AI Powered Soybean Disease Detection Project Gains NVIDIA Support

By Andi Anderson

Soybeans are the most important crop in Ohio, with over 258 million bushels harvested in 2025. Protecting this crop is essential for farmers. When diseases affect soybean plants, quick action is required to reduce damage and avoid unnecessary use of chemicals.

A team of engineers from Ohio State University has received an NVIDIA Academic Grant to develop an advanced system for detecting soybean leaf diseases. The team includes Assistant Professor Hari Subramoni and Professors Scott Shearer and John Fulton. Their project focuses on using artificial intelligence (AI) and edge computing to help farmers make faster and better decisions.

The grant provides powerful computing support, including 32,000 GPU hours on NVIDIA’s cloud platform and two compact supercomputers. These tools will help researchers train and improve AI models and allow data to be analyzed directly in the field. This approach reduces delays and improves accuracy.

“This support from NVIDIA helps us move from a promising proof of concept to a deployable system that can assist Extension personnel and growers with faster, more informed decisions in the field,” said Subramoni. “By combining advanced AI training resources with edge hardware, we can bring analysis closer to where the data is collected and where timely action matters most.”

Currently, detecting soybean leaf disease depends on manual assessments by Extension agents, which can take 24 to 72 hours. This delay often allows diseases to spread further. The new system aims to provide results in real time, helping farmers take quick action.

The project will be tested in four Ohio counties: Trumbull, Champaign, Delaware, and Pickaway. It combines data collected through drones and ground-based tools. The system also uses advanced cyberinfrastructure developed through a national research initiative.

“Our objective is to deliver near real-time, actionable information to Extension educators and agricultural producers,” Shearer said. “Rapid detection and assessment of crop stress enhance the effectiveness and timing of pesticide and nutrient applications, minimizes unnecessary inputs, and enables more precise crop management across Ohio.”

This research shows how modern technology can improve farming practices. The system is expected to reduce unnecessary fungicide use by up to 25 percent. It will also support future developments in precision agriculture, including pest management and irrigation planning.

Photo Credit: gettyimages-zoran-zeremski