Better corn chips with the help of machine-learning

That perfect, crunchy bite of a corn chip starts with knowing exactly how much moisture in the corn. Masa dough, used to make products like corn chips, starts with soaking and cooking the corn grain. This process is called nixtamalization. The moisture content of the dough affects the taste, nutrition, and if it will stick to the machines or not as it goes through the chip making process. Each corn product requires a different moisture content in the dough. 

Nailing down that exact moisture content number is where Professor Candice Hirsch and her team in the Department of Agronomy and Plant Genetics comes in. Over the last decade, they have collaborated with the multinational company PepsiCo on a number of projects to improve corn traits and technologies for the food industry. They created a machine-learning model that can fairly accurately predict the moisture content in masa dough made from the raw grain scanned when it arrives at a grain elevator. The model is called CHIP-NCM, standing for corn hybrid and inbred prediction of nixtamalization moisture content. 

“From a small sample of raw grain, you can see the need to cook longer, cook shorter or change the temperature,” Hirsch said. “In theory, this will reduce food waste, because with every batch they have to run a test batch through first to see how well it’s doing, leading to food waste. Now they can adjust [their cooking methods] preemptively.”

Corn is a key ingredient in many products and not just food: Toothpaste, clothing, tennis shoes, diapers, and children’s mattresses all contain corn products. Hirsch and her group’s work goes beyond the food industry. They are also working to improve the genetics and resiliency of corn, hoping to help adapt this vital source of food to a changing climate. 

“Corn is used in so many things,” Hirsch said. “How do we produce the corn that is going to withstand climate change and be sustainable and have the properties needed for each product that they’re going into? There’s a lot that happens between understanding the variation in the DNA of this species to having a corn chip in the store, right? Our research touches on all of that.”

Central to it all are the students. In Hirsch’s lab and in the field, they engage in hands-on learning, deepening their knowledge, learning to critically analyze problems and data, creating the next generation of scientists. 

“What we do here is so important,” she said. “When our students leave to go into private industry they are primed and ready to actually design the best experiments to find the best hybrids to give to the grower to grow on their farms.”

Educating people on how to think critically, look at data and critically evaluate data, in turn, “lets us also critically evaluate everything.” And that, she says, is one of the most valuable outcomes of education.

This story is adapted from a longer piece originally published in the Spring 2026 issue of Gleanings from the Department of Agronomy and Plant Genetics. 

Explore how CFANS is elevating extraordinary insights and solutions in Minnesota’s cornfields.