Sensors, Algorithms, and Software for Autonomous Navigation of Agricultural Mobile Robots

摘要

The introduction of autonomous mobile robots for agricultural applications represents a technological shift towards more efficient farming practices with lesser dependence on the human workforce. This chapter provides a comprehensive overview of the “Sense, Think, Act” paradigm, with emphasis on some of the hardware components and software packages that are most commonly used in the autonomy of agricultural mobile robots. A relatively detailed insight into various types of sensors is provided, highlighting their use in localization, mapping, Simultaneous Localization and Mapping, and obstacle detection. Different algorithms for path planning, waypoint following, sensor fusion, and collision avoidance are reviewed, along with their key features for ensuring a reliable and safe navigation within agricultural fields and orchards. Furthermore, some of the available packages from the Robot Operating System, followed by a list of robotic simulators, are included to address their applications in algorithm development, testing phase, and validation for real-world deployment. The last section of the chapter discusses the technological limitations, challenges, and potential opportunities related to the commercialization of these robots.