VisionGuard
Real-time computer vision system that detects safety violations on construction sites and warehouses — cutting incident response time from minutes to seconds using edge-deployed YOLO models.
Computer VisionYOLOPyTorchEdge ComputingPython
VisionGuard
A safety monitoring platform that runs optimized YOLOv8 models on edge devices to detect missing PPE, restricted zone breaches, and hazardous conditions in real time. Alerts fire within seconds of a violation, with annotated video frames pushed to a central dashboard. The PyTorch training pipeline supports custom object classes so each deployment adapts to site-specific safety requirements.