Crop-sensing-assisted spraying

Do gap-detection systems make sense in the South African context? By Anna Mouton.

Theoretically, deciduous-fruit growers should be enthusiastic adopters of precision sprayers that promise to improve efficiencies, cut costs, and reduce drift. However, South Africans have been slow to use this technology, despite it not being new.

“Even in the European Union, we haven’t seen as much uptake of precision application systems as we would expect,” said Philip Rebel, crop-protection specialist at ProCrop, speaking at the 2026 Hortgro Technical Symposium.

When the European Food Safety Authority (EFSA) asked European farmers to rank their considerations when buying a sprayer, good deposition was number one. Cost came second, followed by durability, operator safety, pesticide savings, and drift reduction.

“So, they take the easiest and cheapest way to reduce drift, by installing drift-reducing nozzles that meet the minimum drift-reduction threshold,” said Rebel.

Precision application systems are not the easiest and cheapest sprayers, but do they offer a good return on investment? Or are there other reasons for South African growers to consider them?

To investigate these questions, Rebel, in collaboration with AFGRI and Dutoit Agri, conducted trials of a commercially available gap-detection system.

What is crop-sensing-assisted spraying?

Crop-sensing-assisted variable-rate spraying combines two technologies. The crop sensor detects canopy-density differences, and the variable-rate sprayer adjusts spray delivery to match them.

Ultrasonic, infrared, RGB (visible light), or LiDAR (Light Detection and Ranging) sensors can be mounted on the sprayer for real-time scanning. Alternatively, canopy-density differences can be identified by analysing pre-application satellite images.

The crop-sensing data is processed into information that the spray-application system can understand. For example, the sprayer nozzles can close when a gap in the canopy occurs, reducing waste and drift.

“Obviously, the foremost goal of canopy sensing and variable-rate spraying is good deposition because good deposition correlates with biological efficacy,” said Rebel. “You can see this on your farm. If sprayer calibration and setup aren’t optimal, you have problems, even though you use good products and application timing.”

For his trials, Rebel evaluated gap detection using the Smart Apply Intelligent Spray Control System distributed by AFGRI. It scans the orchard with LiDAR in real time, using the data to control solenoids that open and close the spray nozzles. It also has GPS, which enables it to adjust the application rate based on forward speed.

How did the system perform?

Rebel compared applications in full-bearing apple orchards using the same sprayer but with gap detection enabled or disabled. The first trial was at 100% petal drop when the canopy was sparse. The second and third trials were in December when the canopy was full.

To assess deposition, Rebel applied a fluorescent particle tracer according to tree-row volume. After spraying, he collected leaves from the top, middle, bottom, inside, and outside of the canopy, photographed them under ultraviolet light, and quantified deposition using image analysis software.

Trial 1

In the first trial, the gap-detection system was mounted on a low-profile axial-fan sprayer with 14 nozzles per side. Unfortunately, due to the nozzle configuration, one solenoid controlled two nozzles, reducing the system resolution.

The orchard requirement based on tree-row volume was 1 668 litres per hectare. The low-volume calibration for the trial was 650 litres per hectare, and the forward speed was 4.14 kph.

As expected, deposition was significantly higher on the outside than inside the canopy and at the top of the tree. Deposition in the middle and bottom of the canopy didn’t differ significantly, and there were no significant differences due to the gap-detection system.

“We have a deposition benchmark for biological efficacy,” said Rebel. “In this trial, all our deposition values were above the benchmark.”

Trial 2

The setup was the same as for the first trial, but Trial 2 was conducted when the orchard had a full canopy. Deposition was again significantly higher on the outer canopy than on the inner canopy. However, deposition was no longer highest at the top of the canopy.

“More foliage intercepts more droplets on their way to the top of the canopy,” said Rebel. “Overall spray deposition in this trial was lower due to higher interception by leaves, and the greater amount of foliage that needs to be covered.”

The gap-detection system did not significantly alter deposition in the middle, top, inner, or outer canopies. However, deposition in the lower canopy was significantly less with gap detection enabled than disabled.

As for Trial 1, all deposition values exceeded the benchmark for biological efficacy.

Trial 3

The third trial was also in December, but the gap-detection system was mounted on a high-profile sprayer with 12 nozzles per side, each individually controlled by its own solenoid.

Based on tree-row volume, the orchard requirement was 1 780 litres per hectare. The low-volume calibration for the trial was 600 litres per hectare, and the forward speed was 5.2 kph.

“For this situation, the forward speed and air volume were optimised,” said Rebel. “The system reacted much more to gaps because of the increased number of solenoids.”

Deposition was higher outside than inside the canopy, higher at the bottom than in the middle, and higher in the middle than at the top. “The angle from the top of the sprayer to the top of the trees was too large for proper deposition,” commented Rebel.

The only significant difference between deposition with and without gap detection occurred inside the canopy, where greater deposition was observed without gap detection. As for trials 1 and 2, all deposition values exceeded the benchmark for biological efficacy.

The bottom line

The three trials demonstrated that the gap-detection system doesn’t compromise deposition, which was the first consideration of European growers in the EFSA survey. But what about their second consideration: cost?

To evaluate potential savings, Rebel compared spray volume with and without the Smart Apply Intelligent Spray Control System in a range of blocks in the same orchard group (same sprayer calibration) on Paardekloof.

“The Smart Apply System adjusts the amount sprayed based on gaps and orchard uniformity,” he said. “On average, when we had the gap-detection system on, we saved about 9% of the spray liquid in each orchard.”

For a particular orchard, this figure will depend on variables such as sprayer type and setup, and orchard uniformity and age. Additionally, the trials examined only the effect of gap detection, not that of canopy-density detection, because the background calculations underlying the latter are not well understood.

Rebel raised two concerns about the system. Firstly, he questioned the durability of a precision scanning system and its associated components. For example, branches protruding into the work row occasionally caught on the cables, disabling the solenoids, something that would require careful monitoring in practice. Secondly, the system relies on internet connectivity, which is not always reliable on farms.

While crop-sensing-assisted variable-rate spraying can reduce spray drift, Rebel concluded that employing only gap detection in a high-end canopy-sensing system would not yield an attractive return on investment within a reasonable timeframe.

“It will probably get cheaper as the technology develops,” he said. “But, at the moment, most South African farms can do many other things that do not require much technology to improve their spray applications. We need to focus on the basics.”

*

Watch this presentation on the Hortgro YouTube channel.