Conventional laboratory tests in agriculture are slow, expensive and complex.

But new research from the University of Illinois Urbana-Champaign (UIUC) shows that near-infrared (NIR) spectroscopy combined with machine learning is changing the way we analyze commodities like corn kernels and sorghum biomass – with faster, more cost-effective results without sacrificing accuracy.

In two groundbreaking studies published in a UIUC press release, researchers show how NIR technology not only reduces costs but also helps industry analyze key components more quickly. This innovation could have far-reaching implications for everything from food production to biofuels.

Accelerated corn kernel analysis using NIR spectroscopy

“NIR spectroscopy has many advantages over traditional methods. It is fast, accurate and inexpensive. Unlike laboratory analysis, it does not require chemicals and is therefore more environmentally friendly,” explains Mohammed Kamruzzaman, assistant professor in the Department of Agricultural and Biological Engineering (ABE) at the University of Illinois.

In one study, researchers developed a global model for analyzing corn kernels – a critical advancement for the grain processing industry, as moisture and protein content directly affect nutritional value, processing efficiency and price.

Because corn varies greatly depending on where it is grown, the research team collected samples from seven countries – Argentina, Brazil, India, Indonesia, Serbia, Tunisia and the United States – to ensure their model works in different environments. Lead author Runyu Zheng explained: “We combined gradient boosting machines with partial least squares regression. This is a novel approach that produces accurate and reliable results.”

Increasing biofuel production through sorghum biomass analysis

The second study looks at the use of sorghum biomass for biofuel – a renewable and low-cost energy source. Rapid analysis of the chemical composition could revolutionize the biofuel and plant breeding industries.

Md Wadud Ahmed, the lead author, explained: “We first scanned the samples and obtained NIR spectra as a result. This is like a fingerprint that is unique to different chemical compositions and structural properties.”

NIR’s non-destructive, fast and flexible screening capability allows industrial users to analyze samples without interrupting production. Kamruzzaman summed it up: “You can simply take samples for measurement, scan them and then feed them back into the production flow.”

For more information, see the original studies in Food Chemistry And Biomass and bioenergyfunded by the DOE Center for Advanced Bioenergy and Bioproducts Innovation.

• First article: “Optimizing feature selection using gradient boosting machines in PLS regression for predicting moisture and protein in multi-country maize kernels using NIR spectroscopy” (DOI: 10.1016/j.foodchem.2024.140062).
• Second article: “Rapid and high-throughput determination of sorghum (Sorghum bicolor) biomass composition using near-infrared spectroscopy and chemometrics” (DOI: 10.1016/j.biombioe.2024.107276).

This research, supported by the U.S. Department of Energy, represents a critical step toward more efficient and sustainable agricultural and bioenergy practices.