Precision agriculture is not better for the environment — and it often just leads farmers into greater debt
Will robotics and “smart” technology solve the climate crisis and make agriculture more resilient and sustainable?
That’s what the entrepreneurs and tech companies who are launching the Reservoir, a new “on-farm innovation hub” for robotic farm technology in Salinas, California, would like us to think. In their press release for the test farm for “robotics innovation,” a John Deere spokesman involved with the project says that they are “strengthening the resilience, efficiency, and sustainability of high-value crops — with potential to benefit food systems worldwide.”
The Reservoir is only the latest attempt to introduce new, expensive electronic technologies into agriculture. The technologies they are testing fall under the umbrella term of “precision agriculture,” which, according to a General Accounting Office report released on January 31, 2024, includes technologies like remote sensing platforms, in-ground sensors, targeted spray systems, automated mechanical weeders, auto-steering tractors and activity monitors on dairy cows. Although precision agriculture has been promoted by ag tech companies since the 1990s, less than 30 percent of U.S. farmers currently utilize it.
Precision agricultural technologies are usually promoted as increasing efficiency, applying chemicals more precisely, or simply collecting larger amounts of on-farm data that can be used to make management decisions. They usually involve no fundamental change in a farming system beyond adding more electronic sensors or data-analyzing software. Nevertheless, in rhetoric typical of the electronics industry in general, their promoters like to say that precision agriculture is “climate-smart” or better for the environment.
But is it?
Equating “smart” with “green” is one of the greatest marketing triumphs of the electronics industry — and it’s a lie. Computers are not inherently good for the environment — far from it. Electronics manufacturing is an extremely polluting industry, dependent on all sorts of rare elements that cause environmental destruction when they are mined. Electronic waste disposal is similarly problematic, and recycling systems are still inadequate.
And it’s not just the actual electronics in a self-driving tractor or precision fertilizer applicator that have environmental impacts. For any device connected to the internet (as practically all of them are nowadays), a substantial amount of its environmental impact stems from huge data centers filled with computer servers, which currently account for 4.4 percent of all U.S. electricity usage, according to a recent article in MIT Technology Review.
Increased application of artificial intelligence (AI) technologies will greatly increase both the amount of data centers being built and the amount of electricity they use. And fantasies about all those computers being powered by solar and wind don’t jive with the reality that computers require a steady, reliable source of electricity — something intermittent renewables can never provide. Currently, data centers use electricity with a 48 percent higher carbon intensity than the U.S. average.
According to the U.S. Department of Energy, electricity use by data centers has tripled in the last decade and could double or triple again by 2028 — drawing up to 12 percent of total U.S. electricity usage — if AI is adopted at the rates that tech companies are pushing for. That’s why, despite retirements of coal plants and huge increases in wind and solar generation, total U.S. carbon emissions from energy use remained steady in 2024 and total electricity generation reached an all-time high, according to the U.S. Energy Information Administration.
AI is not “climate-smart,” whatever its proponents may say. It may actually be the dirtiest invention of the twenty-first century — but the environmental impacts are invisible to the end user. Any kind of new robotic technology that utilizes AI — for example, a robotic weeder that uses cameras and image-matching technology to distinguish between weeds and crops — will contribute to the demand for more servers and electricity.
Regardless of what their manufacturers claim, no electronic technology — precision agriculture included — is going to significantly reduce carbon emissions. The two biggest users of fossil fuels in agriculture are farm machinery and agricultural chemicals. Any management system that allows the farmer to minimize the number of passes through the field — reduced tillage, for example — can help reduce fuel use. But using GPS to steer the tractor or making an autonomous tractor that the farmer doesn’t have to sit in will not make any difference.
Similarly, any system that reduces chemical inputs will reduce a farm’s fossil fuel footprint. Synthetic nitrogen fertilizer is very carbon intensive to produce, accounting for 2 percent of global energy use, mostly from natural gas. Precision technology may help farmers apply fertilizer closer to where plants will take it up, but it won’t reduce fertilizer use below the recommended application rate. However, a better system of crop rotation that incorporates legumes, manure and cover crops can eliminate the need for nitrogen fertilizer, reducing fossil fuel use far more dramatically than the most efficient applicator anyone can ever invent.
Pesticide manufacture also uses a huge amount of fossil fuels — it takes ten times more fossil fuel to make one pound of pesticide than it does to make one pound of nitrogen fertilizer. Integrated pest management (IPM) techniques can help reduce pesticide use by scouting fields and only applying pesticides if economic damage to the crop seems imminent, rather than spraying on a set schedule. But there’s no need for electronic sensors — a good farmer who knows their land and crops will be better able to make decisions about whether pesticides are necessary than any computer model.
Finally, there’s a significant reason why 70 percent of U.S. farmers have not yet adopted precision agriculture technologies — and it’s not because they’re Luddites. The reason is cost. Farm machinery is expensive enough as it is, and the more electronic doodads get added on, the bigger the price tag. Most farmers are barely breaking even, if not losing money, and the last thing they need is to go into debt buying fancy new electronic toys with dubious benefits.
Bottom line: If you like playing with new technologies, have the time and expertise to repair them when they inevitably malfunction, and have enough disposable income to buy precision ag technology, by all means do so. But don’t be deceived by the shiny marketing rhetoric into thinking that it’s better for the climate or environment, because it’s not. Good eco-ag practices like reduced tillage, cover crops, crop rotations and rotational grazing confer far greater environmental benefits than robotics ever will.
