Researchers at the University of Virginia’s School of Education and Human Development and School of Engineering and Applied Science will leverage artificial intelligence to improve the quality of teachers’ mathematics instruction with a new $1.4 million grant from the Bill & Melinda Gates Foundation.
Through the project, “Artificial Intelligence for Advancing Instruction at Scale (AI2S),” the researchers will build and test an AI model to classify specific elements in videos of teachers’ instruction, from whole-group instruction to individual student work. The team will also develop a dashboard that will give teachers real-time, applicable feedback to improve their instruction.
Classroom observational videos are a very helpful tool for improving teaching skills and yet assessing them is an incredibly time-consuming task.
“Classroom observational videos are a very helpful tool for improving teaching skills and yet assessing them is an incredibly time-consuming task,” said Peter Youngs, professor at the UVA School of Education and principal investigator of the grant. “By leveraging AI to classify the videos, we hope to provide teachers meaningful feedback that’s not only practical but is also easy to access and affordable.”
The grant builds on cutting-edge research from professor and chair of the Department of Electrical and Computer Engineering Scott T. Acton, who serves as a co-principal investigator on the grant and first engineered AI breakthroughs in human action detection technology, applying neural networks to video analysis with remarkable precision.
This is about helping teachers do what they do best — teach — while making professional development more accessible and equitable.
“This is about helping teachers do what they do best — teach — while making professional development more accessible and equitable,” said Acton. “Combining AI technology with real-world educational challenges is an exciting way to impact classrooms across the country.”
By pairing Youngs’ expertise with Acton’s innovations in engineering, the project creates a seamless blend of educational insight and advanced technology.
This collaboration, which includes Jonathan Foster, assistant professor at the University at Albany, ensures the tool isn’t just technically sophisticated but also practical and meaningful for teachers in diverse classrooms.
The multi-modal AI tool uses synchronized video and audio to evaluate classroom dynamics, capturing details like the quality of teacher-student dialogue, cognitive rigor of lessons and instructional patterns. It surpasses traditional systems that rely solely on video or audio, offering richer, more holistic insights.
When paired with coaching, the feedback generated by the tool is designed to improve the skills and quality of elementary school teachers’ mathematics instruction.
The project will also include co-investigator Adam Geller, founder of Edthena, an organization building technology tools that streamline feedback to teachers. Edthena’s “AI Coach” is a next-generation professional learning and feedback tool that utilizes conversational AI to guide teachers through a complete coaching cycle in an online, self-directed setting. Youngs, Acton and Foster will collaborate with Geller and Edthena to develop a teacher dashboard to deliver the real-time, applicable feedback to improve teachers’ instruction.
With an emphasis on equity, the project targets classrooms serving diverse student populations, particularly in under-resourced schools. By making professional development more efficient and scalable, it aims to support teachers in delivering equitable, rigorous math instruction while reducing costs for schools and districts.
This initiative could dramatically expand its impact, with future applications across subjects like English, history and science. With the grant from the Gates Foundation, the AI2S team aims to support meaningful changes in how schools empower their educators.