FIELD BOUNDARY INITIATIVE

Overview

The Field Boundary Initiative is the Taylor Geospatial Engine's first instantiation of the Innovation Bridge Program. It aims to gather experts to collaborate at the intersection of satellite imagery, cloud data sharing and machine learning (ML) for agricultural datasets. The initiative will center around creating and updating field boundaries derived from satellite imagery using ML models and increasing the interoperability of agricultural data. Our goal is to explore, and begin to define, an open data ecosystem, using learnings from open initiatives like Linux Foundation and OpenStreetMap to enable collaboration across diverse organizations that enables an ecosystem of innovation. 


While holding a vision of creating a global field boundary dataset, the work during the Initiative will be on concrete steps that enable progress today, including common data schemas, harmonization of existing datasets (both open and proprietary), ML model development, and making it easier to discover and work with good training data for AI/ML algorithms. The full initiative will run for around six months, funding cutting edge academic research and facilitating the collaboration between industry, academia, NGO's and government. The overall initiative will be primarily virtual, through collaboration on GitHub (for open code & data schemas) and Source Cooperative


We're kicking off the initiative on Feb 26-28, 2024 at the Knight Center on the Washington University in St. Louis campus.  We're holding an in-person workshop with a few sessions available for remote attendance. During the kick-off, we hope to:


If we do our jobs correctly, the kick-off will build bonds and spark fruitful collaboration for the rest of the initiative.


Follow the links for more details on the workshop: