Precision farming creates a buzz as growers try new technology
Adrone flies over the orchard. The small robot, loaded with sensors and cameras, moves in a grid pattern high above the trees.
It picks up things the human eye cannot see: the nutrient content of trees, the health of the soil, how much water there is in parts of the orchard.
With this information, farmers can tell which areas of their farms need watering, which need fertiliser — and they can estimate yields.
Precision farming could be the future in SA, farmers and technology specialists say.
Agriculture in the last quarter helped pull the country out of a technical recession and contributed 0.7 percentage points to the 2.5% GDP growth in the second quarter.
But until the first quarter of 2017, agriculture’s GDP contribution had been declining since the end of 2014.
Declining farming profit-ability and water scarcity — drought, declining rainfall or overdemand for water — has left SA with less than two-thirds of the number of farms it had in the early 1990s, the World Wildlife Fund says in its report Agriculture: Facts and Trends in SA.
Climate change is expected to make water increasingly scarce in coming decades, pushing farmers to better manage their water, fertiliser and crop resources.
"There is quite a bit of precision farming in SA and it is developing rapidly," says Raymond Auerbach, professor of soil science and plant production at Nelson Mandela University in Port Elizabeth.
"The pros are that it allows very accurate application of fertiliser and chemicals — avoiding overapplication — and can also help with water management," Auerbach says.
The cons are that the infrastructure for technology-aided farming can be expensive.
However, as it is a vast emerging field, no one is sure how widely precision farming has penetrated in SA. Some call it "precision agriculture", some "satellite farming", others "site-specific crop management" or "digital agriculture".
Precision agriculture uses new technology — sensors, satellites, drones and GPS — to gather more data about a farm or crop, and the data is used to make more informed decisions that are tailored to a specific farm, part of a farm and crops.
"Grain and oilseed producers in SA are leaders in the uptake of precision farming in the world," says Corné Louw, a senior economist at GrainSA.
Although it is expensive to initially acquire the right equipment, every producer must make decisions in terms of his own financial abilities
"The problem is that it is very difficult to quantify," Louw says.
This is because precision farming can range from using GPS tracking to steering equipment, or collecting nutrient information with sensors from specific plants so that farmers know how much fertiliser their crops require.
Jozua du Plessis, GrainSA’s grain producer of the year for 2016, is a firm advocate of precision farming. "To remain competitive with grain producers in the rest of the world, it is essential," he says.
"Although it is expensive to initially acquire the right equipment, every producer must make decisions in terms of his own financial abilities." His farm in Delmas uses precision farming to keep the soil nutrients stable across its hectares to improve productivity — and thus yields and profits — and to keep better records. "We have to measure to know," he says.
While Du Plessis imported systems to precision farm, several local companies are now offering services.
Aerobotics, a Cape Town-based start-up, is one.
"It doesn’t make sense to blanket-apply treatment — whether it’s fertiliser or irrigation — across your crops," says its co-founder and chief technology officer, Benji Meltzer.
"Precision agriculture says ‘let’s get more precise and use that data to become more prescriptive’," Meltzer says
"It can also track performance, pick up on a problem before you see it with the naked eye. These [commercial] farms are massive.
"It is also possible to track how crops are doing, in terms of health, to get a better understanding of potential yields at the end of the season," he says.
Initially, Aerobotics was a drone company that used computer software, but "now we use any form of aerial data, drones are just one form".
Imaging services for agriculture is a competitive space, he says. The company provides services to about 100 farmers.
Demand is picking up again after a decline: precision farming in SA fell prey to the hype that often follows the emergence of a new technology, Meltzer says.
American consultancy Garner describes technology development as having five stages: the "technology trigger", characterised by a sharp spike in visibility; the technology reaching a "peak of inflated exceptions", the greatest visibility it will obtain, before plummeting into the "trough of disillusionment"; and following an arduous climb up the "slope of enlightenment", a technology reaches a "plateau of productivity".
There has been a boom in sensor-based and remote-sensing technology — agriculture is just one of the applications — and a sharp rise in the number of companies offering sensor-based solutions.
However, having the technology and the data it generates do not necessarily mean that it will suit a farmer’s needs. This resulted in farmers being overwhelmed with the technology’s applications and creating disillusionment, Meltzer says.
"The feeling became quite negative because people had been oversold … but that is changing," he says.
"There was a lack of education about what this technology is and isn’t capable of. When drone technology initially came on the scene, everyone knew it had potential, but no one knew exactly what."
A major aspect of new technology is the algorithms that make the data useful.
One way to do this is with artificial intelligence that allows a computer to perform tasks that were usually only within the ambit of human skill, such as sensing, optimised decision-making and learning.
A current Aerobotics project uses artificial intelligence processes to identify and map trees in an orchard.
"There’s an object recognition or classification [algorithm] that will identify each tree and give you a count across a range of metrics, such as tree height, canopy depth, health of the trees," Meltzer says.
Each tree has unique GPS co-ordinates, and "each time you fly a drone over the orchard, you can pick up outliers and anomalies", he says.
The team at Aerobotics — 13 people, 11 of whom are engineers — is also undertaking research to predict individual tree fruit yields.
"There’s been an explosion on the tech side," Meltzer says.
"The applications have started becoming apparent, and industry is waking up to it."