An Autonomous Irrigation Robot Based on ROS2, YOLOv8 and Distributed ESP32 WMN Architecture

This paper presents the design and implementation of an autonomous indoor houseplant irrigation robot operating under the ROS2 framework. The robot prototype integrates locomotion, water delivery, power management, sensors, and high-level control systems. A YOLOv8-based vision system, trained on a custom houseplant dataset, detects and localizes plants for precise irrigation. Robot localization and mapping use ROS2 SLAM Toolbox with 2D LIDAR, for autonomous navigation and irrigation via ROS2's N av2 path planning. Wireless communication is achieved through an ad-hoc mesh network connecting the main computer with distributed microcontrollers. The prototype demonstrates key autonomous functions, as well as acknowledging limitations, presenting opportunities for future enhancements to improve robustness and practical applicability in home automation.