Electric Plant uses AI to interpret plants’ electrical signals

Electric Plant has developed Bombadil, an AI model it says interprets plants’ electrical signals, and says the system was trained on more than 4 million hours of recordings spanning 100 species and varied environments.

The company’s account, carried in an automated aggregation by newscord.org, presents the model as reading plant health directly from electrical activity rather than estimating it from soil, weather or visual cues. Electric Plant also described the hardware used to gather those signals.

Its Bud sensor attaches to a plant and records electrical activity in its tissue, including signals associated with stress, the company said. The company said it built the hardware and AI components needed to capture and interpret plant signals across species.

The approach grew out of co-founder Justin Mast’s interest in research involving a houseplant connected to a motorized stand, according to TechCrunch. Mast, Electric Plant’s co-founder and CEO, characterized the company’s present work as listening to plants, not communicating with them.

He called listening an initial step toward the possibility of tools that might eventually let people “speak” to plants. Mast said growers do not have a complete picture of what plants need, even inside controlled greenhouses.

He also argued that existing growing technology does not reveal plants’ internal conditions with enough sophistication to keep pace with their real-time needs. Those are Mast’s assessments of the problem Electric Plant aims to address, rather than independently established measurements in the reporting.

Electric Plant said it raised an $11.2 million seed round to bring its sensors and Bombadil to market. TechCrunch reported 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.

Business Wire and StreetInsider headlines also reported the company’s launch from stealth and the $11.2 million financing. Mast claimed that a system of this kind could raise yields while cutting water use by 20% or more. TechCrunch also quoted him describing Bombadil as a particularly diverse model for biological electrical signals.

The reported coverage does not independently validate the performance or water-use claims. The reports emphasize different parts of the project: TechCrunch focuses on Mast’s explanation of the motivation and the potential agricultural benefits, while newscord.org foregrounds the company’s stated training scale, sensor operation and technical architecture.

The material does not establish a product launch date or document measured results from growers using Bombadil.