Real-Time Attention Auditing via Standard Hardware
Cameroonian software engineer and founder Randy Kimbi has created an artificial intelligence EdTech framework designed to mitigate attention loss and drowsiness during online instruction. Developed as a Bachelor of Technology project titled "Intelligent EdTech Framework for Real-Time Attention Auditing and Audio-Injected Drowsiness Mitigation," the system earned an Excellent grade from the evaluation jury at PHIBMAT Douala.
The software leverages standard webcam feeds to monitor facial features and eye aspect ratios in real time, detecting whether a student is actively engaged, distracted, or falling asleep. When the system flags diminished attention or prolonged eye closure, it automatically triggers an audio alert to snap the learner back to the lesson. Unlike conventional learning management tools that only log static video attendance or completion status, Kimbi's framework introduces active behavioral feedback to simulate the intervention of a physical classroom instructor.
EdTech Accessibility and Platform Integration
A central feature of Kimbi's framework is its focus on low-cost accessibility across emerging markets. By operating exclusively through basic computer webcams without requiring specialized biometric sensors, wearable hardware, or high-end graphics processing units, the architecture eliminates significant financial barriers to EdTech adoption in sub-Saharan Africa.
The innovation is slated for integration into Gizami, an online learning academy founded by Kimbi. Gizami aims to equip Cameroonian university students with hands-on computer science, digital marketing, and artificial intelligence skills. Incorporating real-time attention auditing allows the platform to verify active course engagement, addressing long-standing industry challenges surrounding low completion rates and passive screen time in remote education environments.
For further background on computer vision applications and automated system architectures, explore our research coverage on Causal World Models and Modular Agents and view our reporting under AI & Society.
