TANG: ai-driven autonomous nutrigation and irrigation for greenhouse tomatoes using plant electrophysiology

Overview

Vivent Biosignals offers an AI-powered crop management platform that uses plant electrophysiology—the electrical signals plants naturally produce—to enable autonomous nutrigation and irrigation in greenhouse and indoor farming operations. The TANG (Tomatoes – Autonomous Nutrigation) solution combines easy-to-deploy biosignal sensors with machine learning models that interpret plant stress responses in real time, allowing automated dosing systems to adjust water and nutrient delivery based on what the plant actually needs.

In preliminary greenhouse trials, autonomous irrigation driven by plant biosignals matched current best-practice fruit quality and yields while using similar or reduced inputs. Nutrient deficiency models for Calcium, Iron, Manganese, and Nitrogen achieved diagnostic accuracies above 95%, identifying stress well before growers could see visual symptoms. This early detection allows proactive intervention, reducing waste and supporting more sustainable intensive horticulture.

Technical specifications

Core components:

• Electrophysiology sensors: Economic, easy-to-deploy devices that capture real-time plant electrical signals
• Machine learning models: Algorithms trained to recognize drought stress and macro/micronutrient deficiencies (Ca, Fe, Mn, N) from biosignal patterns
• Automated relay system: Hardware that interfaces with existing fertigation infrastructure to drive nutrient dosing and irrigation concurrently
• Cloud-based analytics: 24/7 monitoring platform delivering actionable insights to growers

Validated performance:

• Autonomous irrigation trial (n=16 plants, 7 weeks): fruit quality and yields comparable to expert grower best practice
• Nutrient deficiency detection accuracy above 95% (n=15 plants), with early diagnosis prior to visual symptom onset

Applications:

• Greenhouse tomato production
• Indoor vertical farms
• Perennial crop monitoring
• Biostimulant and crop treatment development
• Plant breeding programs

Technology readiness level

The platform is at an advanced validation stage. Core biosignal sensors, machine learning models, and automated control hardware have been developed and deployed in initial greenhouse trials with positive results. Near-term validation activities include expanded greenhouse trials (n=32 plants per trial) covering additional nutrient deficiencies and drought models, supported by leaf tissue analysis and fertigation system metrics (pH, EC). Hardware refinements to enable concurrent nutrient dosing and irrigation control are underway, followed by multi-model autonomous fertigation trials benchmarked against expert grower best practice. The technology is ready for pilot deployments and collaborative refinement with commercial greenhouse operators.