Final Year Project

Project Title

Real-Time Safety Monitoring System for Industrial Workplaces

Description

Our project enhances workplace safety using computer vision and deep learning to detect hazards in real time. In collaboration with Dawlance, we deployed this system at their Karachi plant, where safety lapses can cause serious accidents. We developed an AI-powered monitoring system to detect unsafe behaviors like missing helmets or restricted area entry, instantly alerting supervisors to reduce workplace accidents and ensure safety compliance. Due to data privacy concerns, we built a fully local solution for on-site processing. We installed and interfaced CCTV cameras while optimizing detection across multiple feeds. Since no suitable datasets existed, we manually created and annotated one, and fine-tuned our model, significantly improving accuracy. Our system enhances hazard detection, ensuring: Faster response times to violations Reduced workplace injuries through proactive monitoring Data-driven insights for long-term safety improvements