U of I leads $6M AI research to reduce global crop losses

MOSCOW, Idaho — University of Idaho has been awarded a $6 million grant from the National Science Foundation (NSF) to develop new data collection technologies and AI models to breed stronger, higher-yielding cereal crops. The research aims to reduce harvest losses that disrupt the global food supply and cost farmers billions of dollars each year.

The four-year NSF Established Program to Stimulate Competitive Research (EPSCoR) award will bring together artificial intelligence experts, engineers, plant biologists and geneticists from U of I, Clemson University and University of Nebraska Medical Center to improve the predictive science behind breeding crops such as wheat, corn and sorghum.  

The research addresses one of agriculture’s persistent challenges: plant stems bending or breaking before harvest — a problem called stalk lodging. According to Bayer Crop Science, lodging causes an estimated 5% to 25% of cereal crops to be lost worldwide each year. When plants fall over because of high winds or other environmental stressors, they become more vulnerable to pests and disease and often cannot be harvested, impacting farming operations and reducing food production and crop quality.

“Two-thirds of the world’s calories come from wheat, corn and rice, all crops that suffer from stalk lodging,” said Daniel Robertson, associate professor of mechanical engineering and the project’s lead investigator. “Our goal is to make plant breeding predictive. Instead of crossing plants and hoping for the best, we want breeders to know which genetic combinations will produce crops that are both high yielding and resilient to high winds and other events.”

The project tackles one of the biggest obstacles in crop breeding, known as the genome-to-phenome bottleneck. Plant breeders can identify millions of genetic markers in a crop’s DNA through laboratory testing. Collecting field data is more difficult because each plant has to be grown and physically measured in the field. The data imbalance makes it difficult to determine which genes influence valuable traits such as stalk strength.

Researchers at U of I are leading the development of harvest-integrated sensing technologies to measure the physical characteristics of plants. Data captured will be used to train AI models on the DNA sequences of plants to reduce variables and identify the genetic regions that influence lodging resistance.

“This is fundamentally an engineering problem as much as a genetics problem,” Robertson said. “Plant scientists have made tremendous advances in understanding crop genetics, but mechanics can tell us why plants fail.”

Robertson and a team of graduate students are modeling several designs for new sensing technologies that could eventually be integrated into commercial combines. Collecting data during harvest could greatly increase the amount of information available to breeders while reducing the time and cost required to collect it.

Each partner institution will contribute specialized expertise. AI researchers at University of Nebraska Medical Center will apply their knowledge of the human genome to build new AI models that analyze plant DNA and identify genetic signatures associated with stronger plants.

Clemson University scientists will investigate how genetics influence stalk strength, connecting plant performance to cellular and molecular processes.  

U of I is partnering with North Idaho College to strengthen pathways for students transferring into engineering programs and collaborating with Brigham Young University-Idaho to introduce more engineering students to graduate research opportunities in the state.

“This research will prepare the next generation of workforce-ready professionals at the intersection of agriculture, engineering and artificial intelligence,” Robertson said.

The NSF award begins in August and will support research through 2030.

Media contact

Alexiss Turner 
University of Idaho College of Engineering 
208-885-7511 
alexisst@uidaho.edu