Evaluation of Over-The-Row Harvester Damage in a Super-High-Density Olive Orchard Using On-Board Sensing Techniques.

Abstract

New super-high-density (SHD) olive orchards designed for mechanical harvesting using over-the-row harvesters are becoming increasingly common around the world. Some studies regarding olive SHD harvesting have focused on the effective removal of the olive fruits; however, the energy applied to the canopy by the harvesting machine that can result in fruit damage, structural damage or extra stress on the trees has been little studied. Using conventional analyses, this study investigates the effects of different nominal speeds and beating frequencies on the removal efficiency and the potential for fruit damage, and it uses remote sensing to determine changes in the plant structures of two varieties of olive trees (‘Manzanilla Cacereña’ and ‘Manzanilla de Sevilla’) planted in SHD orchards harvested by an over-the-row harvester. ‘Manzanilla de Sevilla’ fruit was the least tolerant to damage, and for this variety, harvesting at the highest nominal speed led to the greatest percentage of fruits with cuts. Different vibration patterns were applied to the olive trees and were evaluated using triaxial accelerometers. The use of two light detection and ranging (LiDAR) sensing devices allowed us to evaluate structural changes in the studied olive trees. Before- and after-harvest measurements revealed significant differences in the LiDAR data analysis, particularly at the highest nominal speed. The results of this work show that the operating conditions of the harvester are key to minimising fruit damage and that a rapid estimate of the damage produced by an over-the-row harvester with contactless sensing could provide useful information for automatically adjusting the machine parameters in individual olive groves in the future.

Plant phenotyping relevance

オリーブ樹の収穫前後の構造変化をLiDARで評価し、加速度計と組み合わせて収穫による植物状態・損傷を測定している。センシングによる植物形態評価が研究の主要な技術的構成要素である。

Code and data availability

The article describes LiDAR point clouds, accelerometer data, and self-developed analysis programs (Python, R, LabVIEW), but provides no public deposit, availability statement, or URL for any dataset, code, or model. Only ORCID profiles and the CC-BY license URL appear, none of which are paper-specific assets.

No evidence-backed public reproduction asset is currently recorded.