Automated agriculture: Can robots, drones, and AI save us from starvation?

Agriculture has come a long way in the past century. We produce more food than ever before — but our current model is unsustainable, and as the world’s population rapidly approaches the 8 billion mark, modern food production methods will need a radical transformation if they’re going to keep up. But luckily, there’s a range of new technologies that might make it possible. In this series, we’ll explore some of the innovative new solutions that farmers, scientists, and entrepreneurs are working on to make sure that nobody goes hungry in our increasingly crowded world.

In Thomas Malthus’ seminal — though oft criticized — 1798 work, An Essay on the Principle of Population, the economist took a long view of human history, observing that human populations, when they have an abundance of food, grow until they strain their resources, at which point scarcity sets in. “Famine seems to be the last, the most dreadful resource of nature,” he said. “The power of population is so superior to the power in the earth to produce sustenance for man, that premature death must in some shape or other visit the human race.”

Poetic as Malthus’ prediction was, it did not come to pass. Instead, revolutions in industry and agriculture came, and with them, a substantial increase in the production and trade of food. The optimists won, while Malthus became the patron saint of crackpot economists, a ragged preacher on the streets of history.

Although modern economists and tech utopians view Malthus with scorn, his arguments were reasonable given the evidence at the time. Perhaps the new technologies that enabled the tremendous population growth of the last few centuries did not end the possibility of a Malthusian catastrophe, but merely delayed it? Humanity has been able to grow far, far beyond what Malthus could have imagined — rough estimates put the world population in 1800 at 890-980 million people, while today the world population exceeds 7.4 billion — but how long can this trend continue?

The world may need another agricultural revolution to sustain itself.

The UN expects the world population to reach 8.5 billion by the year 2030, and 9.7 billion by 2050. Much of the world’s population is concentrated in China and India; each of these countries currently have more than one billion people, with the population of India projected to surpass that of China by 2022.

New agricultural technologies and techniques have staved off the sharp famines that Malthus predicted. Though once humans had to plant seeds haphazardly by hand, seed drills enabled farmers to sow them in long, uniform lines. With steam-powered tractors, farmers could plow wide swaths of land, without the need for sluggish oxen. Threshing machines cut down the many hours devoted to threshing by hand.

With the world population climbing to ever more staggering heights, and with economic growth allowing for greater consumption, the world may need another agricultural revolution to sustain itself.

One answer to the problem may lie in automation. While agriculture has become increasingly automated over the last 200 years, the process is only accelerating. New technologies, and droves in particular, are allowing farmers to improve the efficiency of their crops and livestock production, as well as quickly neutralize threats such as disease or drought.

Drones may be the farm workers of the future

“There is some stigma to that word,” Dr. Lav Khot, a professor at Washington State University and specialist in site-specific crop management, says of the term “drone.” Many proponents of the commercial uses of drones prefer terms like “unmanned aerial systems,” a longer phrase that lacks the verbal punch. It’s not hard to see why: For many, the word “drone” evokes military hardware, flying machines that drop bombs or spy on targets from above.

Researchers like Dr. Lav Khot use drones for a variety of projects.

Researchers like Khot are experimenting with different ways that drones can be used in agriculture, beating these swords into plowshares. For growers, one of the most exciting uses for drones is crop imaging. Using drones equipped with multispectral sensors, farmers survey their land, taking images that reveal things like the fertility of specific patches of soil, how much water the crops need, and more.

In the past, farmers had to rely on satellite imaging to get such detailed maps of their land, which was a rather time consuming process. Satellite imaging often involves 14 day intervals, Khot told Digital Trends, and problems can arise due to cloud cover or other factors. With drones “you can get it when you want, hypothetically.”

That quick turnaround isn’t just convenient; in cases where plants are affected by disease or pests, it can be the difference between saving a harvest or watching it die. Speaking to The Guardian, Salman Siddiqui, a researcher from the International Water Management Institute, explains how drone imagery can be used to detect diseases and other stressors in plants. “Photosynthetic activity decreases, and that affects the chlorophyll,” he said — and multispectral imaging can detect these changes early, before the problems become so drastic as to produce signs that humans can see.

This is because chlorophyll, the molecule that gives leaves their green hue, draws in visible light, while reflecting a large amount of infrared light. When a plant is unhealthy, it will reflect more of the visible light hitting it, and reflect fewer infrared waves. Infrared imaging can show farmers if their plants are reflecting abnormal amounts of light, tipping them off to problems.

Agriculture, meet automation

One of Khot’s main projects at the moment focuses on irrigation, experimenting with different irrigation techniques and amounts of water, and using drone imagery to track the subsequent health of plants. Take subsurface irrigation, for example. This is a process where tubes that drip water are placed under the surface of the soil, delivering water directly to the roots of crops. Khot and his team adjust the location of the tubes, testing them at 15 or 30 centimeters below the surface of grapevines.

False-color images of a bean-breeding trial captured by a camera mounted on a drone. (Photo: Lav R. Khot/Washington State University & Phillip N Miklas/USDA-ARS)

They adjust the amount of water used as part of their research, and use drone imaging to map the crops, testing “to see if we can go to 60 percent or 30 percent, and still the plant can grow and produce as good a yield.”

Khot also works with irrigation above the surface, testing sprinklers based at low-elevation (no more than a foot above the surface) and moving the sprinklers closer to the canopy, which should reduce the evaporation that occurs as water moves through the air.

Robotic tendrils are spreading into every aspect of farming.

Khot’s research could take on new importance as climate change develops. “In Washington in 2015, we recorded the highest temperature for several years,” Khot said, a development that may be the result of climate change. To prepare for hotter years, Khot and his fellow researchers are testing different varieties of pinto beans to see how they react to differing amounts of water. This should help them discover which beans will be a better investment for farmers if temperatures rise in the future.

Not all uses for drones involve imaging. One project Khot describes is delightfully simple. Wanting to experiment with a larger model of UAS, a Yamaha RMAX — an unmanned helicopter that has been popular in Japan for spraying — Khot and his team came up with a clever solution to a problem that has plagued Washington farmers.

Cherries are one of Washington’s premier crops, and they face a disturbingly common pest: rain. “When the cherry fruit is ripe, it has high sugar content, and the skin becomes very thin,” Khot explains. When rainwater falls on the cherry orchards, it can sit on the canopy, and the thin-skinned cherries will absorb it. As the cherries drink up moisture, they begin to swell and split.

In Japan, the RMAX has been popular as an efficient, unobtrusive crop-duster. (Credit: Yamaha)

Cherry growers can try to minimize the damage from rainfalls by harvesting cherries quickly, or shaking the branches to cast off some of the water. Khot flew an RMAX at low altitudes — 35 to 50 feet — over orchards to disperse water from the canopies. It’s a more efficient method than manually shaking cherry trees by hand or by fan — and much cheaper than hiring a helicopter pilot to come and do it — but not a foolproof one. “The orchards here in Washington State, not all are flat; we’ve got sloping, and so every year there are some accidents with that,” Khot said.

More than just flying robots

Drones look to be a promising tool for farmers going forward, but they are not the only examples of automation sweeping the agrarian world. Robotic tendrils are spreading into every aspect of farming, bringing a cold touch even to that most intimate part of a farmer’s day: the milking of cows.

Machines handling the milking process seems like a win for farmers. Hand milking is rare today — after all, who wants to spend hours a day squeezing cow teats and suffering the jealous glares of calves? — as farmers have been using machines to pump milk for years now. These machines required human input, however; farmers still had to fasten cups to the cow’s teats, and stand by as the machines pumps out milk.

These vehicles could remove the limitations of the human body from a traditionally grueling industry.

With more modern technology, the milking process can be done without even that slight amount of human input. Advanced milking systems allow the cows to approach a milking robot when they feel like it. The cow, as trained by its owner, steps onto a plate, which initiates the milking process. The machine can recognize the cows by their tags, and if a specific cow has not been milked within a certain period of time, the machine will get to work, disinfecting the udder and attaching suction cups to the teats.

The whole process is convenient for farmers, and allegedly comfortable for the cows. Speaking to the BBC, farmer Robert Veich claims the process makes for happier, more productive animals, saying “Cows respond to comfort. It has gone from 28 litres to 36 litres average a day, without a significant increase in feed cost.”

Dairy farms aren’t the only place that automation has taken hold, either. Lately, it’s even begun to hit fruit orchards – which have been remarkably resistant to mechanization over the years. Until recently, fruit picking has managed to withstand the march of automation, largely because fruits are delicate, and machinery could damage the crops or the trees that produce them. A 2011 issue of Migration News claims that most apple picking is still done by hand, and that the majority of laborer’s time is spent not picking the apples, but rather moving the ladders used to reach them and hauling bags of produce back and forth. As Abundant Robotics CEO Dan Steere explained to Digital Trends, “the main problems are that fruit is hard for computers to see, and it’s delicate. Until now it has not been possible to either reliably identify produce or automate harvesting without damaging the produce.”

Automated tractors can cover wide swaths of land without human input, freeing up labor. (Photo: Case IH)

Despite the difficulties, engineers are looking for ways to make picking more efficient through automation. Steere’s company is working on a machine to pick apples; another company, Energid, has built a machine to pick oranges. Abundant Robotics’ apple picker uses a vacuum tube to suck apples off trees, while Energid’s orange harvesting robot shakes fruit from trees and catches them.

Prehaps the most iconic farming vehicle, the tractor, is undergoing a transformation of its own. Autonomous tractors, such as Case IH’s Autonomous Concept Vehicle, could replace human-operated tractors on many farms. The sleek machine does not even have a seat for a driver. It can travel along predetermined routes programmed by its operator, who can track the tractor’s movements and reroute it, if need be, with a tablet app. The tractor can even sense obstacles in its path, stopping to avoid collisions. Autonomous tractor manufacturers such as New Holland even claim that the vehicles will eventually be able to react to changes in the weather.

Running day or night, automation can make farming a 24/7 process. (Credit: Case IH)

Autonomous farm equipment isn’t simply a convenience, allowing farmers to kick back and sip coffee as robots do their work. Nor is it merely a tool to cut down on the costs of labor. These vehicles could remove the limitations of the human body from a traditionally grueling industry. Machines stand no risk of injury or maiming, only damage that can be repaired. Even more intriguing is that farming could become a 24-hour process, uninterrupted by darkness or the need for sleep.

The traditional image of a farmer as grizzled man in a Carhartt jacket, gripping the wheel of his John Deere with calloused hands, may soon fade into history. In his place, a buttoned-up man managing a fleet of machines with his iPad from the comfort of an office.

Will it be enough?

All these new technologies are dazzling, but the question remains: Will they be enough to sustain the growing maw of consumption? The Food and Agriculture Organization of the UN estimates that, in order to feed a growing and increasingly urban world population, annual grain production will need to increase to 3 billion metric tons; meat production will need to increase to 470 million metric tons. That’s a lot of food.

It would be naïve to conclude that agricultural automation alone will be enough to save us.

Complicating matters is the fact that technology alone does not decide how much food the world eats. In the wake of globalization, trade carries food across borders and oceans. Wheat from Oregon ends up at mills in Asia, beef from Japan may end up on a plate in New York. Even if individual countries can increase food production through the use of automation, the state of international trade deals will determine where it goes. In a world in which nationalist competition seems ascendant, and trade deals stand on shakier ground, the global food trade network may change dramatically.

For these reasons, it would be naïve to conclude that agricultural automation alone will be enough to save us. The problem we face is complex and multifaceted, and no individual technological innovation is likely to change our fate by itself. But despite the fact that there will be no silver bullets in this struggle, we should take solace in the fact that automation is just one arrow in our technological quiver. One breakthrough idea or creative solution won’t be enough, but a hundred of them might do the trick — and if there’s one thing we have in abundance as a species, it’s creative ideas.

Ultimately, the past may be the greatest source of relief. History shows that technology can help feed the world, so we have every reason to believe it can continue to do so in the future.