The Western Cape province in South Africa has been in making headlines of late, with its novel ways of adopting drone technology in its agricultural industry.

Just this past week, agricultural services company Aerobotics revealed how it was working with a partner to deploy drones on a mission to destroy a pest problem at one of the farms, through distributing pest-eating bugs around vineyards at one farm.

Now, a rising agricultural drone start-up in the province has just revealed its recent task, whereupon it was called in to deal with the red-eyed wattle shrub – locally known as rooikrans – which was proving to be a big problem at a farm in the Garden Route District.

The manager at the 3000-hectare floricultural farm said he turned to drone services supplier Integrated Aerial Systems because the red-eyed wattle infestation was proving resistant to other pest elimination efforts.

“Initially introduced to stabilise sand dunes along South African coastal areas, the red-eyed wattle has spread rapidly, threatening local ecosystems, biodiversity, and the economy,” IAS said in a report produced after the operation.

“Its aggressive growth forms dense thickets that outcompete and displace native vegetation, disrupting natural habitats. On the floricultural farm, the red-eyed wattle has encroached on valuable agricultural land and displaced indigenous plants.

“Traditional control methods, such as cutting and controlled burns, have proven ineffective, while biological control methods using tractors and manual efforts have been both tedious and inadequate.”

The challenge was to develop an efficient strategy to manage this invasive species while preserving the surrounding fynbos environment.

The drone start-up was confident that its advanced drone technology would offer a stronger solution, centred around precise identification and targeted management of the infestation; which would only deal with the red wattle and leave the other indigenous vegetation unaffected; as the farm manager had requested.

So, naturally, IAS first took survey drones up around a sample of the affected area to take stock of the situation.

“An aerial approach provides a precise and comprehensive view, significantly reducing the time required compared to traditional land-based scouting methods,” the company says.

“We proposed mapping a test area first, using two different sensors, incorporating artificial intelligence (AI) algorithms to pinpoint the locations of the invasive plants. By integrating precise aerial imagery with geospatial AI, IAS aimed to improve the accuracy and efficiency of the identification process, ensuring that only invasive species are targeted and minimising the impact on the surrounding environment.

“To evaluate the identification process, a 20-hectare section of the farm, characterised by varying densities of red-eyed wattle along with fynbos, was set aside for detailed mapping. This area, with its diverse levels of infestation, was chosen to enable a thorough analysis of all potential infestation scenarios.”

For this part of the operation, the young company used the DJI Phantom 4 Pro RTK and the DJI M210, equipped with a multispectral sensor.

The DJI Phantom 4 captured high-resolution RGB imagery, which provides a clear visual representation of the area suitable for general mapping purposes.

The rugged M210 meanwhile, with its multispectral sensor collected data across multiple spectral bands, gave the team detailed insights into plant health and species composition but requiring specialised equipment and involving higher costs.

“Using both drones allowed for a comparative evaluation of their effectiveness in identifying the invasive species. The mapping was conducted as close to noon as possible to minimise the effects of shadows and ensure the clarity and accuracy of the imagery.”

After collection, the drone data was fed into a software program to generate orthomosaics, which resulted in two key products: an RGB orthomosaic from the Phantom 4 and a multispectral orthomosaic from the M210. The data analysis involved AI-powered tools designed to streamline the identification and classification of areas on an orthomosaic.

“These tools significantly enhance resource efficiency by accurately identifying and classifying areas for spot spraying. The process results in a high-accuracy prescription map that can be generated quickly.”

The data analysis returned the results that showed that the AI tools had effectively identified the red-eyed wattle by recognising its darker green hue, denser canopy, and greater height.

“We are also well-equipped to facilitate targeted herbicide applications, minimising the impact on non-target species and ensuring effective removal of the invasive plants. Additionally, the total affected area can be quantified and exported, providing a comprehensive report of the treatment operation.”

IAS went on to add that they used a geospatial AI tool that allows for fine-tuning identified areas by editing shapes, creating buffers around detected regions, adding areas of interest, or removing non-targeted zones.

These custom shapes can then be exported as files and uploaded directly to the spraying drone’s controller, ensuring precise and targeted herbicide application, and significantly enhancing resource efficiency. By concentrating on areas with invasive species, the tool reduces the impact on surrounding native vegetation and decreases herbicide use, potentially saving up to 80 percent in treatment resources.

At the completion on the operation, IAS noted the following; “The study’s findings provided the farm manager with valuable insights and practical tools for managing the invasive red-eyed wattle on the property.

“The use of drone technology and geospatial AI tools enabled precise identification and mapping of the infestation. The high-resolution imagery allowed for accurate classification of problem areas, crucial for planning targeted herbicide applications. This precision optimised resource use, significantly reducing the quantity of herbicide required, while ensuring that native vegetation and the surrounding ecosystem were minimally impacted.

“At Integrated Aerial Systems (IAS), we specialise in helping farmers identify invasive species through the use of advanced drone aerial imagery and AI analysis. Our workflow enables us to provide detailed maps and targeted spraying that improve invasive species management.

“By applying this method farmers can effectively target and manage problem areas, preserving their land’s natural landscape and biodiversity while maintaining agricultural productivity. The successful implementation of these technologies offers a reliable and efficient strategy for ongoing invasive species control, ensuring sustainable farm management and environmental conservation.”