Row Navigation using LiDAR in Autonomous Agricultural Vehicles

Agricultural automation has become indispensable for increasing efficiency and productivity while reducing labor dependency and minimizing operational errors. However, autonomous navigation of agricultural vehicles faces significant challenges, especially in perennial crops, where GNSS signal degradation compromises positional accuracy and operational safety. In this context, LiDAR sensors emerge as an efficient solution to complement positioning systems, enhancing obstacle detection and vehicle guidance in the field. This work presents the development and validation of a LiDAR-based plant row detection algorithm applied to autonomous navigation in agricultural environments. The system was implemented on a mobile robot controlled by an algorithm designed to identify obstacles and define the optimal trajectory between plant rows, ensuring the robot’s centralization on the route and avoiding collisions. The tests were conducted in a simulated environment, where PVC pipes of different diameters were arranged to represent crop plants. The algorithm processed LiDAR data to identify row boundaries and dynamically adjusted the robot’s trajectory based on a geometric obstacle-free triangle model, ensuring continuous and safe navigation. The results demonstrated that the system successfully navigated the plant rows without collisions, in both straight and curved trajectories, validating its effectiveness. The proposed approach can be integrated into autonomous agricultural vehicles, contributing to the advancement of precision agriculture and improving field navigation.