AI-powered autonomous spraying robot for precision orchard applications

In this study, an electric and autonomous orchard spraying robot, named OrcBOT, was developed, modeled, and prototyped for precision orchard spraying. The system integrates electrostatically charged nozzles with YOLOv5-based real-time canopy detection, enabling highly precise and variable-rate pesticide application through independent nozzle control. Spraying operations are coordinated using stereo cameras and RTK-GPS navigation, while nozzle activation is managed by a central electronic control unit based on canopy structure. The robot is capable of both remote-controlled and fully autonomous operation, with monitoring and control accessible via smartphone and tablet applications. Field trials conducted in apple orchards using food dye as a tracer demonstrated an average droplet size of 150-170 µm, classified as fine spray according to ASAE S572.1. Canopy coverage averaged 55%, reaching up to 57% under optimal operating conditions (2 bar, 1 km/h, 10 kV). These findings demonstrate the effectiveness of OrcBOT in fine pulverization applications and underline its potential as a sustainable and practical solution for precision orchard spraying.