Level-4 Autonomous Robot for Late-Night Patrolling, Crime Monitoring, and Live Streaming

Author : Gracy Bandaru

Abstract :This paper presents a Level 4 autonomous robotic system for late-night patrolling, crime surveil- lance, and live video streaming, integrating advanced technologies like LIDAR for spatial tracking, encoder motors for precise movement, and a 360-degree rota- tional camera for real-time monitoring. Utilizing ROS 2, SLAM, and Gazebo, the system creates accurate virtual environments for navigation in complex areas. It employs an LSTM model trained on the UCF Crime Dataset for detecting suspicious activities and YOLO for real-time weapon detection. Upon detecting threats, the system sends immediate alerts with GPS. Experimental results validate its effectiveness, demonstrating its po- tential as a scalable and reliable automated security solution

Keywords :Autonomous robotics, LSTM, crime surveillance, ROS 2, SLAM, UCF Crime Dataset, suspicious activity detection.

Conference Name :International Conference on Engineering & Technology (ICET-25)

Conference Place Srinagar India

Conference Date 5th Apr 2025