Learning Analytics Through Video Analytics and Wearable Sensors for Real-Time Attention Monitoring in Classrooms

In traditional classroom settings, monitoring students’ attention and engagement levels remains a persistent challenge. This paper presents an action research study that integrates real-time learning analytics using video analytics and wearable sensors to support instructional interventions and enhance learning experiences. The system captures group-level attention and activity data through skeleton-based video processing and wearable smartwatches, analyzing trends without storing personal video recordings. It provides instructors with instant feedback via a dashboard, enabling adjustments such as pausing for quizzes, initiating group discussions, or offering breaks. The automated setup supports up to fifty students per session and emphasizes group analytics over individual tracking to respect privacy. The study demonstrated a significant improvement in average attention levels—from 50% to 70%—through tailored pedagogical strategies informed by AI-driven analytics. This work aligns with Sustainable Development Goal (SDG) 4 on inclusive and quality education and illustrates the feasibility of deploying scalable, privacy-conscious AI systems in live classroom environments.