Inquiries
in motion.

This participatory design initiative develops an AI-enhanced Multimodal Learning Analytics (MMLA) platform for collaborative K-12 STEM classrooms, co-designed with a transdisciplinary team of researchers, educators, and developers. Grounded in sociocultural perspectives, the project examines how MMLA platforms can be responsibly integrated into secondary learning environments, attending to the cultural, social, and material contexts in which collaborative STEM+AI learning unfolds. The project also builds a foundational multimodal dataset to support computer vision research for video-based MMLA.

AI is reshaping education, yet most AI in Education (AIED) systems are built on Western, monolingual, and individualistic models of learning that overlook the diverse ways students engage in STEM practices. The Global Responsible AI in Education Collaborative reimagines AIED through multimodal, multilingual, and globally grounded lenses, leveraging AI to make the invisible visible and the silenced heard. This participatory design project brings together researchers, practitioners, community partners, and families from four hubs, Virginia, New Mexico, Brazil, and South Korea, to examine how AI can support more equitable K-12 STEM education and develop research priorities and actionable guidelines for stakeholders.

This project builds methodological capacity in computer vision and spatiotemporal deep learning for AI-enhanced Multimodal Learning Analytics in K-12 STEM classrooms.

This project develops an openly licensed, interactive textbook on learning theories to advance affordability, equity, and student success across education programs. Existing commercial texts are costly, fragmented, and misaligned with digital and AI-enhanced learning contexts. Replacing multiple required textbooks with a single open resource will reduce student costs and improve instructional coherence. Aligned with VIVA Open priorities, the textbook integrates diverse perspectives, inclusive practices, and interactive case-based learning. Its modular, open design supports sustainability, regular updates, and cross-institutional adoption.