Smart Night Patrolling Robot using YOLOv8

Abstract

The traditional patrolling involves the security guards or police or to monitor CCTV packages at the commercial area. Robotics systems can provide an efficient and cost-effective solution to security issues, and can enhance traditional security measures. The smart night patrolling robot is a prototype designed to enhance security measures in gated communities, industrial premises, and apartment complexes during the night. It incorporates multiple sensors such as a sound sensor, GPS, ultrasonic sensor, and IR sensor, powered by a rechargeable battery to detect the sound or object in the surrounding along with the location. The proposed system implemented YOLOv8 model for object detection from the detected data. If the object is deemed suspicious or if any human activity is detected, the robot sends an emergency message to a specified chat ID using the Telegram API. The model achieves better accuracy of 92% for detecting objects in night vision compared to other versions of YOLO Model. It is designed to operate autonomously, freeing up security personnel to focus on other tasks. The robot’s ability to detect and investigate suspicious activity makes it a valuable addition to any security system.