New technology options for deciduous-fruit farmers

Embracing technology can help growers farm more efficiently and sustainably. By Jenny Underhill.

“Cigarette lighters are amazing, aren’t they? But what do they have to do with agriculture?” asked Arie van Ravenswaay, Assistant Director for Innovation, Technology Design and Transfer at the Western Cape Department of Agriculture.

Van Ravenswaay explained that a group of factories in Shao Dong, China, collectively produce up to 10 billion lighters annually. These companies have reduced the plastic in their lighters to such an extent that they can fit 1 000 lighters into a carton. The remaining plastic provides just enough pressure to keep the combustible liquid safely inside.

“It’s all about reducing margins and increasing precision as well as maintaining quality. That’s the type of thinking that we want to bring to agriculture,” explained Van Ravenswaay.

Multispectral sensors and drone technology

According to Van Ravenswaay, drones are an important tool in precision agriculture. They are often equipped with multispectral cameras or sensors that monitor crop health, detect disease, and optimise irrigation.

Standard cameras record only red, green, and blue light. In contrast, multispectral sensors capture wavelengths invisible to the human eye, such as near-infrared and shortwave infrared.

Multispectral indices are simple mathematical formulae that compare two or more wavelength bands. For example, the leaf chlorophyll index can estimate leaf chlorophyll content to assess a crop’s photosynthetic capacity.

“A dying leaf reflects light very differently from a leaf that is alive. It is easy to pick up these differences using multispectral indices,” said Van Ravenswaay. “The technology is so sensitive that you can identify when a plant is going into stress days or even weeks before symptoms become visible to the naked eye. Multispectral imaging can help us to intervene before damage occurs.”

A multispectral drone service provider can regularly fly over a farm, enabling growers to develop multispectral maps that document growth over weeks, months, and years. The images can also show the negative impacts of disease or insect damage. However, the images cannot be used to identify the pathogen itself, only its effect.

“When you implement a new technology, you can uncover things that you did not know were possible,” said Van Ravenswaay. He and his team were asked to survey a polluted river system. Although they were not sure what they were looking for, they took drone footage of the area. The standard camera image showed nothing in the river system. However, the multispectral image of the river system showed a plume of polluted water entering the river from a buried pipe.

Innovative spray technology

Drones can be extraordinarily effective for applying chemicals to orchards. The blades of a drone create a vortex that moves inwards continuously, pushing a column of air downwards, which carries the chemical spray with it. The drone’s vortex also induces canopy movement, improving spray coverage.

“What if we could implement a similar spray technology that comes without the hassle of flying?” asked Van Ravenswaay. “Drone regulations are a headache.”

He believes that autonomous rovers are the new generation of tractors. Autonomous rovers can form part of a broader orchard technology system that includes sensors, drones, AI platforms, and monitoring equipment. Information from different sources, such as LiDAR, multispectral and hyperspectral imaging, electrical conductivity sensors, soil probes, and temperature sensors, can be integrated.

“Growers can use the data to perform quantitative measurements while they are spraying and adapt the protocol according to what is happening in the orchard,” said Van Ravenswaay. “The primary goal is early disease detection.”

Creating tools and platforms using AI

The multitude of data collected from sensors, drones, and rovers will become unwieldy. Van Ravenswaay is convinced that AI can be used to develop platforms and tools to help growers make sense of their data. Implementing AI and technology can help growers farm more efficiently, sustainably, and profitably.

“It’s important to know the underlying code and how it works, but you do not need to be a programmer or have a big team,” said Van Ravenswaay. To test the potential of AI, his department worked with Microsoft to develop a retrieval-based AI system that can analyse uploaded documents, answer questions, and perform calculations based on data gathered in the Western Cape.

Many growers believe that the goal of AI is a fully autonomous farming system, but this need not be the case. “In South Africa, I do not believe that autonomous technology should replace people entirely,” stated Van Ravenswaay. “Instead, we should create systems where people, AI, sensors, drones, and autonomous machines work together.”

Technology can continuously collect data, identify patterns, and make recommendations, while people provide experience, judgement, and an understanding of local conditions that machines cannot always interpret correctly. This creates a feedback loop in which technology supports decision-making.

“I firmly believe that South Africans should not simply copy another country’s technology,” said Van Ravenswaay. “We should adopt useful technologies, adapt them to our own farming realities, and build systems that combine human expertise with technological innovation.”

*

Watch this presentation on the Hortgro YouTube channel.