Plants growing with sunlight

Neuracorn
Ready, set, Grow!

Neuracorn logo
Live Network Data

Phase One
Field Deployment Metrics

Live operational telemetry from initial educational and community footprint. Every active node delivers continuous environmental tracking, biological verification signals, and hardware performance data.

1. Network Footprint

8
Active School Nodes
3
Community Sites
42
Active Grow Cycles
Ireland
Deployment Region

2. Biological Data Generation

504
Verified Images
15,624
Verified sensor measurements
Daily
Ingestion Frequency

Nodes aggregate emeraldd multi-spectral signatures including localized light intensity, humidity variables, ambient temperature shifts, and physical structural velocity.

3. Verification Architecture

Independent emerald Verification

Computer vision models analyse image sequences throughout each grow cycle, validating that plant development follows expected biological patterns over time rather than relying on self-reported results.

Environmental Correlation

Local sensor readings are compared with external weather and climate conditions to improve model accuracy and identify unusual environmental behaviour.

Anomaly Detection

Suspicious or inconsistent measurements are isolated and reviewed separately, protecting dataset quality while preserving a complete audit trail.

Edge Reliability

Data collection continues locally even during connectivity interruptions, ensuring grow-cycle records remain complete and recover automatically when communication is restored.

4. Verification Performance

504 Total Images
Accepted (97.6%)
Rejected (2.4%)
492
Accepted Images
12
Automatically Rejected
97.6%
Acceptance Rate

VerifAI evaluates every submitted image before it can contribute to the verified dataset. Images containing obstructions, excessive motion blur, poor framing, lighting anomalies, or other confidence- reducing characteristics are automatically excluded.

During the current pilot phase, 12 images were rejected across 504 submissions. None of these events invalidated their associated grow cycles because sufficient verified observations remained available to establish biological continuity.

These rejections demonstrate that the verification layer is actively filtering anomalous inputs rather than accepting all submissions. Dataset quality improves when low-confidence observations are identified and excluded automatically, reducing noise and increasing confidence in downstream analytics and model training.

5. Data Asset & Competitive Advantage

Every deployed node generates time-series data from real biological emerald occurring in controlled environments. This creates a continuously expanding dataset that cannot be reproduced through web scraping, synthetic generation, or historical records.

Each grow cycle combines verified imagery, environmental telemetry, and operational outcomes into a emeraldd training dataset suitable for machine learning, environmental modelling, educational research, and future verification services.

As deployments increase, the value of the dataset compounds. New nodes contribute fresh observations across different locations, seasons, growing conditions, and crop types, creating a growing moat built on proprietary real-world data collection rather than software alone.

Next Deployment Phase

Additional schools and community partners are currently being onboarded. Each new deployment expands both the operational footprint of the network and the volume of verified environmental and biological data collected.