A robot army is beginning its march across rural US, promising to transform the future of food. Twenty-five intelligent machines were last month dispatched to the Midwest and the Mississippi Delta, where they are to travel over newly planted fields at 19kph, annihilating baby weeds.
Produced by John Deere and created by the start-up Blue River Technology Inc, these robotic weeders look much like standard industrial sprayers at first glance, but each is rigged with an intricate system of 36 cameras and a mass of tiny hoses. They use computer vision to distinguish between crops and weeds and then deploy with sniper-like precision tiny jets of herbicide onto the weeds — sparing the crop and ending the common practice of broadcast-spraying chemicals across billions of acres.
The “See and Spray Ultimate” robots are expensive, enormous, wildly complex machines accessible only to industrial-scale farmers, but within a few years their effect on the environment and human health could be nothing short of spectacular. They are in the vanguard of a wave of reimagined agricultural equipment that is expected to help farmers produce more food on less land with radically reduced chemical applications.
Photo: AP
Intelligent machines can treat plants individually, eventually applying not just herbicides, but pesticides, fungicides and fertilizers on a plant-by-plant — rather than field-by-field — basis. This kind of hyper-precision might do more than ratchet down agrochemical usage, also allowing for more diversity and crop-mixing on fields, so that larger farms can begin to mimic natural systems.
Meanwhile, robotic planters and combines are already showing yield improvements of up to 2 percent, and robotic harvesters could eventually alleviate increasingly grueling farm work and labor shortages.
Robots on farms, for all their environmental and ethical promise, raise plenty of concerns — some valid, others spurious. They would add cost and complexity to farming equipment, making farmers increasingly reliant on “big ag” companies such as John Deere.
In the beginning, they would reinforce the dominance of large industrial operators while eluding the local small and midsize farmers who are essential to sustainable and resilient food systems.
So as the era of artificial intelligence (AI) in farming dawns, manufacturers, US President Joe Biden’s administration and investors should be thinking about how to develop this market responsibly.
Funding should be steered to the development of smaller, more affordable machines, while also supporting a rental economy that enables local and midsize farmers to lease, if not own, this next-generation equipment.
The US Department of Agriculture (USDA) should also create rebate and tax credit programs to help farmers affordably trade out old machinery for new.
See and Spray is one of seven AI products that John Deere has in development, including robotic planters, self-driving tractors and combines that meticulously separate wheat from chaff. All are equipped with dozens of cameras and algorithm-crunching data processors that examine, analyze and measure every plant and seed on a field.
“We’re doubling down, tripling down on investment in robotics and machine learning,” said Jorge Heraud, John Deere’s vice president of automation and machine autonomy.
Having grown up working on and weeding his grandparents’ tomato farm in Peru, Heraud founded Blue River Technology, which John Deere acquired in 2017 along with its See and Spray prototype for US$305 million. In five years, Heraud has helped grow John Deere’s AI team to 400 people from 50.
Many skeptics question whether this equipment would ever be widely adopted.
John Deere says it already has more demand than it is ready to meet: Heraud decided to release only 25 in its first fleet, because the company is still honing the financing and servicing model.
John Deere is charging an upfront price that it would not disclose — it is at least the cost of a standard sprayer of the same size, about US$500,000 — plus an ongoing per-acre fee that might be charged monthly or annually and includes software upgrades and maintenance.
Heraud plans to increase the fleet by a factor of 10 annually, so that by 2025 the company would have thousands of robotic weeders on the market.
The worry that intelligent machines will simply make industrial farms bigger and farmers lazier, less responsible stewards of the land is unfounded.
These kinds of advanced technologies have extraordinary potential to help farmers improve the health of their soil and the quality of the food they produce by drastically reducing the use of harmful chemical herbicides such as glyphosate, Dicamba and 2,4-D.
Twenty billion gallons (75.7 billion liters) of herbicide are applied annually by sprayers worldwide across 404.69 million hectares of farmland.
When See and Spray technology is integrated into all sprayers sold by the company, which Heraud said could happen within a decade, the volume of herbicide deployed on these farms could plummet to 4 billion gallons.
Future generations of the equipment might also be able to significantly curtail the use of fossil fuel-derived fertilizers, which, when over-applied, fuel climate change.
Governments can help allay concerns with incentives: The California Air Resources Board provides a helpful model, offering farmers rebates for upgrading their existing machinery to models with cleaner engines.
The USDA and investors can also encourage the development of a rental economy by incentivizing and funding young companies such as Nutrien Ag Solutions in the US and Hello Tractor in Africa that function like the Ubers of agriculture, enabling small and midsize farmers to lease or acquire fractional ownership of next-generation farm equipment without having to maintain it or learn the technology.
Venture capital can be directed to support new players in the AI market. One young start-up, Earthsense, is developing robots the size of microwaves that rove around farms removing weeds. John Deere is also working on smaller, more affordable machines.
For better or worse, the era of AI agriculture has arrived, and if investors and government officials do their part to support the responsible development and adoption of this technology, the result would be nothing short of a paradigm shift toward sustainable farming.
Virtually every aspect of food production, from planting to processing, could be revolutionized, making it feasible to feed a hotter, more populous world.
Amanda Little is a Bloomberg Opinion columnist covering agriculture and climate. She is a professor of journalism and science writing at Vanderbilt University.
This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.
SEEKING CLARITY: Washington should not adopt measures that create uncertainties for ‘existing semiconductor investments,’ TSMC said referring to its US$165 billion in the US Taiwan Semiconductor Manufacturing Co (TSMC, 台積電) told the US that any future tariffs on Taiwanese semiconductors could reduce demand for chips and derail its pledge to increase its investment in Arizona. “New import restrictions could jeopardize current US leadership in the competitive technology industry and create uncertainties for many committed semiconductor capital projects in the US, including TSMC Arizona’s significant investment plan in Phoenix,” the chipmaker wrote in a letter to the US Department of Commerce. TSMC issued the warning in response to a solicitation for comments by the department on a possible tariff on semiconductor imports by US President Donald Trump’s
The government has launched a three-pronged strategy to attract local and international talent, aiming to position Taiwan as a new global hub following Nvidia Corp’s announcement that it has chosen Taipei as the site of its Taiwan headquarters. Nvidia cofounder and CEO Jensen Huang (黃仁勳) on Monday last week announced during his keynote speech at the Computex trade show in Taipei that the Nvidia Constellation, the company’s planned Taiwan headquarters, would be located in the Beitou-Shilin Technology Park (北投士林科技園區) in Taipei. Huang’s decision to establish a base in Taiwan is “primarily due to Taiwan’s talent pool and its strength in the semiconductor
Industrial production expanded 22.31 percent annually last month to 107.51, as increases in demand for high-performance computing (HPC) and artificial intelligence (AI) applications drove demand for locally-made chips and components. The manufacturing production index climbed 23.68 percent year-on-year to 108.37, marking the 14th consecutive month of increase, the Ministry of Economic Affairs said. In the first four months of this year, industrial and manufacturing production indices expanded 14.31 percent and 15.22 percent year-on-year, ministry data showed. The growth momentum is to extend into this month, with the manufacturing production index expected to rise between 11 percent and 15.1 percent annually, Department of Statistics
An earnings report from semiconductor giant and artificial intelligence (AI) bellwether Nvidia Corp takes center stage for Wall Street this week, as stocks hit a speed bump of worries over US federal deficits driving up Treasury yields. US equities pulled back last week after a torrid rally, as investors turned their attention to tax and spending legislation poised to swell the US government’s US$36 trillion in debt. Long-dated US Treasury yields rose amid the fiscal worries, with the 30-year yield topping 5 percent and hitting its highest level since late 2023. Stocks were dealt another blow on Friday when US President Donald
RSS