Overview
Child Fall Detection System is an AI-powered safety monitoring solution that detects child falls in real time using pose estimation, edge-friendly computer vision, and instant alert delivery through IoT communication pipelines.
Problem to Solve
Build a low-latency and reliable child safety monitoring system that can automatically detect fall incidents and notify caregivers immediately, while remaining affordable, scalable, and practical for home or daycare deployment.
Solution Approach
Developed an AI-based fall detection pipeline using MediaPipe pose estimation to identify body movement patterns associated with child falls, then integrated the vision system with ESP32-CAM for real-time image capture and MQTT for lightweight message transmission between devices and backend services. Designed and implemented an instant alert mechanism using the Telegram Bot API so caregivers receive immediate notifications when a fall event is detected, while applying edge computing principles to improve responsiveness, reduce latency, and support efficient system operation in resource-constrained environments.
Impact
Delivered a functional real-time monitoring prototype that combines AI, IoT, and instant messaging for practical child safety surveillance, demonstrating a cost-efficient and scalable approach suitable for early-warning systems in households and daycare settings.