Researchers from the Technical University of Munich (TUM) and the University of Cologne have unveiled an innovative AI-driven learning system using webcam-tracked eye movements to transform math.
Designed to tailor problem-solving hints and exercises, this technology promises to make individualized learning accessible to more students.
Imagine a classroom with standard webcams and computers—no high-end gear required.
Professor Achim Lilienthal from TUM explains how eye-tracking data reveals patterns in how students solve math problems. These insights are displayed as heatmaps, where frequent focus areas glow red and fleeting glances appear green.
The AI then classifies these visual patterns, identifying whether students excel or need extra help. Based on this data, the system suggests targeted exercises and learning videos tailored to each student’s needs.
Screen time is linked to childhood development delays
“This combination of webcam eye tracking, learning strategy recognition, and automated teacher support reports is unprecedented,” says Professor Maike Schindler from the University of Cologne. Her research, which began under the KI-ALF project funded by Germany’s Ministry of Education, focuses on assisting students struggling with basic arithmetic skills.
However, the benefits aren’t limited to those with difficulties—high-achieving students can also receive custom challenges to advance their learning.
The AI system’s real-world impact is already being seen at Wulfen Comprehensive School in North Rhine-Westphalia. A math skills assessment showed that one-third of Year 5 students struggled with arithmetic.
Now, with the help of this AI system, five students at a time can receive personalized lessons, effectively expanding the reach of individual teacher support.
Professor Lilienthal also highlights how this innovation bridges the gap between cost and precision. While traditional eye trackers are prohibitively expensive, the research team has fine-tuned standard webcams to achieve a similar level of accuracy. By combining expertise in robotics and mathematical education, they’ve created a scalable solution for schools worldwide.
Journal References
- Introduction to eye tracking in mathematics education: interpretation, potential, and challenges; Maike Schindler, Anna Shvarts & Achim J. Lilienthal; Educational Studies in Mathematics (ESM); 3-2025; DOI: 10.1007/s10649-025-10393-1
- Structure Sense in Students’ Quantity Comparison and Repeating Pattern Extension Tasks: An Eye-Tracking Study with First Graders; Demetra Pitta-Pantazi, Eleni Demosthenous, Maike Schindler, Achim J. Lilienthal, and Constantinos Christou; Educational Studies in Mathematics (ESM), 2024, pp. 1 – 19; DOI: 10.1007/s10649-023-10290-5
