Agri-Sense: Autonomous Tomato Farming Robotic Vehicle

Abstract

The precision greenhouse cultivation of tomatoes is often hindered by the expensive labor costs, the easy spread of diseases, and the unavailability of robust digital technologies. This paper proposes Agri-Sense, a comprehensive autonomous robotic system that covers the entire scope of greenhouse management. The system includes a robotic platform that allows the crops to be traversed autonomously without the need for human input. The ripeness and leaf disease detection are done using advanced YOLO-based computer vision techniques. The robotic arm has 4 degrees of freedom, which includes a modular gripper and a precise sprayer system for disease treatment. This reduces the usage of chemicals and protects the healthy crops from disease. To solve the problem of connectivity in rural greenhouse environments, the system includes a sensor mesh using LoRa sensors and a Software-Defined Networking controller. This controller uses Quality of Service and VPN tunnels to provide robust communication services even in the presence of unreliable connectivity. Experimental results showed that the proposed system bridges the gap between passive monitoring and autonomous greenhouse management. The system is robust and efficient in greenhouse management.