Planting the future: Mizzou researchers put AI to work on the farm

July 8, 2026
Contact: Eric Stann, [email protected]

Farmers are getting more tools in their toolbox, thanks to new research from the University of Missouri that shows how they can tweak planting practices to make the most of every acre.

Traditionally, farmers have applied a one-size-fits-all approach to planting. But that strategy may be leaving profits behind, as not every part of a field performs the same.

With the help of digital technology and artificial intelligence, Mizzou researchers are exploring strategies that allow farmers to manage different parts of a field based on their unique conditions.

"Fields might look the same from the road, but they're not," Jasmine Neupane, assistant professor of agricultural systems technology at Mizzou's College of Agriculture, Food and Natural Resources and co-corresponding author, said. "Some areas have better soil and moisture, while others are more prone to erosion or nutrient loss."

Smarter planting with AI

In many cases, the strategy of planting more seeds to boost production can increase costs without improving returns. To address this, Neupane and her colleagues used an AI model to analyze data from two Ohio farms and show how variable-rate seeding (VRS) can help farmers strike a better balance.

Instead of planting the same number of seeds across an entire field, VRS allows equipment to adjust seeding rates in real time based on each area's yield potential. The researchers trained their AI model using common field data - including soil samples, elevation and years of yield records - to generate more location-specific recommendations.

"AI helps farmers choose the right planting rate for different parts of the field," Neupane said. "It also helps them adjust how much fertilizer and crop protection they use, leading to lower costs and better overall results."

This targeted approach can also help farmers use resources more efficiently.

"It keeps farmers from applying nutrients or chemicals unnecessarily," Neupane said. "That helps prevent runoff and other environmental impacts, protecting nearby soil and water."

Mixed results

Researchers studied corn and soybeans - two of the most common crops in the United States - and found key differences in how they respond to the new approach.

Corn showed consistent and predictable results. The model identified where higher planting rates paid off and where they didn't, making it a strong candidate for immediate use in precision agriculture.

Soybeans, however, proved to be more complicated because the plants can adapt to the season based on weather conditions. This resilience makes it harder to predict how changes in seeding rates will affect final yields. In many cases, environmental factors such as rainfall and temperature had a greater impact than planting decisions alone, suggesting more research is needed before the model can deliver consistent recommendations for soybean farmers.

This summer, Neupane plans to expand the research to fields at Mizzou's Digital Agriculture Research and Extension Center. For her, the motivation behind doing this work is deeply personal.

Growing up in Nepal, Neupane witnessed firsthand the challenges farmers face with small land holdings and limited access to technology. Her goal is to use digital tools and data-driven insights to make farming more accessible and efficient for those growing food around the world.

As AI tools continue to advance, she hopes these technologies will help farmers better understand their fields and make more informed decisions.

"When you really understand what your field is telling you, you can manage it much more strategically," Neupane said.

The study, "Leveraging machine learning and geospatial analysis to determine agronomic and economic optima for variable-rate seeding in corn and soybean," was published in the Agronomy Journal.

Co-authors are Bishwoyog Bhattarai at Mizzou; Asbin B K at Texas A&M University; and John Fulton, Neha Joshi and Sami Khanal at The Ohio State University. Khanal and Neupane are co-corresponding authors. Asbin B K was a graduate student at Mizzou during the study.

MU College of Agriculture, Food and Natural Resources

Build a Resilient World