The research behind
GreenMind.

🌱 Relevance to GreenMind

Flexible organic multi-electrode arrays map Venus flytrap action potentials at high resolution. The study demonstrates precise, non-invasive measurement of plant bioelectrical signals, providing a technical foundation for GreenMind's sensing approach.

🌱 Relevance to GreenMind

Peter Gloor shows that bioelectrical signals from a purple-heart plant can classify environmental conditions (85.4% accuracy) and human emotions (73%). The study demonstrates the potential of machine learning for analysing plant signals of the kind recorded by GreenMind.

🌱 Relevance to GreenMind

Peter Gloor demonstrates measurable differences in plant bioelectrical activity when a person moves nearby. Random-forest classifiers achieve 62.7% accuracy on 2,978 samples across three plant species, providing further evidence that plant signals respond to environmental changes.

🌱 Relevance to GreenMind

Electrical impedance measurements on plant stems detect water stress with up to 95% correlation with soil moisture. This low-cost approach measures plant needs directly, aligning with GreenMind's sensing approach.

🌱 Relevance to GreenMind

Electrophysiological signals are used to detect nutrient deficiencies in greenhouse plants. This supports using the sensing approach behind GreenMind to investigate nutrient stress as well as water stress directly in plants.

🌱 Relevance to GreenMind

This foundational study shows that plants respond to environmental stress through action and variation potentials, including responses to pests, pesticides and heavy metals. It provides a scientific basis for using plants as living biosensors, as explored by GreenMind.

🌱 Relevance to GreenMind

A comprehensive review of methods for detecting heavy metals in plants, including non-invasive measurements. This is relevant to GreenMind because heavy-metal exposure affects plant biomarkers and electrical signals.

🌱 Relevance to GreenMind

Hedera helix is fitted with a PhytoNode wearable that records electrophysiological signals outdoors for five months. AutoML classifiers achieve an F1 score of up to 95%, demonstrating the potential of plant electrical signals for environmental monitoring under real conditions.

🌱 Relevance to GreenMind

Machine-learning classifiers (random forest and SVC) distinguish stages of tomato ripening from electrophysiological signals with more than 80% accuracy. The study shows that the plant electrome, as recorded by GreenMind, can also reflect developmental processes in fruit.