LOS ANGELES, California, September 2, 2026 — Agriculture Secretary Brooke Rollins on Tuesday announced a pilot project aimed at modernizing the way the government estimates U.S. crop acreage and production. The program relies on satellite imagery, geospatial tools, and artificial intelligence. It comes after months of criticism from farmers and dealers over uncertain federal information.
A Response to Growing Frustration
Following a month of listening sessions with growers, Rollins announced the plan at the Farm Progress Show in Boone, Iowa. Staff cuts at the agency were criticized by farmers and grain traders alike. At the same time, they cited unusually large revisions between final corn acreage estimates and previous projections.
Rollins told the crowd, “We will do a test and see how better satellite imagery works.” According to her, the U.S. Department of Agriculture will coordinate the effort with NASA and other federal agencies to expand reporting without putting additional burdens on farmers.
The listening sessions highlighted widespread concern over how estimates and revisions are built. Farmers reported being surveyed repeatedly without a noticeable improvement in accuracy. According to Rollins, that feedback was built right into the pilot.
The Importance of the Numbers
Information from the National Agricultural Statistics Service can move commodity markets within minutes after release. Reports also impact the sale price farmers receive for their harvests, such as when grain prices plunged following final USDA numbers.
The drop arrives against the backdrop of a farm economy already experiencing instability, as monitored by educational outreach from Iowa State University Extension and Outreach. Even survey data may be less reliable, economists say, because of staff cuts at the agency. The new pilot is designed to reduce reliance on standalone reporting.
Reports filter through to futures markets nearly instantaneously after publication. Such data is used by traders and lenders to price contracts or extend credit. Even small changes can turn the price of an entire crop around in a day.
Less Paperwork, More Machine Learning
The plan consists of traditional farmer polls complemented by geospatial tools and crop models. The USDA said the aim is to reduce the number of times it asks farmers for identical information. It will also study machine learning to fine-tune yield models.
Officials pledged to be more forthcoming about survey response rates and data limitations. That marks a change, as farmers noted having little visibility into how estimates were constructed. “We’re making these changes in order to continue re-establishing confidence in the numbers,” Rollins said.
The USDA characterized the technology push as additive rather than a major overhaul of practices. For now, this pilot will run in parallel with the existing survey architecture. Farm groups had lobbied for just this kind of gradual, tested deployment.
What Comes Next for Farmers
USDA did not provide a timeline for when satellite data will begin influencing monthly official reports. The agency said it will continue collecting farmer input as the project moves forward. Changes in methodology require backtesting and a large set of representative data before full adoption.
Grain traders will keep a close eye on developments since market movements hang on USDA numbers. While farm groups hailed the announcement, they added that results will count more than promises. Whether confidence in federal crop data recovers before next year’s harvest could depend on the pilot’s success.
Commodity markets have been subject to extra volatility, as farm income broadly faces a slump this year. Better government data might help dissipate some of that volatility over time. For the moment, farmers say they will evaluate the plan by action rather than words.
