LINCOLN, NEBRASKA, US — A University of Nebraska–Lincoln plant scientist will help develop a digital twin of a cornfield, allowing researchers to quickly test what-if scenarios related to corn performance.

UNL’s James Schnable, who received nearly $1 million from the National Science Foundation (NSF), is collaborating with researchers at Iowa State University and Purdue University, who also received NSF funding for the project for a total of $2 million.

“Once we build these digital twins, we can use high-performance computing to simulate and explore how whole fields of corn with different properties would behave — how resilient they are to wind and how efficiently they use water and capture light, for example — without having to actually grow the corn field,” said Schnable, Nebraska Corn Checkoff presidential chair and professor of agronomy and horticulture.

The twins would help researchers overcome real-world constraints on the number of field tests researchers can run and more quickly identify the planting arrangements and hybrids with the most potential.

“If we can weed through a million different combinations and find the most promising ones, we’re going to make a lot faster progress than if we’re just picking 100 each year to test,” he said.

In 2021, eight institutions including Nebraska united to establish the AI Institute for Resilient Agriculture, based at Iowa State, with the goal of building virtual replicas of crops and fields. Through that initiative, Schnable has been collaborating with Iowa State’s Baskar Ganapathysubramanian, an expert in artificial intelligence. For this project, the pair teamed up with Bedrich Benes of Purdue, who focuses on 3D simulation and digital reconstruction of plants.

The trio’s combined expertise and Nebraska’s facilities are enabling the team to overcome longstanding barriers to twinning crop fields. To this point, researchers have relied on crop growth models — complex series of equations — to predict crop performance. But collecting enough data to adapt these models to new hybrids is labor and time intensive.

Nebraska’s LemnaTec High-Throughput Plant Phenotyping System, housed at Nebraska Innovation Campus’ Greenhouse Innovation Center, automates much of the data collection process and produces high-resolution data across a plant’s life cycle, which Benes can use to twin the plants.

The team also will use Ganapathysubramanian’s approaches to model ray tracing — a sophisticated technique for simulating light’s behavior in digital images — to track light distribution throughout the canopy.

“Through these tricks that Baskar has come up with in terms of actually getting ray tracing level data, and having access to really high-performance computing, is why we’re able to do this when it was not previously feasible,” Schnable said.

The digital twin will enable the researchers to pinpoint the planting arrangements that are most ideal for capturing light, minimizing water loss and increasing productivity. They also will enable optimization of the corn plants themselves — one of Schnable’s areas of expertise. The tool will allow users to identify the ideal physical characteristics of corn, such as the number, angle, length and width of leaves, in certain environments.

Then, using inverse procedural modeling and quantitative genetics techniques, Schnable will identify the genes that control these physical characteristics.