Autonomous Navigation in Vineyards with Deep Learning at the Edge

Abstract

With the rapid growth of the world population over the past years, the agriculture industry is asked to respond properly to the exponential augmentation of global demand for food production. In the past few years, autonomous agricultural field machines have been gaining significant attention from farmers and industries in order to reduce costs, human workload, and required resources. Nevertheless, achieving sufficient autonomous navigation capabilities requires the simultaneous cooperation of different processes; localization, mapping, and path planning are just some of the steps that aim at providing to the machine the right set of skills to operate in semi-structured and unstructured environments. In this context, the presented research exploits later advancement in deep learning and edge computing technologies to provide a robust and fully integrable local planner for autonomous navigation along vineyards rows. Moreover, the devised and tested platform necessitates only of low range and low-cost hardware with minimal power and bandwidth requirements. The machine learning algorithm has been trained and tested with acquired images during different field surveys in the north region of Italy. Then, after performing an optimization process, the overall system has been validated with a customized robot platform in the appropriate environment.