Autonomous Agricultural Monitoring Robot for Efficient Smart Farming

Abstract

The demand for precise agricultural monitoring grows as the global population increases. Robotic systems have emerged as invaluable tools, particularly in smart farming, with the potential to enhance crop efficiency. This paper presents the implementation of an autonomous robotic monitoring system tailored to agricultural fields. Focused on comprehensive tomato plant monitoring, including ripeness assessment, disease detection, and environmental monitoring, the study addresses the limitations of time-consuming and costly traditional methods. The proposed robotic system offers an affordable and efficient alternative to boost productivity rates. Built on the Robot Operating System (ROS) framework, the robot's navigation system utilizes the A* algorithm and Dynamic Window Approach (DWA) planners. Deep learning techniques were employed (YOLOV5) for autonomous 3-level ripeness classification and detection of nine tomato diseases. The integration of Internet of Things (IoT) technology enabled wireless soil monitoring through the robot. The robot achieved high accuracy in detecting the ripeness level in real-world validation.