Stereoscopic plant-protection system integrating UAVs and autonomous ground sprayers for orchards

Front Plant Sci. 2022 Oct 31:13:1040808. doi: 10.3389/fpls.2022.1040808. eCollection 2022.

Abstract

For orchard plant protection, conventional large machines and small sprayers are practically restricted by either narrow planting intervals with dense leaves or their inadequate penetration power, which leads to an unsatisfactory effect of spray. This paper proposes a stereoscopic plant-protection strategy that integrates unmanned air and ground sprayers to spray different parts of canopies to improve uniformity. In order to verify the proposal, a stereoscopic plant-protection system (SPS) was developed, consisting of a small swing-arm sprayer and a T16 plant-protection Unmanned Aerial Vehicle (UAV). Then, optimal operation parameters were determined by Computational Fluid Dynamics (CFD) and orthogonal experiments, and the uniformity was finally quantified by trials. CFD and orthogonal experiments showed that a swing-arm angle of 60° and a forward speed of 0.4 m/s were optimal for the ground sprayer, whilst a height of 2.0 m from the top of canopies and a forward speed of 1.0 m/s were appropriate for the UAV. The trial results showed that the density of vertical droplet deposition varied from 90 to 107 deposits/cm2 in canopies, and the uniformity was 38.3% higher than conventional approaches. The uniformity of top, bottom, inside and outside canopies was significantly improved. Meanwhile, the density of droplet deposition on both sides of leaves in all test points exceeded 25 deposits/cm2, able to meet the standard of spray. This study provides a practical approach for uniform pesticide spray to large-canopy fruit trees. Moreover, the high flexibility of plant-protection UAVs and the significant trafficability of small swing-arm sprayers can solve the problem of large machine entering and leaving orchards.

Keywords: UAV; UGV; computational fluid dynamics (CFD); orchard; stereoscopic plant-protection; uniform spraying.