EAGER: Scaling AI Literacy through Edge AI Computing in K-12 Convergent Research

Artificial intelligence (AI) increasingly shapes daily life, and many high school students encounter AI by engaging as users of these tools through apps. While these experiences introduce AI concepts, opportunities to apply AI in authentic, real-world scientific inquiry are limited. This project addresses that gap by creating a low-cost Edge Artificial Intelligence Kit that allows students to participate directly in the scientific process: collecting and analyzing data, testing hypotheses, and interpreting results in natural settings. By using real-time computer vision to observe and track honey bees, students explore AI system behavior and performance while developing computational reasoning and cross-disciplinary STEM skills. These experiences provide meaningful, hands-on engagement with both AI and science, preparing students for thoughtful participation in a technology-rich world. Accordingly, this project contributes to the goals stated in the Dear Colleague Letter NSF 25-035 regarding advancing AI education for the American youth. This project develops and studies a model for integrating real-time edge-based computer vision systems into field-based secondary STEM education. Research-oriented environmental monitoring prototypes are adapted into a modular, containerized software and hardware platform that can be assembled and deployed by non-technical educators without local programming expertise. Tenth-grade students engage the full Artificial Intelligence pipeline, including sensor calibration, edge deployment, data collection, and evaluation of system performance under real-world conditions. The research examines how low-cost, offline edge computing infrastructure supports inquiry-based, cross-disciplinary learning across computer science and the natural sciences. Outcomes include a scalable classroom-ready platform, structured instructional modules, and empirical evidence on how accessible edge Artificial Intelligence systems can sustain rigorous, data-centric STEM engagement at scale. By enabling schools to contribute locally collected environmental data to a shared dataset, the project also strengthens connections between education and community-relevant scientific research. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria. NSF Award ID: 2619327 | Program: 01002627DB NSF RESEARCH & RELATED ACTIVIT | Principal Investigator: Remi Megret | Institution: University of Puerto Rico-Rio Piedras, SAN JUAN, PR | Award Amount: $118,000 View on NSF Award Search: https://www.nsf.gov/awardsearch/show-award/?AWD_ID=2619327 View on Research.gov: https://www.research.gov/awardapi-service/v1/awards/2619327.html