Autonomous Litter Detection & Picking Drone

This paper suggests a drone system for automated trash discovery and collection that integrates robotic manipulation and computer vision to combat environmental declination in hard- to- reach places. The YOLOv8 object identification algorithm, which has an average perfection of 89.7 in a variety of outside circumstances, is used by the proposed system for real-time waste identification. Drone is incorporated with a specially designed 3- DOF robotic arm powered by MG996R servo motors to recoup waste. This arm is capable of precise and stable grasping and retrieval operations during flight or stationary hovering. The drone navigates autonomously using GPS and onboard sensors, while the YOLOv8 algorithm continuously processes camera input to detect and track waste. Experimental evaluations demonstrate the system's effectiveness in accurately detecting and retrieving litter across different terrains and conditions.