A System for Efficient Path Planning and Target Assignment for Robotic Swarms in Agriculture

With a growing world population, the demand for food grows accordingly, putting agricultural production under continuous pressure for more productivity. However, traditional farming approaches are reaching their limits, due to constraints in money and human effort. A solution to this problem lies in automation via the use of emerging technologies, which can accelerate procedures and operations without the need for additional human resources. Suitable modern technologies for various tasks of precision agriculture include the use of the Internet of Things (IoT) via lightweight IoT-enabled robotic systems, which can form swarms for operating in parallel. Path planning for these robotic swarms is a complex, NP-hard problem. This paper addresses the problem of efficient path planning of robotic swarms, formulating the problem as a specific type of Vehicle Routing Problem (VRP). Various state-of-the-art algorithms are employed to solve this VRP, in order to decide on the best approach for different agricultural topologies, tasks, and number of robots available. An end-to-end system is proposed and evaluated, using the Internet/Web as an infrastructure and communication medium, taking GPS input data from map providers, identifying and applying the most suitable algorithm for the specific landscape and task; and finally producing GPS coordinates as routes for the robots to follow. Recommendations for further improvements are discussed.