligns with open science practices and promotes ethical accountability.
• Adherence to FAIR Data Principles: Your submission should follow the FAIR (Findable, Accessible, Interoperable, Reusable) data principles to the best degree possible to ensure that your data and models are responsibly shared and can be easily accessed and reused by the research community.
• Environmental Considerations: Consider the environmental impact of your submission. This includes optimizing your models for efficiency to reduce energy consumption and ensuring that any geospatial or sustainability challenges addressed in your work contribute positively to societal or environmental outcomes.
• Acknowledgment and Citation of Sources: Properly cite any data, models, and code that you did not create. Give credit to the original creators and ensure you are using these resources in line with any licenses or terms of use.

By following these guidelines, you will help ensure that your submission adheres to high ethical standards, respects privacy, promotes computational reproducibility, and contributes positively to the broader research community.

Submissions will be reviewed by a panel of experts from academia, industry, and relevant fields, ensuring a fair and comprehensive evaluation process.

Participants will have access to various support and resources, including:

• The I-GUIDE Platform for data access and high-performance computing resources.
• Webinars and virtual consulting office hours with experts to guide and assist participants.
• Community forums for discussion and collaboration.

For any questions or support, participants may contact the challenge organizers at spatial-ai-challenge@i-guide.io. There will also be regular webinars and virtual consulting office hours to help participants.

Please send additional questions to this email address: spatial-ai-challenge@i-guide.io.