AI-Assisted Robotic Eggplant Harvester with a Four Degree of Freedom Robotic Arm

Abstract

This research proposes a unique four-degree-of freedom robotic arm intended for eggplant picking, addressing significant difficulties in Sri Lankan agriculture such as growing labor prices, skilled labor shortages, and inefficiencies in manual harvesting. The research intends to build and manufacture a prototype robotic harvester that leverages a YOLOv8-trained model for object recognition to recognise eggplants and calculate their 3D coordinates. These coordinates are processed in Python and transferred to an Arduino microcontroller, where joint angles are determined using inverse kinematics equations based on the arm's specifications. The servo motors are then controlled to position the end effector for accurate harvesting, with a cutting mechanism completing the operation. The equipment is installed on a movable cart equipped with object-following technology for greater mobility and contains storage for gathered goods. Powered by a 3S battery management system with Lithium Ion batteries, the harvester is adapted to criteria acquired from the Palwehera Government Seed Farm, Sri Lanka to ensure its performance in local agricultural circumstances. This unique robotic technology is intended to assist Sri Lankan farmers overcome labor-related issues, cut costs, and enhance output, with future research concentrating on enhancing efficiency and adapting the system for other crops.