Low-cost electronic nose and AI models for monitoring plant-soil microbial activity

Overview

The proposed solution leverages a low-cost, portable electronic nose (e-nose) technology integrated with advanced machine learning models to revolutionize the monitoring of plant-soil microbial activity. By capturing and analyzing volatile organic compounds (VOCs) as multivariate "fingerprints," this device serves as a non-invasive proxy for microbial processes and plant physiological status. The e-nose is designed for use in both greenhouse and field environments, enabling continuous and non-destructive monitoring of soil, root zones, and plant tissues. This technology offers a scalable, customizable platform for diverse agricultural applications, aiding in the early detection of plant stressors and optimizing agricultural management practices.

Technical specifications

• E-nose technology: Utilizes a small array of gas sensors to detect VOCs as voltage signals.
• Machine learning integration: AI models translate sensor data into quantitative indicators of microbial and plant activity.
• Deployment capabilities: Suitable for greenhouse and field settings with potential for UAV and UTV platform integration.
• Non-invasive and cost-effective: Provides a low-cost alternative to traditional monitoring methods without destructive sampling.

Technology readiness level

This technology is currently at TRL 5, having been validated in controlled environments and is progressing towards field testing. Future validation efforts will enhance model robustness and establish practical deployment protocols for wider agricultural use.