Stanford engineers leverage AI and mechanical testing to bridge the texture gap for better plant-based alternatives, potentially paving the way for a sustainable gastronomic future.
In a breakthrough study that could revolutionize the future of plant-based diets, Stanford University engineers have combined mechanical testing with machine learning to more accurately replicate the texture of animal meats in plant-based alternatives.
The findings, published in the journal npj Science of Food, suggest that such methods can expedite the development of plant-based meats that closely mimic their animal counterparts, promising to make plant-based diets more appealing to devoted meat-eaters.
“We were surprised to find that today’s plant-based products can reproduce the whole texture spectrum of animal meats,” senior author Ellen Kuhl, a professor of mechanical engineering at Stanford, said in a news release.
The environmental benefits of reducing our reliance on animal meats are well-documented. Industrial animal agriculture is a significant contributor to climate change, pollution and habitat loss, among other issues.
According to one study, plant-based meats generally have half the environmental impact of animal meats. Yet, persuading die-hard meat lovers to switch remains a challenge. About a third of Americans, in one survey, indicated they were “very likely” or “extremely likely” to buy plant-based alternatives.
“People love meat. If we want to convince the hardcore meat eaters that alternatives are worth trying, the closer we can mimic animal meat with plant-based products, the more likely people might be open to trying something new,” lead author Skyler St. Pierre, a doctoral student in mechanical engineering, said in the news release.
Revolutionizing Food Texture Testing
Historically, food texture testing methods have been inconsistent and often not shared with the broader scientific community, limiting collaboration and innovation. This is where the Stanford team’s approach stands apart.
The research, which began as a class project by St. Pierre, involved analyzing the mechanical properties of both animal and plant-based hot dogs, sausages, turkey and tofu.
The researchers used a machine to pull, push and shear the samples — actions that simulate chewing — and processed the data through a newly designed neural network.
“These three loading modes represent what you do when you chew,” Kuhl added.
The results were compelling. The mechanical tests showed that plant-based hot dogs and sausages behaved similarly to their animal counterparts in terms of stiffness and texture, findings that were echoed by human testers in a subsequent survey.
A Data-Driven Future for Plant-Based Innovation
“What’s really cool is that the ranking of the people was almost identical to the ranking of the machine,” added Kuhl. “That’s great because now we can use the machine to have a quantitative, very reproducible test.”
This convergence suggests that machine learning can play a crucial role in advancing plant-based meat development. The Stanford team envisions using AI to generate recipes for plant-based meats with precisely desired properties, reducing the trial-and-error phase typical in food science.
The potential impact of this research extends beyond the lab. By sharing their data online, the team hopes to foster greater scientific collaboration and accelerate innovation in the field of plant-based foods.
“Historically, some researchers, and especially companies, don’t share their data and that’s a really big barrier to innovation,” St. Pierre added. Without a shared database and collaborative efforts, though, “how are we going to come up with a steak mimic together?”
The project is ongoing, with new tests already being performed on veggie and meat deli slices, and plans to assess engineered fungi developed by Vayu Hill-Maini, an assistant professor of bioengineering at Stanford.
“If anybody has an artificial or a plant-based meat they want to test,” Kuhl said, “we’re so happy to test it to see how it stacks up.”