Reinforcement Learning Agents in Precision Agriculture

In the domain of precision agriculture, the convergence of Artificial Intelligence and precision techniques has revolutionized farming practices. This, has marked an era of enhanced efficiency, productivity, and sustainability through exploiting scientific fields such as Machine Learning, Big Data, Computer Vision, Internet of Things and Robotics being applied in agriculture more and more. Given the frequent need for ongoing decision-making inside dynamically changing environments, precision agriculture emerges as a prime domain for the application of Reinforcement Learning techniques, with multiple future promising outcomes. We analyze the literature and identify three main application categories that have been focused on, namely crop management, water management and robot control. Moreover, we highlight the importance and advantages of exploiting Reinforcement Learning in agriculture and highlight the gaps that exist in the literature. Lastly, we propose a general framework towards the future applications of Reinforcement Learning in the field of agriculture.