PaperlayerView exact revision on arXiv

Source-linked AI summary

DS2-Based Cross-Data-Space Interoperability for Precision Agriculture

Katerina Kyriakou, Ilias Syrigos, Ioannis Moutsinas, Panagiotis Tzimotoudis, Thanasis Korakis

arXiv:2609.17185v1cs.DC

TL;DR

Precision agriculture remains constrained by fragmented data and the need to preserve farmers’ sovereignty while enabling interoperability. The paper instantiates DSIA between DigiAgro and AgroScience, producing an automated cross-data-space workflow that delivers field-specific irrigation decision support. The framework supports secure, policy-controlled sharing of agricultural data and services for crop-management recommendations within the demonstrated scope.

  • Problem

    Fragmented agricultural data and requirements for provider control, trust, metadata, and contractual terms limit interoperable precision-agriculture data sharing.

  • Method

    The paper instantiates DSIA as a modular interoperability ecosystem connecting DigiAgro and AgroScience Data Spaces through requirements, onboarding, data sharing, transformation, and service components.

  • Results

    The demonstrated workflow transfers field data to an AgroScience service that returns field-specific irrigation recommendations including irrigation depth, water volume, timing, duration, and confidence.

  • Takeaways & Limitations

    The framework supports secure, sovereign, and transparent agricultural data sharing while keeping farmers in the decision loop over irrigation recommendations.

Abstract

from arXiv · show

Despite the strategies of modern precision agriculture to leverage the integration of legacy agricultural systems, the challenges of IoT data fragmentation, farmers' sovereignty preservation, and limited interoperability still persist. This paper presents our work, conducted within the Horizon Europe DS2 project (DataSpace, DataShare 2.0), that applies an interoperability-oriented framework supporting participants of different agricultural data spaces to share data products and services under secure, sovereign, and transparent methods. The suggested methodology follows a layered reference architecture, where each layer consists of independent operational modules that facilitate the inter-sector data exchange between DigiAgro and AgroScience Data Spaces. The result of this work is an automated ecosystem for sharing diverse farm IoT measurements, satellite images and metrics, weather forecasts, and analytics services across distinct data spaces, aiming to generate accurate recommendations on crop practices, such as irrigation schedules, that farmers and agronomists will rely on to increase crop production while maintaining sustainability.

I. INTRODUCTION

Precision agriculture combines fragmented data from independently managed sources, but sharing must also preserve data sovereignty and trust. This paper instantiates DSIA to connect DigiAgro and AgroScience for secure cross-data-space irrigation support.

  • I. INTRODUCTION: Fragmented IoT, weather, imaging, and analytics data constrain applications such as irrigation planning and crop monitoring.
  • I. INTRODUCTION: Data sharing requires provider control over discovery, access, and subsequent use, while consumers need metadata and explicit contractual terms.These requirements make agricultural data sharing a challenge involving trust and data sovereignty, not only technical integration.
  • I. INTRODUCTION: The paper instantiates DSIA to connect DigiAgro and AgroScience Data Spaces for cross-data-space agricultural data exchange.DigiAgro provides field IoT observations, while AgroScience provides analytics, weather forecasts, satellite images, and metrics.
  • I. INTRODUCTION: The work maps interoperability requirements to DSIA concepts and components.
  • I. INTRODUCTION: A feasibility assessment demonstrates the instantiated workflow through an irrigation decision-support service.

II. RELATED WORK

Prior work has explored cross-data-space interoperability, but comprehensive end-to-end agricultural implementations remain limited. The paper addresses this gap through an intermediary architecture supporting sovereign discovery, policy enforcement, and automated transactions.

  • II. RELATED WORK: Existing cross-data-space interoperability approaches include collaborative, federated, and intermediary-based Data Spaces, each with distinct integration or governance characteristics.Collaborative approaches can introduce substantial technical integration overhead, while federated approaches use centralized governance.
  • II. RELATED WORK: Comprehensive architectures for cross-data-space exchange and widely reported end-to-end implementations remain areas of ongoing research.Related Horizon projects have addressed AI-driven metadata brokers and automated FAIR data pipelines.
  • II. RELATED WORK: The proposed approach enables sovereign discovery, usage-policy enforcement, and automated data transactions with minimal deployment overhead through containerization.
  • II. RELATED WORK: The architecture is grounded in stakeholder needs translated into functional, technical, and operational requirements.Requirements were elicited through collaboration with representatives of the two agricultural data spaces.

A. Functional Requirements

The requirements define the capabilities, safeguards, and deployment conditions needed for interoperable exchange between DigiAgro and AgroScience. They cover publication, transfer, transformation, result delivery, security, sovereignty, and governance.

  • A. Functional Requirements: FR1–FR4 support data-product discovery, cross-data-space transfer, validation and transformation, and decision-support result delivery.Transfers need not require either data space to replace its existing infrastructure, and results may be machine-readable or visual.
  • A. Functional Requirements: TR1–TR6 require standardized transactions, identity validation, machine-readable contracts, data-model interoperability, secure communication, and modular deployment.
  • A. Functional Requirements: OR1–OR3 require data sovereignty, an active tripartite collaboration agreement, and legal and policy compliance support.Cross-boundary transactions can proceed only after the agreement between data-space operators and the DSIA operator is established.
  • A. Functional Requirements: The reference architecture uses a trusted intermediary and decentralized governance to exchange agricultural data while retaining the autonomy of existing data spaces.
  • A. Functional Requirements: The irrigation scenario combines farmer IoT telemetry, usage-policy enforcement, on-the-fly transformation, and containerized deployment on legacy-system hosts.

A. Interoperability Governance System

The interoperability system organizes governance, secure transfer, heterogeneous-data transformation, and AI-driven support into coordinated subsystems. Governance provides the administrative mechanisms for binding participating data spaces.

  • A. Interoperability Governance System: The Interoperability Governance System validates collaboration agreements, issues member identities, manages participant registries, and defines operational rules.Its processes cover operational, collaboration, membership, and conflict management.
  • A. Interoperability Governance System: The system comprises subsystems for secure data transfers, data-pipeline orchestration, and AI-driven support services.
  • A. Interoperability Governance System: The Data Sharing subsystem coordinates contract negotiation, enforcement, and data transfer through an interoperable Connector and supporting trust, sovereignty, and traceability mechanisms.
  • A. Interoperability Governance System: The Data Pipeline subsystem performs on-the-fly transformations across heterogeneous ontologies, languages, and formats during data-sharing transactions.

3) Supporting Services:

The use case combines two complementary agricultural data spaces with modular interoperability services to support data-driven crop management. Independent components and marketplace-delivered services accommodate legacy systems while enabling data sharing and analytics.

  • 3) Supporting Services:: Supporting services help organisations integrate legacy data spaces through automated data-offering preparation, discovery assistance, and trust and compliance analysis.These services include AI-driven offering tools, chatbots for querying and filtering, and risk-analysis tools.
  • 3) Supporting Services:: A service-deployment marketplace automates provider-service preparation, packaging, and publication, while allowing participants to browse, install, and deploy containerized modules.The modules are designed as standalone solutions for existing architectures.
  • 3) Supporting Services:: The applied framework targets sharing agricultural, environmental, weather, and satellite data across legacy systems to improve crop productivity, management, and sustainability.Its evaluation uses an agricultural cross-sector data-sharing scenario centered on an irrigation decision-support service.
  • 3) Supporting Services:: DigiAgro supplies field measurements and crop-monitoring data, while AgroScience provides AI/ML recommendations and third-party weather, satellite, and metric streams.Their cooperation operates within the AgroNIT ecosystem, which preserves participant control over offered data and services.
  • 3) Supporting Services:: The framework uses independent DSIA components and services that can be selected through a marketplace and adapted to participating organisations' requirements.Core modules support data transfer, usage-policy enforcement, and analytics of system execution.

A. Data Acquisition Layer

The data acquisition layer collects and standardizes farm IoT telemetry and retrieves external datasets such as satellite imagery and weather forecasts. These resources are then cataloged and exchanged across data spaces to support agronomic analytics.

  • A. Data Acquisition Layer: Farm sensor measurements are collected and standardized through marketplace-deployed modules before becoming data-space resources.The acquisition process integrates raw IoT data into usable resources through cooperating modules.
  • A. Data Acquisition Layer: The Data Retrieval Module exposes natural-language REST API queries to retrieve third-party datasets such as satellite imagery and weather forecasts at runtime.It reduces the technical complexity of accessing external provider data.
  • A. Data Acquisition Layer: The Catalog, DSIA Connector, and Orchestrator enable discovery, negotiation, peer-to-peer transactions, and automation for sharing acquired resources across data spaces.The shared sensor data supports training and optimizing ML models for agronomist-oriented recommendations.
  • A. Data Acquisition Layer: Usage policies are evaluated during the exchange while automated contract agreements are initiated and mutually accepted to support secure, privacy-preserving transfers.This operationalizes policy control during cross-data-space sharing.

C. Governance, Identity & Logging Layer

The governance layer adds risk assessment, policy control, identity, and immutable auditing to cross-data-space transfers. These components protect data-space rules and make negotiations, workflows, and transfers traceable.

  • C. Governance, Identity & Logging Layer: The framework integrates federated identity, usage-policy enforcement, and immutable logging to strengthen trust verification, regulatory compliance, and security.These purpose-specific components complement the data-sharing modules that perform the actual transfer.
  • C. Governance, Identity & Logging Layer: The Policy Creation component expresses data-access rules in ODRL, compares usage-control policies at runtime, and enforces compliance with data-space rulebooks.It supports governance and fine-grained data-sovereignty control.
  • C. Governance, Identity & Logging Layer: An immutable blockchain ledger records contract negotiations, orchestration workflows, and data transfers for auditing and transparency.The logging mechanism is intended to protect farmer sovereignty by preserving an auditable event history.

D. Onboarding & Deployment Layer

The onboarding and deployment layer verifies participating organisations and exposes shared data products, services, and software modules through a centralized marketplace. Local containerized infrastructure then runs the interoperability stack across heterogeneous environments.

  • D. Onboarding & Deployment Layer: Each data space must onboard to DSIA before deploying its containerized modules across heterogeneous infrastructure.Onboarding is a prerequisite for secure inter-sector exchange between DigiAgro and AgroScience.
  • D. Onboarding & Deployment Layer: The Portal verifies and registers organisations and data spaces, assigns authorized roles, and provides the administrative entry point to the ecosystem.After onboarding, administrators access the centralized Data Marketplace.
  • D. Onboarding & Deployment Layer: The Data Marketplace exposes inter-sector IoT data, crop-monitoring measurements, recommendation services, and deployable software modules.This gives participant administrators a centralized catalog of data products and services.
  • D. Onboarding & Deployment Layer: The Inter-sector Dataspace Toolkit is deployed locally to provide a scalable, data-space-agnostic host environment for Kubernetes and containerized interoperability modules.The final deployment occurs after IDT establishment to reduce installation complexity.

E. Analytics Layer

The Analytics Layer transforms raw farm and crop-monitoring data into standardized, linguistically harmonized inputs for algorithms and personalized farmer decision support.

  • E. Analytics Layer: Raw IoT farm data and crop-growth monitoring data are transformed into user-oriented intelligence for personalized decision support.The layer is intended to deliver personalized decision support to farmers.
  • E. Analytics Layer: The Curation Module transforms and refines farm datasets into standardized formats before downstream processing.
  • E. Analytics Layer: The Culture and Language Module applies automated vocabulary translation and localized concept mapping to create a uniform structure for algorithms.
  • E. Analytics Layer: The Model Development Toolkit lets agronomists and data analysts design, develop, and evaluate reusable ML models for crop management and water requirements.

VI. PRECISION AGRICULTURE CROSS-DATA-SPACE DATA SHARING

The paper demonstrates a DSIA-based workflow for sovereign data exchange between DigiAgro and AgroScience, producing field-specific irrigation recommendations from heterogeneous agricultural data. The workflow combines one-time policy and collaboration setup with recurring retrieval, calculation, and farmer-controlled use of recommendations.

  • Cross-data-space setup: Cross-data-space operation begins with policies, collaboration credentials, marketplace subscription, and field selection authorizing controlled access to selected data products.The setup supports discovery, contract negotiation, and acquisition across the two independently governed data spaces.
  • Cross-data-space setup: Recurring executions reuse valid Endpoint Data References and contract identifiers, avoiding repeated catalogue discovery and contract negotiation.If an EDR expires, is revoked, or becomes invalid, the orchestration process renews it or restarts the required procedures.
  • Irrigation decision-support scenario: Field-specific recommendations combine DigiAgro IoT observations with AgroScience weather forecasts, satellite imagery, and vegetation indicators.The irrigation service uses these complementary resources to estimate crop water requirements for a selected field.
  • Irrigation decision-support scenario: Irrigation estimation accounts for crop evapotranspiration, effective rainfall, root-zone soil-water storage, soil-dependent horizons, system efficiency, and salinity-related leaching.When satellite metrics are unavailable, the service may use a crop coefficient and reference evapotranspiration instead.
  • Recommendation delivery: The service returns field-associated evapotranspiration, crop coefficient, irrigation depth, water volume, start time, duration, and confidence, while farmers may approve, modify, postpone, or reject recommendations.Cancelling the subscription, removing a field, or revoking authorization terminates the corresponding contracts and prevents later access to that field’s data.

VII. CONCLUSIONS

The paper presents an interoperability architecture that bridges existing agricultural data spaces through modular technologies, supporting data sharing and irrigation recommendations. It promotes crop-management decision-making aimed at improving production yield and sustainability, while identifying empirical evaluation and semantic interoperability as future priorities.

  • The interoperability ecosystem bridges DigiAgro and AgroScience Data Spaces to orchestrate field-data collection and generate accurate irrigation recommendations.The architecture uses modular independent technologies to connect edge IoT devices with cross-data-space analytics.
  • The framework promotes intelligent crop-management decisions intended to optimize production yield and ensure sustainability.
  • Future work will empirically evaluate performance and scalability while strengthening semantic interoperability through standardized ontologies and links to external European initiatives.The planned evaluation targets transaction latency and computational resource consumption across field-deployed IoT devices.
Loading 2609.17185v1…