Purpose

This use case addresses the integration of real-time agricultural data, AI services, and cross-sector collaboration within a federated digital environment. It focuses on precision farming in Northern Greece, where diverse actors, such as farmers, agronomists, researchers, service providers, and the agri-food industry, interact through a shared infrastructure. The scope is intentionally broad in technological reach yet specific in its application context: improving crop productivity, resource efficiency, and decision-making, while maintaining full control over data assets.

​

Requirements Co-creation and Architecture Validation

Precision agriculture delivers IoT-intensive requirements with sensor and satellite data, creating the most datarich validation environment.

​

Offered Service

The service supports precision farming by offering AI-driven recommendations on irrigation, crop monitoring, and sourcing logistics. It helps farmers, agronomists, and cooperatives make real-time, data-informed decisions to boost productivity and sustainability. Delivered through the AgroNIT web platform, the service combines dashboards, AI models, and a federated data marketplace. Core data products include IoT sensor data, high-res crop images, weather forecasts, and satellite indices. These are processed across two dataspaces: DigiAgro data space (run by UTH) and AgroScience data space (run by CROWNEST).

​

Dataspaces

Two distinct dataspaces support this system: DigiAgro DS and AgroScience DS. AgroScience DS, which is operated by CROWNEST SA, delivers analytics, decision support tools, and consultancy services by integrating external datasets, such as satellite imagery, short-term forecasts, agronomic baselines, and making them actionable. The DigiAgro DS, operated by the University of Thessaly, serves as a dedicated infrastructure for the acquisition, management, and controlled sharing of agricultural field data.

​

Complex Data Lifecycle and Regulatory Compliance Testing

Precision Agriculture creates federated learning scenarios requiring distributed data control across agricultural cooperatives.

​

Non-functional Properties

• Data sovereignty: Farmers retain ownership and control over their data at all times.

• Privacy preservation: Sensitive and proprietary data is not shared unless explicitly permitted; Federated Learning reduces exposure.

• Compliance: All operations align with GDPR and European data governance frameworks (IDSA, GAIA-X).

• Scalability: The system architecture supports increased data volume, additional farms, and new service providers.

• Interoperability: Based on open standards and modular integration, supporting future cross-sector extensions.

​

Functional Properties

• Real-time monitoring of environmental and crop conditions via IoT sensors and cameras.

• Visualization of weather forecasts, field metrics, and crop imagery.

• Discovery, publication, and consumption of datasets and digital services within a federated marketplace.

• Secure data exchange between independently operated dataspaces (DigiAgro DS and AgroScience DS).

• Role-based access and dynamic policy enforcement for data usage and service invocation.