AI-Driven Security System with Automated Light Control for Energy Efficiency

Energy consumption in commercial buildings is significant and adopting energy-saving measures reduces costs and environmental impact. Traditional systems, like PIR sensor-based lights, often operate inefficiently, while surveillance cameras are mostly used for post-event analysis rather than real-time intervention. The proposed automatic corridor lighting and monitoring system enhances energy efficiency, safety and security in buildings such as colleges, research centers and hospitals. Using the You Only Look Once version 8(YOLOv8) model and an Light Dependent Resistor(LDR) sensor, lights are turned ON only in the dark when a person is detected. Security is strengthened by using Multi-Task Cascaded Convolutional Networks(MTCNN) with FaceNet for facial recognition, sending real time images of unauthorized individuals to authorities via Telegram. A smoke sensor detects hazards, activating a buzzer to warn occupants. This integrated system addresses energy, security and safety challenges effectively.