Raman spectroscopy platform for early, non-destructive detection of crop pathogens in tomato plants

Overview

This solution offers an innovative, rapid, and cost-effective approach to high-throughput screening of crops under pathogen attack. It leverages Raman spectroscopy (RS) as a non-destructive tool for in-field detection of plant pathogens, enabling discrimination between healthy and infected plants before visible symptoms appear. By integrating portable RS instruments with machine-learning algorithms, the platform supports precise calibration of pesticide use, contributing to sustainable crop management systems. Initial validation has demonstrated high accuracy in detecting two major tomato viruses, with strong potential for extension to additional devastating pathogens.

Technical specifications

Core technology:

• Raman spectroscopy for non-destructive, in-field pathogen detection
• Portable RS instruments enabling field deployment
• Machine-learning algorithms trained on RS spectral profiles for classification of healthy versus infected plants
• Remote sensing structure for scalable, real-time crop monitoring

Validated performance:

• TYLCSV (Tomato yellow leaf curl Sardinia virus): accuracy greater than 70% at 14 days post-inoculation, before symptom onset
• TSWV (Tomato spotted wilt virus): accuracy greater than 85% at 8 days post-inoculation, before symptom onset
• RS profiles also distinguish between different viruses, providing pathogen-specific diagnostic information
• Virus quantification validated via qPCR in controlled greenhouse conditions

Pathogen scope (planned):

• Tomato brown rugose fruit virus (EPPO A2 quarantine pathogen)
• Tomato leaf curl New Delhi virus (EPPO Alert List)
• Fusarium spp., Verticillium spp., and Alternaria spp.

Collaboration: Developed in partnership with the Italian Institute for Metrological Research (Torino), contributing metrological expertise for calibration and standardization.

Technology readiness level

The technology has completed initial proof-of-concept validation using a bench RS instrument, demonstrating accurate pre-symptomatic detection of two tomato viruses in controlled greenhouse conditions. The next phase involves upgrading to portable RS instrumentation and expanding pathogen coverage to include additional viruses and pathogenic fungi. Unsupervised and supervised data analyses will be applied to refine machine-learning models. Field validation across commercial tomato varieties is planned to confirm performance under real-world agricultural conditions. The platform is positioned at an early-to-mid stage of development, moving from controlled-environment validation toward field-ready deployment.