Navegación autónoma de un UGV para la detección de cacaos enfermos y sanos

Abstract

This project presents an agricultural control and monitoring system designed for a differential robot, based on software architecture developed in ROS2. This system is designed to enable autonomous and precise navigation within a cacao plantation, under the supervision of an operator. The project integrates GPS-based navigation algorithms, allowing the robot to move efficiently and record relevant crop information in real-time. In addition, an artificial vision system based on deep neural networks was designed, capable of identifying cacao fruits with signs of disease. The information from each detection is associated with geographical coordinates, creating a map of critical points within the plantation. This way, the farmer can accurately know the location and condition of the affected fruits, aiding in timely decision-making Validations are presented within a simulated cacao plantation environment, demonstrating the system's ability to maintain reliable detection. This system contributes to the advancement of robotics applied to precision agriculture, highlighting the use of autonomous robots as effective tools for monitoring and early diagnosis in cacao crops. Keywords: Agricultural monitoring, differential robot, ROS2, simulated environments