Plant Signals Translated: New AI Listens to Flora

Electric Plant, a startup co-founded by Justin Mast and neuroscientist Lee von Kraus, is developing a novel way to understand plant needs. The company has spent three years collecting data to train an AI model called Bombadil. This model interprets the subtle electrical signals within plants, aiming to translate them into actionable insights for growers.

The company recently announced a $11.2 million seed funding round. This investment, led by Onto Ventures and Climactic, will support the market launch of their sensors and the Bombadil AI. Mast believes this technology will provide greater visibility into plant requirements than current methods.

Existing tools for monitoring plant needs, such as soil probes and satellite imagery, are sophisticated but often lack real-time accuracy regarding a plant's internal state. Electric Plant's sensors, priced at tens of dollars, are a fraction of the cost of conventional equipment that can run into thousands. These sensors attach to plants, connecting to living tissue to detect microvolt changes.

Data from the sensors is wirelessly transmitted to a cloud-based platform where Bombadil analyzes it. The interpreted results are then sent to customers, enabling adjustments to irrigation or fertilization schedules. Mast claims this system can increase crop yields and reduce water consumption by over 20%.

Bombadil has been trained on data from over 125 plant species, including ferns, tomatoes, and dahlias. Mast highlights this as the most diverse model created for biological electrical signals, extending beyond what has been achieved in the animal kingdom. Applications are anticipated in agriculture and forestry, with initial customers yet to be disclosed.