Electric Plant's AI Model Bombadil Aims to Decode Plant Electrical Signals
Startup raises $11.2 million seed round to build sensors and a language model that reads the electrical activity inside plant stems
By AXY AI Desk ·
Key points
- Electric Plant has built Bombadil, an AI model that aims to interpret electrical signals in plant stems and relay information about plant needs to growers.
- The startup says its sensors cost tens of dollars, compared with existing tools that generally cost thousands.
- The sensors attach to plants with stainless steel screws that penetrate the cambium and send data to a base station over a low-power wireless network.
- Electric Plant raised an $11.2 million seed round led by Onto Ventures and Climactic.
- Independent testing will be needed to confirm how accurately the model interprets plant signals in real growing conditions.
Electric Plant has built an AI model that it says can translate the faint electrical activity moving through plants into information growers can use. The startup, co-founded by Mast, who serves as co-founder and chief executive, and neuroscientist Lee von Kraus, has developed a large language model called Bombadil. The company says the system can interpret the subtle electrical signals running through a plant's stem and relay details about the plant's needs to the people who care for it. Electric Plant has also raised an $11.2 million seed round to bring its sensors and the Bombadil model to market, with Onto Ventures and Climactic leading the round.
The company's central claim is that plants are constantly communicating through electricity, and that people have had very limited ways to listen. Mast told TechCrunch that the idea took hold after he read scientific literature about a houseplant wired to a motorized stand. According to the research he described, when the plant needed more light, electrical signals moved through its living tissues. A control board detected the current and moved the plant closer to the light source. “I’m like, what the heck?” Mast said. “That was the entry to the rabbit hole for me.”
Why Plants Produce Electrical Signals
Plants have long been known to generate electrical signals, although they work differently from the nerve impulses in animals. Plant cells move charged particles such as potassium, calcium and chloride across their membranes, and these movements create small voltage changes. Researchers have studied these signals for decades and have linked them to responses such as wounding, changes in temperature, and reactions to light and water. Scientists still debate how much of this activity carries specific information, and how it is interpreted, which is part of what makes Electric Plant's approach ambitious.
For most of the history of agriculture and forestry, growers have had to infer what plants need from indirect evidence. Electric Plant's pitch is that the electrical activity offers a more direct window into the plant's condition than the tools currently in common use.
The Problem With Existing Plant Monitoring Tools
Mast said that even in the most controlled greenhouses, farmers and growers do not have full visibility into what their plants require. Traditionally, farmers, growers and foresters have relied on coarse tools that respond to the plant's demands for water, nutrients or light. Those tools include soil probes, handheld spectroscopes, which measure how materials interact with light, and satellite imagery.
Mast acknowledged that these tools are useful. “That technology is sophisticated, but its ability to get at the inside of the plant isn’t very sophisticated,” he said. “Most of that stuff still lags quite a bit behind the real-time needs of the plant.” In other words, tools that measure soil, leaves or the sky from outside can reveal stress only after it has already affected the plant. A tool that reads signals from inside the stem could, in principle, detect changing needs sooner.
Cost is another factor. Mast said the startup's sensors are significantly cheaper than existing tools, which generally cost thousands of dollars. “This is tens of dollars,” he said. If that price gap holds in commercial use, it could make continuous monitoring practical for smaller farms and greenhouse operations that cannot afford expensive equipment, though the company has not yet published independent comparisons of accuracy.
How the Sensors and Bombadil Model Work
According to the company, the sensors are about the size of a regular iPhone folded lengthwise. They are attached to a plant with stainless steel screws. Those screws must penetrate the cambium, which is the layer of living tissue just beneath the bark, in order to connect to the microvolt changes occurring within the plant. A microvolt is one millionth of a volt, so the signals being measured are extremely small. Data from the sensors is sent to a base station over a low-power wireless network, where it can be processed.
The data is then used to train and run Bombadil. Electric Plant spent the past three years collecting data to build the model. Because the company started with existing equipment to create its training data, it used those devices to teach Bombadil to interpret electrical signals in about 125 different plant species. Mast said the company expects the model to outperform those instruments over time. “We know the model will surpass a lot of that equipment,” he said.
A large language model is a type of AI system trained on very large amounts of data to recognise patterns and generate outputs. Electric Plant's version applies that pattern-recognition approach to plant electrical data rather than human text. That means the system's usefulness depends heavily on the quality and variety of the data it has learned from, and on how well its interpretations match what happens to plants in real growing conditions. The company's public statements do not yet describe how independently those results have been checked.
Founders, Funding and Investor Backing
The company's founding partnership brings together a startup founder and a neuroscientist. Lee von Kraus has worked with the Defense Advanced Research Projects Agency, known as DARPA, the research arm of the United States Department of Defense that funds advanced science and technology projects. Mast and von Kraus founded Electric Plant to pursue the idea that plant electrical signals could be decoded.
The $11.2 million seed round is an early-stage financing round, typically used by young companies to build a first product and prove that it works. The round was led by Onto Ventures and Climactic, with participation from Ponderosa, Endeavor Catalyst, Ludlow, Michigan Rise, The Fund, Schox VC, and Dug Song. The money is intended to help the company bring its sensors and Bombadil to market.
What the Technology Could Mean for Growers and Foresters
The most immediate use case described by the company is agriculture. Greenhouse operators, who already control light, water and nutrients closely, could in theory use a real-time reading of plant stress to adjust conditions more precisely. Foresters, who manage large areas where soil and satellite data may only show general patterns, could also be interested in tools that report on individual trees. The company has not published details on pricing, deployment timelines or specific customers in the source material available.
The founders go further in their ambitions. They believe the model could one day lead to a tool that lets people “speak” to plants. “We’re listening more than we’re speaking, but listening’s the first step,” Mast said. That goal is far more speculative than the current sensor and monitoring pitch. Translating a plant's electrical state into a clear message about its needs is one challenge; building a reliable two-way interface would be a much larger one. Analysts who follow agricultural technology would likely want to see evidence that Bombadil's readings consistently predict plant outcomes before accepting that framing.
What to Watch Next
Several questions will determine whether Electric Plant's approach becomes a practical tool. The first is accuracy. The company says Bombadil can interpret signals from about 125 species and will improve on existing instruments, but independent testing across different climates, crops and growing systems would show how well those claims hold. The second is cost and deployment. Low-cost sensors that require screws to be placed in the cambium may raise practical questions about how easily they can be installed across large numbers of plants, and how long they remain reliable.
The third question is whether plant electrical signals can be interpreted in ways that help growers make decisions rather than simply generating data. Monitoring is only useful if it leads to timely action. For now, Electric Plant's launch offers a clear proposition: plants send electrical signals, sensors can capture them more cheaply than before, and an AI model can attempt to turn that activity into information about what a plant needs. Whether that proposition works in the field will be the test for the company, its investors and the growers it hopes to serve.
Questions and answers
What is Electric Plant's Bombadil AI model?
Bombadil is a large language model built by Electric Plant to interpret the subtle electrical signals moving through a plant's stem. The company says it can relay information about a plant's needs to growers and is trained on data covering about 125 plant species.
How do Electric Plant's sensors work?
The sensors, about the size of a regular iPhone folded lengthwise, are attached to plants with stainless steel screws that penetrate the cambium. They send data to a base station over a low-power wireless network.
How much funding has Electric Plant raised?
Electric Plant recently raised an $11.2 million seed round led by Onto Ventures and Climactic. Other participants include Ponderosa, Endeavor Catalyst, Ludlow, Michigan Rise, The Fund, Schox VC, and Dug Song.
Can people actually talk to plants with this technology?
The founders say the model could one day lead to a tool that lets people speak to plants, but they describe this as a long-term goal. The current focus is on listening to and interpreting plant signals.
Topics: Electric Plant, Bombadil, AI model, plant electrical signals, agriculture technology, seed funding, plant sensors
Sources: TechCrunch AI