Localization, Navigation and Activity Planning for Wheeled Agricultural Robots – A Survey

摘要

High cost, time intensive work, labor shortages and inefficient strategies have raised the need of employing mobile robotics to fully automate agricultural tasks and fulfil the requirements of precision agriculture. In order to perform an agricultural task, the mobile robot goes through a sequence of sub operations and integration of hardware and software systems. Starting with localization, an agricultural robot uses sensor systems to estimate its current position and orientation in field, employs algorithms to find optimal paths and reach target positions. It then uses techniques and models to perform feature recognition and finally executes the agricultural task through an end effector. This article, compiled through scrutinizing the current literature, is a step-by-step approach of the strategies and ways these sub-operations are performed and integrated together. An analysis has also been done on the limitations in each sub operation, available solutions, and the ongoing research focus.