If plants could talk, they might be able to help us better understand environmental conditions. A new sensor was designed to do just that by using AI to listen to and translate plant-speak into data humans can understand.
Bud, from Electric Plant Co., is a lightweight, low-cost handheld sensor that attaches to plants through two probes. It detects and interprets the electric signals that plants naturally produce. Two stainless steel screws hold the sensor in place, and it can be affixed to botanicals whose stem is at least an inch thick. When used on trees, it’s sensitive enough reach into the cambium layer, which divides the bark from the wood.
Like all living things, plants produce electricity, and the company says by tracking and decoding the electric signals that plants give off in response to environmental stimuli, the sensor can gather data about conditions such as hydration and stress. “I actually think of this as kind of a dongle,” Electric Plant’s cofounder and CEO, Justin Mast, tells Fast Company.
“The plant itself is the sophisticated technology,” Mast says. “Plants and trees have millions of little microreceptors that are processing information about the world around it and then they’re coordinating a lot of their physical responses through electrochemical signals, and so that’s what we’re tapping into and listening to.”
Other companies, too, are working to translate plant needs into plain language. One of them, Vivent Biosignals, a Switzerland-based plant intelligence company that also uses AI, announced it surpassed more than 2 million monitored plant days in July. The founder of a U.K. startup called Smarty Plants, which makes sensors for houseplants that go in the soil, pitched his product on an episode of a BBC business reality show that aired last month.
Electric Plant emerged from stealth with $11.2 million in seed funding on Tuesday. Its artificial intelligence model, called Bombadil, is trained on more than 4 million hours of electrical signals from more than 1,400 plants across 100 species, ranging from tomatoes, coleus, and common garden plants to sequoias, apple trees, almond trees, and oaks.
The model estimates seven physiological stress markers from a plant’s electrical signals, like the operating efficiency of photosystem or the temperature differential between leaves and air, and turns those into insights about plant conditions. In a world that’s getting warmer and with prices for things like fertilizer on the rise, the technology offers the possibility of more efficient planting with a higher success rate.
Mast, who also founded the direct-to-consumer houseplant delivery service Bloomscape, says while working there he learned firsthand how hard it is to tell people what plants need at any given time. That depends on factors like how much sunlight a plant gets, what part of the world you’re in, how big the plant is, and whether it’s near a radiator.
“It’s so contextual,” he says, and that also goes for forestry and agriculture.
“There you have people who are way more skilled, but it’s actually really hard to scale that kind of attention and that kind of understanding [of] what plants need and want,” Mast says. Changing conditions like watering schedules or nutrients can take weeks or months to show up visually, but with its system, Electric Plant promises faster knowledge that unlocks operational possibilities.
“Fundamentally, it’s just about getting access to the inside language that’s flowing through plants and trees so that we can better understand the experience of the world and how to better care for them,” Mast says.