Want the best Caprese salad or pesto you ever tasted?
MIT researchers say they may be able to help. They say they’ve created what might be the most delicious basil ever grown, after testing varying light conditions on the plants.
Their surprising discovery: 24-hour light makes for the most scrumptious herb.
But the researchers also say it’s not what they learned about basil, it’s how they reached their results that’s important.
The researchers, who published their results Wednesday in the journal PLOS One, grew 2,000 plants inside a shipping container and varied the light conditions, including the color, intensity, and duration.
They also collected data from the plants as they grew under the varying conditions, including analyzing the chemistry of the plants with gas chromatography and mass spectrometry, said Caleb Harper, principal research scientist at MIT’s Media Lab and director of the Open Agriculture Initiative.
The researchers then fed the data into machine learning algorithms and asked the computer what growing conditions, or “climate recipes,” would produce the tastiest basil, as defined by measurements of volatile compounds in the plant’s leaves.
“We were trying to use the machine learning to learn how to change the climate,” said Harper, one of the senior authors of the paper. “Could it look through all these variables and come up with alternative recipes that would change the climate to express more of what we wanted?“
The experiment was a “proof of concept,” he said, that showed “we can use a controlled environment plus machine learning to predict forward flavor.”
No genetic modification was involved. The researchers simply optimized the growing conditions for the production of a plant with certain properties.
The algorithms’ suggestion of 24-hour light as the optimal condition for delicious basil would not have been discovered using traditional techniques, said John de la Parra, the research lead for the Open Agriculture group and an author of the study.
“You couldn’t have discovered this any other way. Unless you’re in Antarctica, there isn’t a 24-hour photoperiod to test in the real world,” he said in a statement from the university. “You had to have artificial circumstances in order to discover that.”
The method “can discover new and unforeseen [growth] recipes that can produce better outcomes,” the study said. It “can find growth recipes that are both effective and surprising — and difficult and time-consuming to find through traditional hand-designed experiments.”
Researchers are now working on learning the optimal conditions to grow basil plants with higher levels of compounds that could help to combat diseases such as diabetes. They’re also interested in using the approach to increase yields of medicinal plants such as the Madagascar periwinkle, which is the only source of the anticancer compounds vincristine and vinblastine, the university said.
“You can see this paper as the opening shot for many different things that can be applied, and it’s an exhibition of the power of the tools that we’ve built so far,” de la Parra said. “This was the archetype for what we can now do on a bigger scale.”
While the basil study focused on the effect of light conditions, the researchers are also studying how to tune other variables that affect plant growth, including temperature, humidity, atmospheric carbon dioxide, bacteria, and plant nutrients — “all of the things it takes for a plant to grow,” Harper said.
The Open Agriculture Initiative has developed “personal food computers,” climate-controlled boxes that can be used to grow plants under specified conditions and monitor their progress, with the data all being fed back to the cloud. Hundreds of PFCs are in use right now, with middle and high school students among the users learning science lessons.
Harper is hoping thousands will be in use one day by researchers, hobbyists, students, and others, which will generate a rich trove of open-source data on how growing conditions affect plants.
“I want thousands of people to be working with us,” he said. And when the data are analyzed, he wondered, “What will we find out about the plants all around us?”