Automation and robotics in the cultivation of pome fruit: Where do we stand today?

Using deep learning for pruning region detection and plant organ segmentation in dormant spur-pruned grapevines

Even though mechanization has dramatically decreased labor requirements, vineyard management costs are still affected by selective operations such as winter pruning. Robotic solutions are becoming more common in agriculture, however, few studies have focused on grapevines. This work aims at fine-tuning and testing two different deep neural networks for: (i) detecting pruning regions (PRs), and (ii) performing organ segmentation of spur-pruned dormant grapevines. The Faster R-CNN network was fine-tuned using 1215 RGB images collected in different vineyards and annotated through bounding boxes. The network was tested on 232 RGB images, PRs were categorized by wood type (W), orientation (Or) and visibility (V), and performance metrics were calculated. PR detection was dramatically affected by visibility. Highest detection was associated with visible intermediate complex spurs in Merlot (0.97), while most represented coplanar simple spurs allowed a 74% detection rate. The Mask R-CNN network was trained for grapevine organs (GOs) segmentation by using 119 RGB images annotated by distinguishing 5 classes (cordon, arm, spur, cane and node). The network was tested on 60 RGB images of light pruned (LP), shoot-thinned (ST) and unthinned control (C) grapevines. Nodes were the best segmented GOs (0.88) and general recall was higher for ST (0.85) compared to C (0.80) confirming the role of canopy management in improving performances of hi-tech solutions based on artificial intelligence. The two fine-tuned and tested networks are part of a larger control framework that is under development for autonomous winter pruning of grapevines.

Supplementary Information

The online version contains supplementary material available at 10.1007/s11119-023-10006-y.

A novel approach to combine food production with carbon sequestration, biodiversity and conservation goals

Land use conflict is a major contributor to unsustainable deforestation rates, with agriculture being the primary driver. Demand for agricultural output is forecast to increase for years to come and the associated deforestation is a key driver in global declines of biodiversity. Moreover, deforestation is contributing to instability of agricultural production systems and reduces our ability to mitigate anthropogenically driven climate change. There is urgency in reducing this land use conflict and the cultivation of ectomycorrhizal fungi (EMF) may provide a partial solution. As an example, here we focus on Lactarius indigo, an edible and historically appreciated species with distribution in the Neotropics and Nearctic. Exploring the geographic spread and associated climate preferences, we describe how cultivation of this species can be combined with forest-based biodiversity and conservation goals. Detailing a full methodology, including mycelium production and how to create trees that may produce the fungus, we explore potential benefits. Combing data from the emerging field of EMF cultivation with nutritional studies, we show that a protein production of 7.31 kg per hectare should be possible, exceeding that of extensive pastoral beef production. In contrast to commercial agriculture, L. indigo cultivation may enhance biodiversity, contribute to conservational goals and create a net sink of greenhouse gases whilst at the same time producing a similar or higher level of protein per unit area than the most common agriculture use of deforested land. With such startling and clear benefits, we call for urgent action to further the development of such novel food production systems.

Modeling of Individual Fruit-Bearing Capacity of Trees Is Aimed at Optimizing Fruit Quality of Malus x domestica Borkh. 'Gala'

The capacity of apple trees to produce fruit of a desired diameter, i.e., fruit-bearing capacity (FBC), was investigated by considering the inter-tree variability of leaf area (LA). The LA of 996 trees in a commercial apple orchard was measured by using a terrestrial two-dimensional (2D) light detection and ranging (LiDAR) laser scanner for two consecutive years. The FBC of the trees was simulated in a carbon balance model by utilizing the LiDAR-scanned total LA of the trees, seasonal records of fruit and leaf gas exchanges, fruit growth rates, and weather data. The FBC was compared to the actual fruit size measured in a sorting line on each individual tree. The variance of FBC was similar in both years, whereas each individual tree showed different FBC in both seasons as indicated in the spatially resolved data of FBC. Considering a target mean fruit diameter of 65 mm, FBC ranged from 84 to 168 fruit per tree in 2018 and from 55 to 179 fruit per tree in 2019 depending on the total LA of the trees. The simulated FBC to produce the mean harvest fruit diameter of 65 mm and the actual number of the harvested fruit >65 mm per tree were in good agreement. Fruit quality, indicated by fruit's size and soluble solids content (SSC), showed enhanced percentages of the desired fruit quality according to the seasonally total absorbed photosynthetic energy (TAPE) of the tree per fruit. To achieve a target fruit diameter and reduce the variance in SSC at harvest, the FBC should be considered in crop load management practices. However, achieving this purpose requires annual spatial monitoring of the individual FBC of trees.

Advances in Agriculture Robotics: A State-of-the-Art Review and Challenges Ahead

The constant advances in agricultural robotics aim to overcome the challenges imposed by population growth, accelerated urbanization, high competitiveness of high-quality products, environmental preservation and a lack of qualified labor. In this sense, this review paper surveys the main existing applications of agricultural robotic systems for the execution of land preparation before planting, sowing, planting, plant treatment, harvesting, yield estimation and phenotyping. In general, all robots were evaluated according to the following criteria: its locomotion system, what is the final application, if it has sensors, robotic arm and/or computer vision algorithm, what is its development stage and which country and continent they belong. After evaluating all similar characteristics, to expose the research trends, common pitfalls and the characteristics that hinder commercial development, and discover which countries are investing into Research and Development (R&D) in these technologies for the future, four major areas that need future research work for enhancing the state of the art in smart agriculture were highlighted: locomotion systems, sensors, computer vision algorithms and communication technologies. The results of this research suggest that the investment in agricultural robotic systems allows to achieve short—harvest monitoring—and long-term objectives—yield estimation.