Development of a Lidar-Based Automatic Steering System for Agricultural Robotics

This thesis proposes a new steering system for agricultural machinery with the aim of improving the automation capabilities of farming robots. Accurate and reliable autonomous machinery has the potential to provide significant benefits to the efficiency of farming operations, but the existing systems for performing one of the most essential automation functions, autonomous steering to keep machinery on the proper course, each have drawbacks that impact their usability in various scenarios. In order to address these issues, a new lidar-based system was developed for automatic steering in a typical farm field. This approach uses a two-dimensional lidar unit to scan the ground in front of the robot to detect and steer based on farm tracks, a common feature in many farm fields. This system was implemented and evaluated, with results demonstrating that the system is capable of providing accurate steering corrections.