Enter Note Done
Go to previous page in this tab
Session
  • Presentation | GC31C: Next-Gen GeoAI: Scalable and Research-Driven Machine Learning Applications for Environmental Impact I Oral
  • Oral
  • Share Icon
    Map Icon
    Bookmark Icon
  • GC31C-04: Deploying GeoAI Systems in Real-World Agricultural Applications (invited)
    • Schedule
    Notes
  • Location Icon271
    NOLA CC
    Set Timezone
    •  
    View Map

Generic 'disconnected' Message
Author(s):
Ana M Tarano, Arizona State University (First Author, Presenting Author)
Gabriel Tseng, McGill University
Ivan Zvonkov, University of Maryland
Catherine Nakalembe, University of Maryland
Amna Elmustafa, Arizona State University
Tristan Grupp, World Resources Institute
Inbal Becker-Reshef, University of Maryland
Hannah Kerner, Arizona State University


Before deployment, developers have to consider user priorities, infrastructure constraints, and data governance. Our approach evaluates success not only by model performance, but by usability, adoption, and operational fit.


This presentation highlights projects that use components from previous geospatial AI deployments (i.e., Presto, Fields of the World, Street2Sat, and crop-mask). These open-source and cloud-based tools allow rapid prototyping and adaptability to new users, tasks, regions, or data constraints. We use design thinking principles to bridge the gap between research and real-world applications with 4 examples.


We will show how we create impactful solutions that address both research and practical needs of applications by valuing stakeholders and different disciplines.




Scientific Discipline
Neighborhood
Type
Main Session
Discussion