AI-Powered Fruit Harvesting System Using a Robotic Arm for Precision Agriculture

This paper introduces an AI-based fruit harvesting system designed to improve efficiency and reduce labor demands in agriculture. Integrated with the Smart Agricultural Robot Bulldog (SARDOG), the system utilizes a six-degrees-of-freedom robotic arm controlled by a Raspberry Pi 5 and an NVIDIA Jetson Orin for advanced fruit detection and localization. The YOLOv8 deep learning model facilitates real-time identification and harvesting of fruits through precise kinematic control and seamless hardware communication. A robust dataset was compiled and extensively trained to ensure high accuracy in fruit classification and positioning. Experimental results confirm the system’s ability to accurately detect and harvest various types of tree fruits, significantly optimizing the fruit-picking process. This integration into SARDOG provides a stable, efficient platform for agricultural automation, marking a significant advancement in precision farming. The research highlights the transformative potential of AI and robotics in addressing global food demands while promoting sustainable agricultural practices.