Hit enter to search or ESC to close

Innovations in Agriculture: Harvesting the Benefits of Autonomous Technology

Automation innovation in agriculture

It’s 4:30 a.m., and on a farm somewhere in Texas, a farmer wants to ignore the alarm clock, but can’t. There’s so much work to be done out in the fields, and it needs to be finished before the incoming storm. Imagine, instead, that with the swipe of an app, that farmer could instruct an autonomous tractor to leave the barn, drive itself to a designated field, carry out the tillage, crop maintenance, or harvesting work that needs doing, and return to the barn when it’s done, regardless of the time of day.

If you’ve been keeping up with coverage of CES 2022, you’ll know this is no longer a vision, but very much a reality, after John Deere announced at the Las Vegas tech show that its Autonomous 8R Tractor will be available for farmers by the end of the year.

Agriculture is no stranger to automation. Since the mid-1990s, GPS-guided autosteer technology has enabled farmers to drive their combine harvesters in straight, repeatable rows. Automated feeding machines have ensured time-controlled, precise delivery of food per animal on livestock farms for at least 15 years. It was only a matter of time before farmers would be offered an autonomous tractor—but as we will see, autonomous tractors are just the start of a new era of human-machine collaboration in a sector exploring the potential of new technologies that could facilitate precision farming, improve productivity, address the environmental impact of existing farming techniques, and improve quality of life for farmers.

Feed the world

Agriculture is a sector under immense pressure to increase output in the face of growing economic and environmental challenges; it’s an area where autonomous technology could make a difference to everyone, now; and the business case adds up.

According to the Washington, D.C.-headquartered Population Reference Bureau (PRB), the global population will reach 9.7 billion by 2050. That rapidly rising population needs to be fed, but the challenges of supporting such growth in food production are matched by the challenges of a declining and aging workforce in agriculture globally, further hindered by political and pandemic-related restrictions on transient workforce availability. “When it comes to harvesting, humans are extremely efficient, but one of the reasons for automation in agriculture is the lack of labor,” explains Wilfried Aulbur, a Senior Partner at Roland Berger, a strategy consultancy. Aulbur notes that while demand for agricultural products is rising, fewer people are working in farming than they used to, leaving farmers with little choice but to consider turning to automation. “Nobody wants to invest in an autonomous harvester, or a complex fruit picking machine, but without farmworkers, you end up with unharvested crops. And that’s when the use case for robots makes sense.”

Indeed, in a recent article, Jahmy Hindman, Chief Technology Officer at Deere & Co., wrote, “The farm is poised to be the place where fully autonomous vehicles first breakthrough to profoundly impact every one of the 7.7 billion people on the planet.”

Just as it does for moving people and goods, autonomous technology offers numerous advantages for farming: it can work through the night; it doesn’t get tired, or sick; and it is highly efficient, especially when multiple units are deployed in what’s known as a “swarm.” All of this can take considerable administrative, operational, and physical pressure off farmers, enabling them to focus on value-add work, and ultimately, improving their quality of life.

Comparing cars and tractors – like apples and pears?

Self-driving vehicles operate within an Operational Design Domain (ODD)—that is, the operating parameters of a self-driving system that define the speed limits, road types, weather conditions, and geographical boundaries within which the technology is designed to operate based on its sensor capabilities.

Clearly, cars and autonomous tractors have very different ODDs. Autonomous road vehicles need to operate within the rules of the road, alongside other vehicles, and alongside vulnerable road users such as pedestrians, cyclists, and scooter riders; by contrast, farm equipment needs to operate at much lower speeds, in open fields where most of the rules of the road don’t apply, and there’s no oncoming traffic. It’s easy to see why the use of autonomous vehicles is forecast to scale particularly quickly in agriculture. 

That said, autonomous agricultural machinery needs to work alongside farmworkers, and the public, too. “Just as with autonomous cars, we need to solve for safety, because the autonomous technology will always be used in a cooperative environment between robots and people,” notes Aulbur. There may also be hikers, or children playing near the field; and the farmer will need to jump behind the wheel if there’s a public road on the way to the field.

Applications of autonomous technology in farming

The main applications for autonomous farm technology are the preparation and plowing of the soil, known as tillage, as well as milking, harvesting, weeding, and pest control.

Autonomous plowing

To understand the advantages of autonomous plowing, take the example of the aforementioned Autonomous 8R Tractor, a full-size tractor designed for autonomous tillage. Farmers can pilot the tractor manually or use their mobile devices to remotely assign the tractor to till soil in a field in autonomous mode, monitoring its status via the John Deere Operations Center Mobile app. The vehicle is equipped with a variety of technology systems, including six pairs of stereo cameras, and the tractor can remotely contact the Operations Center for help if it encounters obstacles that it cannot identify.

Autonomous milking

Given the vast quantities of milk produced and consumed around the world, it’s hardly surprising that milking has been automated for a number of years. Although most farms have several thousand cows, herds at some of the world’s largest dairy farms run into the tens and hundreds of thousands.

“One Russian farmer that I know has 150,000 cows, producing 3.9 million liters of milk a day,” says Aulbur. “The cows are tagged, and whenever they feel they need to be taken care of, they walk to one of the milking stations where a robot attaches the milking equipment. The cow gets milked, and then moves out. The process is fully autonomous.”

Autonomous spraying and Weeding

The application of herbicides and pesticides is a hugely inefficient aspect of farming. Not only are the products expensive to buy, but broadcast application, sometimes known as “spray and pray,” covers a large area of crops and requires greater volumes than targeting individual weeds, and there’s no way of avoiding the run-off pollution caused by spray bouncing off leaves. 

Precision is a major factor in deploying autonomous technology, says Aulbur. “The precision piece refers to quantities of nutrition, water, pesticides, and herbicides. Where do I put what, rather than putting everything equally? Precision enables application down to a plant level.” That’s why, in 2017, John Deere acquired Blue River Technology, which had developed targeting technology called See & Spray, using computer vision and machine learning to detect and differentiate plants and weeds, and to target herbicide only at the weeds.

But what if you could do this on a per-plant basis, without chemical or physical intervention? Seattle tech start-up Carbon Robotics has a solution which it claims is better than spraying chemicals, or weeding mechanically or manually. The Autonomous LaserWeeder trundles through fields of crops at 5 mph, and can kill 100,000 weeds per hour, covering up to 20 acres a day. Autonomous navigation is possible day and night, achieved using a combination of computer vision, front and rear cameras, positional GPS accurate to within 6 inches, and lidar sensors for obstacle detection.

An alternate take on chemical-free weed killing has been developed by Small Robot Company. The Tom monitoring robot maps a field of crops and identifies individual weeds, which the company’s robot weeding prototype, Dick, then destroys with electrical charges, akin to small lightning strikes.

Harvest Assist

Perhaps the most critical area of farming is the harvest. Working the land is pointless without the right timing and efficiency of harvesting the grain, crops, fruit, or vegetables. Although increasing levels of automation have been deployed in combine harvesters for a number of years, fruit and vegetable picking is of major interest to autonomous technology developers seeking to solve the complexities of identifying good quality and ripe fruit, rejecting rotten fruit, and carefully handling the picked fruit.

Global production of fresh fruit is rising, with almost 900 million tons of fresh fruit produced annually. However, according to the UN Food and Agriculture Organization, around 30 percent of global food loss occurs at the agricultural production and harvest stage, and among the reasons for this are farmers abandoning crops or failing to complete harvests due to lack of labor availability. With the fresh fruit market alone worth $615 billion in 2021, crop waste creates not only a moral dilemma but also a huge cost for farmers; and that makes it a highly lucrative market for autonomous solutions.

“The key thing is understanding the problem that you’re solving,” says Aulbur. “For the farmer, it’s about quality, speed, and cost per unit—that is, the cost per crate of apples, or grapes, or the cost per head of lettuce. The farmer suffers a massive discount on the price of an apple that is not perfectly ripe, and the same goes for bruises and blemishes. Yield functions for agriculture drop rapidly.”

Using artificial intelligence (AI) can also help farmers guarantee the accuracy and quality of their products, explains Inma Martinez, Co-Chair and Project Lead AI in Agriculture & Farming at Global Partnership on Artificial Intelligence (GPAI). “In the case of specialty and perishable crops such as fruit and vegetables, knowing with precision when to harvest is of utmost importance. Many such crops are picked early to mature during transport. Here is where machines deliver the most accurate prediction, and sensors in each box that can guarantee levels of maturity act as an insurance policy. Supermarkets only purchase produce ready to be sold, and refuse what is not ripe, so farmers get the short end of the stick if their produce is audited by human eyes. But accurate data means buyers cannot renege on contracts, and farmers’ produce has to be bought as ordered.”

For those out in the fields, one harvest assist solution is Burro, which uses follow-me technology to guide an automated wagon into which pickers can load fruit, enabling them to focus on picking without the weight and distraction of shifting a manual wheelbarrow.

Burro is designed to assist human pickers. But as the challenges of hiring sufficient numbers of fruit pickers increase, farmers have begun to look to tech to ensure their crops don’t go to waste on the branch. “Picking an apple is a complex task,” explains Aulbur, “because you need the right green and red ratio to determine the ripeness of the apple.”

That’s why Tevel Technologies has developed AI-powered orchard drones that can identify and pick ripe fruit. Tevel’s “flying autonomous robots” are tethered to a vehicle which trundles slowly along orchard rows, stopping as required to enable the fruit-picking drones to fly up to the branches to identify, pick, and deliver ripe fruit to the wagon below. AI perception algorithms and cameras classify and select the appropriate fruit, which is plucked by a picking arm. The drones can work day and night, calculating the yield as they work, and the company’s website shows a large orchard with a fleet of tethered drones operating in unison. Tevel says its solution is intended “to complement existing fruit pickers and not merely to substitute them.”

Cultivating a new era of farming

From autonomous tractors and automated dairies, to plant-identifying weeding machines, and apple-picking drones, agriculture is at a major turning point in its evolution. Advances in the automation of agricultural technology are significant, but autonomous technology for farming is task-specific and lacks human versatility, notes Aulbur, adding, “What works for strawberries doesn’t work for blueberries. What works for blueberries doesn’t work for grapes. And what works for Romaine doesn’t work for another type of lettuce.” 

Being limited to mission-specific solutions complicates an investment decision for those running mixed farms producing crops and livestock. Here, farmers may be interested in farming as a service, a fast-growing business model which helps farmers deploy autonomous technology on a rental-style basis, without needing to absorb the high upfront costs of acquisition. Farmers struggling to hire sufficient labor get to harvest their crops, and supermarkets—and their customers—get high-quality produce.

Autonomous technology is being introduced into agriculture through the automation of existing equipment, but longer-term, not only farming equipment but the farming techniques themselves are likely to evolve around autonomy. So when you next bite into an apple or sip on a glass of Chablis, you might just be enjoying the work of an autonomous technology coder—and a well-rested farmer.

Choose your lane

How Autonomous Vehicles Distinguish Between Bicycles and People Who Ride Them

How Autonomous Vehicles Distinguish Between Bikes and People

When it comes to how autonomous vehicles see the world, humans come first, literally. Autonomous vehicles (AVs), like the kind operated by Pittsburgh-based Argo AI, use Machine Learning to detect and classify the objects in their surroundings, identifying people...
Why The League of American Bicyclists is optimistic about autonomous vehicles

Why a Leading Cycling Advocacy Group Is Optimistic About Autonomous Vehicles

As autonomous vehicle use grows, AV companies and the League of American Bicyclists are collaborating on how to ensure cyclists and motorists can share the roads safely, even if the “motorist” is artificial intelligence software. As part of the...

Self-Driving Is Arriving Right On Time. Just Like Ice Cream Did

Seven years ago, I was a self-driving skeptic. Not of the technology. Of all the “experts” promising autonomous vehicles would be everywhere by 2020. You didn’t need to be Nostradamus to know that was ridiculous. All you needed was...
Illustration of a futuristic parking deck turned into a mixed-use space, with AVs driving by

How Autonomous Vehicles Could Help Transform Parking Lots

Researchers say it’s likely that autonomous vehicles (AVs) can help reduce the need for parking lots, opening more room for grass and trees and other elements of nature. It may not seem like it when you’re circling the block...
An illustration of an Argo autonomous vehicle in teal against a backdrop of buildings, a bicyclist, and research papers

7 Big Breakthroughs From Argo Research at CVPR 2022

The 2022 Conference on Computer Vision and Pattern Recognition (CVPR 2022) is nearly here. Thousands of computer scientists, software engineers, and researchers from around the globe will gather in New Orleans to review and discuss their latest work in...

Researchers Predict the Future With Lidar Data

Researchers at Carnegie Mellon University’s Argo AI Center for Autonomous Vehicle Research, a private-public partnership funded by Argo for advancing the autonomous-vehicle (AV) field, say they have come up with a way to use lidar data to visualize not...

Must Reads