Articles

Targeted Blossom-Green Fruit Thinning with Variable Rate Sprayer

Blossom or green fruit thinning is a critical task for crop load management in apple production.

Save for later Print

Updated:

May 16, 2023

Skip to the end of the images gallery

Image: Long He, Penn State

Skip to the beginning of the images gallery

Normally, as high numbers of flowers per tree set fruit, a reduction in the number of fruits per tree by thinning is necessary, both for optimal fruit development and good return bloom. Thinning can be carried out during bloom (flower thinning), but it is more often done later (fruit thinning) because growers want to be certain that fruit set is adequate. Manual thinning, chemical thinning, and mechanical thinning are typically the ways that growers used to thin the fruit crops for both blossom and green fruit thinning.

The primary goal of this study is to conduct targeted and precision spray for blossom/green fruit thinning with better efficacy of chemical usage and crop load management. In the year 2022, a robotic apple blossom thinning system was developed to apply target and precision spraying. The outcomes of the study provided guideline information for developing a full-scale robotic chemical thinning system for apple growers.

Robotic Chemical Blossom Thinning System Development

A cartesian robotic system was designed for precision apple blossom thinning consisting of three major components: 1) a color stereo camera (ZED2, Stereolabs Inc., San Francisco, CA) based machine vision system, 2) a cartesian system-based manipulator, and 3) a spraying end-effector with a solenoid valve (55295-1-12, TeeJet Technologies, Glendale Heights, IL) and flat-shaped nozzle (AIXR11003, TeeJet Technologies, Glendale Heights, IL), shown in Figure 1.

Figure 1. Overall description of the Cartesian robotic spraying system and its three major components. Image: Long He, Penn State

Machine Vision System for Flower Cluster Detection

An algorithm was developed to process the images acquired by the machine vision system. Figure 2 shows an example of the detection results obtained with the developed Mask R-CNN with instance segmentation approach. The original output of the instance segmentation model contained pixel-leveled masks together with a bounding box indicating that flower clusters were detected (Figure 2, left). In this study, since the most critical information needed from the vision system was the pair of x-coordinates of each flower cluster, which can be directly extracted from the bounding box, the visualization code was modified so that it only showed the bounding box in the output images (Figure 2, right).

Two photos of the segmentation of apple flower clusters — Figure 2. Mask R-CNN-based instance segmentation of apple flower clusters, a) Original output with predicted masks and bounding boxes included, b) Bounding boxes were extracted. Image: Long He, Penn State

Field Test and Evaluation

Three different treatments, cartesian robotic sprayer (CRS), boom-type sprayer (BTS), and air-blast sprayer (ABS), were tested on the same day for precision apple blossom thinning. In total, 90 trees were reserved for the experiment, including 54 experiment samples and 36 buffer trees. Three consecutive trees were selected as one experiment sample, and six replicates were used for each treatment. To prevent chemical deposition from other treatments, two additional buffer trees were included at the beginning and the end of each experiment sample. All the other trees within the same row were left as un-thinned control (UTC) to help compare the effectiveness of thinning. Due to the height limitation of the cartesian manipulator, the CRS only covered one section (two limbs) of apple trees at a time, while both BTS and ABS were eligible to cover the entire canopy.

Three photos showing the cartesian robotic sprayer, a boom-type sprayer, and an air-blast sprayer — Figure 3. Left: Cartesian robotic sprayer (CRS); Middle: Boom-type sprayer (BTS); Right: Air-blast sprayer (ABS). Image: Long He, Penn State

The proposed cartesian robotic sprayer (CRS) was compared with an air-blast sprayer (ABS) and a boom-type sprayer (BTS) in the application of apple blossom thinning. Three measurements were recorded during the field test: time used for spraying, usage of chemical (lime-sulfur), and the number of green fruit set before June Drop. Table 1 listed the performance of the robotic sprayer by comparing it to the other two systems. The chemical usage for the robotic sprayer system was less than 50% of the other two sprayers. Since we only had one nozzle in this robotic sprayer, the time per tree was much longer than the other two. In the future, if multiple nozzles are used to cover the entire tree height at a time, then the spray efficiency can be greatly improved.

Table 1. Comparison among three different sprayers for spraying time and chemical usage
Spraying System	Average Spray Time per Tree (s)	Total Chemical Usage (gal)
Cartesian Robotic Sprayer (CRS)	18.63	0.6
Air-blast Sprayer (ABS)	2.42	1.8
Boom-type Sprayer (BTS)	3.28	1.1

Summary

Under the cartesian robotic spraying thinning approach, an average of 2.4 flowers remained in each cluster and turned to green fruits, which met the expectations of this experiment. There was no significant difference in the number of green fruits between CRS and ABS, while both of them obtained much less green fruit set compared to BTS and UTC. The results indicated that the CRS was able to achieve relatively higher effectiveness of thinning while reducing the usage of chemical thinner by over 50%.

Agenda

Authors