← All research storiesFarming & food

Peer-reviewedFarming & food

Machine learning maps where smart farming already pays off: about 42% of the world’s cropland

Using data from more than 4,000 farm sites and machine learning, researchers estimated where tools like weeding robots, drones and GPS-steered tractors make economic sense today.

How far alongPeer-reviewedReviewed by independent experts for a journal before publication. Peer review is a quality check, not a grade of evidence strength: confidence still depends on study design, limitations, replication and the wider evidence.
How soon could it reach you?A few years outPromising, but needs more testing or engineering first. Our rough editorial guess is several years, with a lot of uncertainty.
Where it comes fromPNASUniversity of Bonn; ETH Zurich; University of California, Berkeley · Oct 19, 2026Read the original →

What the scientists did

“Smart farming” means using technology to give each part of a field just what it needs. The team looked at five kinds: self-running irrigation, weeding robots, drones and other field-mapping tools, machines that decide how much fertilizer or pesticide to apply, and GPS steering for tractors.

No global database of where these tools are used existed, so the researchers contacted more than 8,000 smart-farming experts and searched scientific papers, finding more than 4,000 locations where the tools are in use. They then used machine learning to find which local conditions go along with adoption, and combined that with regional economic forecasts to build a world map of where smart farming pays off.

Words to know

Smart (precision) farming

Using sensors, data, robots and automation so each part of a field gets the right amount of water, fertilizer or care.

  • Self-running irrigation
  • Weeding robots
  • Drones that map fields
  • GPS-steered tractors

Machine learning

A type of AI that finds patterns in data instead of following rules written by people.

  • Finding which conditions, like farm size or income, go with farms adopting new tools
  • Recommending videos based on what you watched

Environmental governance

How strong and well enforced a region’s environmental rules are. The release doesn’t list specific rules; common examples are below.

  • Limits on fertilizer or pesticide use
  • Rules protecting rivers and groundwater

What they found

The study estimates smart farming already pays off on 42% of the world’s cropland, a share that could rise to 50% within 20 years. With stricter environmental rules worldwide it could reach 75%; with looser rules it would fall to 31%.

The strength of environmental rules was “by far the most important predictor” in the models. Payoff is highest where incomes and education are high, fields are large and rules are strict, such as Western Europe, Canada, the U.S. and Australia. Potential is limited in sub-Saharan Africa, partly because today’s machines are built for large farms. Smart irrigation is often profitable where water is scarce.

42%of global cropland where smart farming already pays off
75%possible share with stricter environmental rules
4,000+smart-farming locations in the dataset

Why it matters to you

More of the food you eat comes from farms that use drones, robots and GPS-steered tractors to save water, fertilizer and pesticide. This study suggests government rules play a big part in where those tools spread, which affects water quality, food production and jobs in farming communities, including careers for people who like both technology and the outdoors.

Where this could lead

Smarter farm policy

Governments could use maps like this to see where rules or support would make smart farming worth it.

1–3 years

Tools for small farms

Because today’s machines suit large farms best, the findings point to a need for cheaper, smaller technology.

3–10 years

These are possibilities the research points toward, not promises. Most early findings take years of testing before they reach everyday life, and some never do.

Why scientists care

Farmers need to grow more food while polluting less. Knowing where precision tools already make economic sense, and why, helps scientists and policymakers focus on places where the environment and farmers’ profits can both win.

Keep in mind

The map is a model built from expert surveys and published examples, not a complete count of farms, so it may miss places. The news release doesn’t describe which machine learning methods were used or how they were checked, and “pays off” is an estimate that can change with local costs, wages and training.

How far along is the science?

  • Peer-reviewedReviewed by independent experts for a journal before publication. Peer review is a quality check, not a grade of evidence strength: confidence still depends on study design, limitations, replication and the wider evidence.
  • PreprintA manuscript shared publicly, often before or alongside journal review. Check its separate journal and review status (each story says what we know). Results may change.
  • Conference talkPresented to other scientists at a meeting. Usually short and early; a full paper may come later.
  • Agency reportPublished by a science agency such as NASA, NOAA or NIH. Reviewed internally, but not by an outside journal.
  • Press releaseAn announcement from a university or company. Useful, but always check the research it describes.

How soon could it reach you?

  • Could reach you soonCould show up in products, services or advice soon. Our rough editorial guess is about 1 to 3 years, not a forecast from the research.
  • A few years outPromising, but needs more testing or engineering first. Our rough editorial guess is several years, with a lot of uncertainty.
  • Long-term scienceBasic research that builds knowledge. Real-world uses, if any, are far off.
  • Small stepA small update to earlier work. Useful to researchers, but not a change you would notice yet.

Every story links to its original source: the paper itself, or the conference program when no paper is out yet. Explanations are written with AI, usually from the paper’s abstract (each story says what it was written from), then checked against the source. We only use journals, recognized preprint servers, science agencies, scientific conferences and established science news outlets.