Precision weeding in sugar beet farming: UAV monitoring of robotic systems

Abstract

Weeding robots are revolutionizing sugar beet farming. The performance of these systems is typically evaluated based on weed control efficacy (WCE), crop damage, and herbicide savings. While robots demonstrate WCE comparable to conventional methods, optimizing their performance requires improvements in crop and weed detection, reduced crop damage, and identification of critical operational timings. This study employed high-resolution RGB UAV imagery to evaluate seven robotic weeding strategies by tracking plant status pre- and post-weeding. Results showed that UAV-based analyses effectively identified strategies with the highest crop losses (up to 8%). However, UAV-based statistical analysis closely matched traditional scoring methods in differentiating weeding strategies. These findings underscore the potential of UAV technology as a powerful tool for evaluating and optimizing weed control systems contributing indirectly to the reduction of fungicide use.