Winning contest entry shows AI's agricultural potential
Lincoln, Nebraska, April 8, 2026 — Nipuna Chamara, research assistant professor in biological systems engineering at the University of Nebraska–Lincoln, won a category of the 2025 Testing Ag Performance Solutions competition using artificial intelligence to make decisions for him.
The takeaway? AI can be an invaluable agricultural tool when paired with farmers’ experience and know-how.
“If a person like me, who’s not a farmer, can use AI to win a competition like this, imagine what a seasoned farmer, with decades of experience and knowledge, could do with this tool,” Chamara said.
The TAPS program is a university-led, real-world competition where participants manage actual plots of corn and soybeans for an entire growing season. The three judging categories are highest yield, highest profitability and highest input use efficiency. Participants make many of the same decisions for their TAPS plot that they make at home — such as seed selection, irrigation, pest control and grain marketing — but in a low-risk environment where they can experiment without harming or diminishing crops on their working farms.
Each team is given a plot or plots of land on which to grow their crops. Extension educators provide the teams with data, such as moisture levels and soil health, after which teams decide how much and when to water, fertilize or make other such choices. The results are tallied in September, and the winners are announced in January the following year.
In 2025, the TAPS fields were located at the Research, Extension and Education Centers in North Platte and Mead, Nebraska.
“The average grower has about 40 growing seasons to improve their operation,” said Chris Proctor, a Nebraska Extension educator who helps manage the contest. “As soon as the seed goes in the ground, they’re kind of locked in for that year. Within TAPS, last year we got 116 teams competing, so that’s 116 growing seasons’ worth of decisions all in one. So, in some ways it accelerates the learning iterations that are possible within the season.”
After a conversation with his doctoral supervisor, Yufeng Ge, and teaming up with faculty with expertise in nitrogen management, irrigation management, agronomy and agricultural economics, Chamara first entered the competition in 2024, using OpenAI’s ChatGPT to help him make decisions. He found decision-making in a real-world environment difficult and not practical with the AI models available.
At the time, ChatGPT had no real-time access to data, so Chamara and his team entered all the data manually, uploading information on crop type, soil health, moisture, weather and more. They then asked the AI simple close-ended questions about whether it was a good time to plant, fertilize, etc., and followed the recommendations given.
Proctor said he and the other competition managers didn’t see the use of AI as cheating and were more curious than anything.
“At that time, the excitement was building around AI, and I still don’t know that all of us really had our heads wrapped around what it was,” Proctor said. “At that point, it was much more conceptual than practical. AI claimed to do a lot of things, so let’s just see what happens. I didn’t have an expectation that it was just going to run away with the competition.”
That year, Chamara and his team placed seventh in the yield category.
For the 2025 competition, Chamara grew three corn plots and one soybean plot and immediately noticed two key differences with AI:
> It had become more advanced, pulling in real-time data from the internet on which to base its recommendations, even considering recent news about commodity price fluctuations.
> It already had data from the previous year’s competition to serve as a foundation for adding the current year’s information.
To capitalize on these advances, Chamara and his team would upload new data, such as the weekly grain report, and ask close-ended questions with a specific objective in mind.
The AI would then pull information from the internet, combine it with Chamara’s new and old data and recommend certain actions followed by an explanation of its reasoning. For example, as Chamara was competing in the profitability side of the contest, the AI suggested he lock in prices for his corn early because, at the time, market prices for crops were fluctuating drastically with the introduction of new tariffs.
Chamara said that relying solely on AI does run risks, as it can sometimes base recommendations on faulty or incorrect information found online, and he stressed the importance of turning to reliable sources such as grain reports and extension publications. However, he said, countries like the United States and Canada have a long history of collecting agricultural data and sharing it openly with the public, which benefits AI decision-making.
At the conclusion of the 2025 competition, Chamara and his team earned first place for the highest corn yield in the Mead sprinkler corn competition. They recently published the research outcomes in the journal Artificial Intelligence in Agriculture.
For the 2026 competition, Chamara said he would like to focus on profitability and sustainability and see how AI fares. He also would like to see an app that lets farmers connect AI to the sensors around their farm, automatically uploading real-time data daily to provide them with the most up-to-date information on how to accomplish agricultural goals.
“I think the growers that have more robust digital records and data sets can train AI to be more useful because now, all of a sudden, it gives a context to AI,” Proctor said. “If it has that backlog of reference, that’s where I think the power is.”
Chamara said if farmers are willing to try using AI, they already have the domain knowledge to couple with it.“Like a person who has Google, or a person who uses the library, we can use AI as a tool to make us more powerful in processing data,” he said.