The WheatWatcher project aims to develop digital technologies for monitoring nutrients, chemical and biological actors in soil and plants. On-line and portable soil sensing platforms will be developed to enable the collection of high-resolution data on soil chemical and fertility parameters both spatially and stationary.
Proximal soil sensing will provide soil analysis for agricultural purposes, including in-situ spectrometry, geophysical and electrochemical sensing methods, and an automatic sampling kit based on remote sensing-generated sampling plans. The project utilizes non-invasive sensors, such as electromagnetic induction sensors and ground-penetrating radar, to extract information on soil physical properties like water content and soil texture, and invasive sensors, such as vis-NIRS, for in-situ measurements and various detection methods. Specifically, ultrasound-assisted NaOH extraction will be incorporated for efficient phosphate analysis and utilizes ion-selective electrodes for rapid and precise soil nutrient analysis (Mg2+, NO3-, NH4+, K+, Ca2+, Cl-, EC) and pH measurement. Wet chemical sensors are integrated into the multi-sensor platform to measure various soil parameters such as P, K, OC, moisture, Ca, Mg, and Na using vis-NIR spectroscopy. This multi-sensor ‘on-line’ kit can be fit onto different soil equipment e.g., tillage, planters & seeding machine and operate in any depths between 5 and 50 cm.

Digital soil monitoring system:
+ Successful for measuring organic carbon, moisture, total nitrogen, clay and organic matter.
+ Provides real-time insights into soil properties, such as composition, nutrient levels, and contaminants.
+ An innovative in-situ extraction chamber that performs soil analysis without organic solvents which reduces reliance on costly and time-consuming laboratory-based analyses.
+ Machine learning model will map soil health status.
– Less accurate for pH, phosphorous, calcium cation exchange capacity and magnesium.

The WheatWatcher project develops a mobile crop sensing platform that integrates multispectral and hyperspectral cameras to capture crop images in various spectral bands. This enables real-time assessment of crop health and stress, aiding in the early detection of issues like nutrient deficiencies and diseases. By analySing spectral data, the platform provides actionable insights, optimiSing resource allocation, and improving crop yields while minimiSing environmental impact.

The decision support system records and analyzes data to help land managers make site-specific management decisions, such as soil remediation and precision soil and crop treatments.

The decision support system receives data from:
1) users
2) sensor platforms
3) various available data sources, including UAV, satellite, historical records, national and European soil maps (LUCAS)

Integrating a machine learning model with the decision support system provides rapid insights into wheat growth behavior by modeling current and future crop development.