Autonomous Robotic Pruning in Orchards and Vineyards: a Review

TL;DR

The paper surveys a decade (2014–2024) of autonomous robotic pruning in orchards and vineyards, addressing labor-cost pressures and uneven adoption of large machinery. It analyzes advances across perception, skeletonization, pruning-point estimation, simulation, and hardware, highlighting the shift from classical vision to deep learning, 3D reconstruction, and learning-based control. It emphasizes modular system designs, end-to-end demonstrations, and the remaining challenges in cost, speed, data requirements, and sim-to-real transfer. The findings suggest that integrated, robust pruning robots with advanced perception and adaptable hardware can bridge manual and mechanized operations, enabling precise, scalable pruning in diverse terrains.

Abstract

Manual pruning is labor intensive and represents up to 25% of annual labor costs in fruit production, notably in apple orchards and vineyards where operational challenges and cost constraints limit the adoption of large-scale machinery. In response, a growing body of research is investigating compact, flexible robotic platforms capable of precise pruning in varied terrains, particularly where traditional mechanization falls short. This paper reviews recent advances in autonomous robotic pruning for orchards and vineyards, addressing a critical need in precision agriculture. Our review examines literature published between 2014 and 2024, focusing on innovative contributions across key system components. Special attention is given to recent developments in machine vision, perception, plant skeletonization, and control strategies, areas that have experienced significant influence from advancements in artificial intelligence and machine learning. The analysis situates these technological trends within broader agricultural challenges, including rising labor costs, a decline in the number of young farmers, and the diverse pruning requirements of different fruit species such as apple, grapevine, and cherry trees. By comparing various robotic architectures and methodologies, this survey not only highlights the progress made toward autonomous pruning but also identifies critical open challenges and future research directions. The findings underscore the potential of robotic systems to bridge the gap between manual and mechanized operations, paving the way for more efficient, sustainable, and precise agricultural practices.

Figures (15)

• Figure 1: Pie chart of the distribution of the addressed crops in the analyzed papers.
• Figure 2: Distribution of the analyzed papers in the period of time from 2014 to 2024. The orange bars represent the paper using machine learning, showing a neat increase in recent years.
• Figure 3: Scheme of the autonomous robotic pruning pipeline. The above part of the scheme depicts the perception-related tasks, while the bottom part shows the tasks related to manipulator design and control.
• Figure 4: This plot shows the number of papers addressing individual tasks in the autonomous pruning pipeline (bottom left) and combinations of tasks (top). The manipulator and end-effector design tasks are grouped together, as they are consistently paired in the reviewed papers.
• Figure 5: Overview of four different tasks of computer vision for autonomous pruning: (a): detection of branches on pseudo-color image extracted from PCD zhang2018branch, (b): semantic segmentation of different parts of cherry trees you2022semantics, (c): bud detection on grapevine oliveira2024enhancing, (d): panoptic segmentation of the structure of different parts of grapevine and different budswilliams2023modelling.