Microwave mobile scanning device for non-intrusive soil carbon measurement

Overview

Soil carbon, both organic and inorganic, plays a critical role in agricultural productivity and climate regulation, yet measuring it accurately and affordably remains a major challenge. Conventional soil carbon testing requires physical sampling, laboratory processing, and significant time and cost, limiting the scale and frequency of measurements. This solution addresses that gap with a smartphone-sized, mobile microwave scanning device that transmits microwave signals into the soil, captures the unique reflectance response, and uses on-device machine learning to identify and quantify soil carbon in real time. By enabling rapid, non-intrusive, in-field measurement, the technology supports better soil management decisions, helps track carbon losses from erosion and oxidation, and provides a scalable tool for monitoring soil health across farms, rangelands, and restoration sites.

Technical specifications

• Microwave sensing approach: Generates microwave signals directed at the soil and receives the reflected signal, which varies based on the structure of soil organic and inorganic carbon.
• Edge AI processing: A compressed deep neural network runs locally on an embedded edge computing device, allowing the system to process, label, and predict soil carbon content in real time without requiring cloud connectivity.
• Form factor: Designed as a portable, smartphone-sized handheld unit suitable for field use by agronomists, researchers, and land managers.
• Drone integration potential: The portable device can be mounted on a drone to function as a microwave soil carbon camera, enabling aerial, large-area scanning and scaling toward cost-effective regional soil carbon assessment.
• Validation status to date: The underlying microwave sensor and supervised deep learning pipeline have been validated on other media, including soil moisture and lithium battery charge status, demonstrating that the sensing and inference approach works. Controlled experiments for soil carbon are currently underway, and a publication is in preparation.

Technology readiness level

The technology is at an early-to-mid stage of development. Core microwave sensing and deep learning components have been demonstrated on related applications, and controlled validation experiments targeting soil carbon are in progress. The next phase focuses on building a portable prototype that combines microwave signal generation, reflectance capture, and on-device machine learning inference, followed by drone-based aerial scanning for scale-up. The research team estimates a one-year development timeline with the right support and is seeking funding and collaboration to advance the device toward field-ready deployment.