Speaker
Description
Autonomous weeding robots have the potential to significantly enhance agricultural productivity by reducing the labor and chemical inputs required for weed management. The effective design and testing of such robots are crucial for their successful deployment in real-world applications. The goal of this research is to create and assess an autonomous weeding robot, with an emphasis on functional requirements, hardware and software architecture, and an obstacle detection perception system. The first step in the development process is a functional analysis, which uses a FAST (Function Analysis System Technique) diagram to depict the essential requirements. Although the operating system, middleware, and application-specific algorithms are part of the software architecture, the hardware architecture consists of sensors (camera and radar), actuators, and controllers. The perception system employs the YOLO v5 algorithm for obstacle detection. Covering all essential functions ensures efficient guidance in the design process. With the strong performance of the YOLO v5 algorithm in a range of test settings, the perception system exhibits ideal accuracy and reliability in obstacle detection.This study presents a structured approach to the design and evaluation of an autonomous weeding robot. The proposed functional analysis, hardware and software architecture, and perception system testing provide a solid foundation for future improvements and real-world applications. The findings highlight the potential of the robot to improve agricultural efficiency and reduce dependence on manual labor and chemical weeding methods.
