Periodic Reporting for period 1 - FARMWISE (Future Agricultural Resource Management and Water Innovations for a Sustainable Europe)
Berichtszeitraum: 2024-01-01 bis 2025-06-30
The FARMWISE project operates at the intersection of agriculture, water management, and climate change, addressing challenges that threaten Europe’s environmental health and food security. Agriculture is a major consumer of water and contributes to pollution through nutrient runoff, leading to eutrophication and habitat degradation. Climate change intensifies these pressures through more frequent droughts and floods. In this context, sustainable and resilient water management practices are urgently required.
FARMWISE develops socio-technical solutions that combine advanced data science with participatory governance. It integrates AI-driven modelling and decision support with insights into social, economic, and political systems. The project tests 11 agricultural innovations across eight diverse case study regions representing different climates, farming systems, and regulatory contexts. This ensures scientifically robust, relevant, and scalable solutions.
Objectives are to:
• Develop a FAIR-compliant data infrastructure unifying agricultural, hydrological, and climate data.
• Build and validate AI and physics-based models simulating management and climate scenarios.
• Co-create governance strategies through stakeholder engagement and systems thinking.
• Empower local actors with training in systems thinking and facilitation.
• Deliver AI-powered decision support tools integrating environmental, socio-economic, and policy data.
Integration of Social Sciences and Humanities (SSH)
FARMWISE embeds SSH as a core pillar, recognising water management as a social as well as technical challenge. WP1 uses tools such as the COSMIC platform, stakeholder mapping, participatory workshops, and governance analysis to align technologies with real-world needs. By identifying policy barriers and behavioural drivers, and fostering dialogue, the project ensures outputs are socially robust, ethically sound, and more likely to be adopted.
Pathway to Impact
Impact is designed to scale from local innovation testing to systemic change. Solutions validated in case studies generate insights transferred to policy recommendations and decision-support tools. Contributions align with the EU Green Deal, Farm to Fork Strategy, EU Biodiversity Strategy, and Water Framework Directive. Environmental benefits include cleaner water and resilient ecosystems, while economic benefits support farm sustainability. Societally, FARMWISE strengthens collaborative water governance and bridges science, policy, and practice.
FARMWISE develops socio-technical solutions that combine advanced data science with participatory governance. It integrates AI-driven modelling and decision support with insights into social, economic, and political systems. The project tests 11 agricultural innovations across eight diverse case study regions representing different climates, farming systems, and regulatory contexts. This ensures scientifically robust, relevant, and scalable solutions.
Objectives are to:
• Develop a FAIR-compliant data infrastructure unifying agricultural, hydrological, and climate data.
• Build and validate AI and physics-based models simulating management and climate scenarios.
• Co-create governance strategies through stakeholder engagement and systems thinking.
• Empower local actors with training in systems thinking and facilitation.
• Deliver AI-powered decision support tools integrating environmental, socio-economic, and policy data.
Integration of Social Sciences and Humanities (SSH)
FARMWISE embeds SSH as a core pillar, recognising water management as a social as well as technical challenge. WP1 uses tools such as the COSMIC platform, stakeholder mapping, participatory workshops, and governance analysis to align technologies with real-world needs. By identifying policy barriers and behavioural drivers, and fostering dialogue, the project ensures outputs are socially robust, ethically sound, and more likely to be adopted.
Pathway to Impact
Impact is designed to scale from local innovation testing to systemic change. Solutions validated in case studies generate insights transferred to policy recommendations and decision-support tools. Contributions align with the EU Green Deal, Farm to Fork Strategy, EU Biodiversity Strategy, and Water Framework Directive. Environmental benefits include cleaner water and resilient ecosystems, while economic benefits support farm sustainability. Societally, FARMWISE strengthens collaborative water governance and bridges science, policy, and practice.
In its first 18 months, FARMWISE has built strong foundations for data-driven and co-created water management.
Data Infrastructure and Tools (WP2)
A review of 135 European water datasets identified 61 with long-term records. An open-source tool with 17 API wrappers was created to unify access. Deliverable 2.1 presented the FARMWISE database design, storing internal data and outputs from innovation testing, and a prototype mapping tool for visualising water availability and quality.
Innovation Assessment and Modelling (WP3 & WP4)
WP3 analysed 11 agricultural innovations improving soil moisture, nutrients, and reducing runoff. Deliverable 3.1 defined monitoring variables. WP4 applied AI, including Random Forests, to 27 climate indices identifying stress thresholds for potato yields. A hydrological model (HydroGeoSphere) tested management and climate scenarios, showing Controlled Drainage with Subirrigation could reduce water use without yield loss. WP3 developed a pipeline for 4,000 nitrate leaching simulations and pesticide transfer scenarios.
Decision Support and Policy Modelling (WP5)
An AI framework was validated to improve water quality monitoring, producing reliable risk maps and transferable predictions. The Visual Decision Support System (VDSS) was advanced with React architecture, six visualisation components, and a Query-to-Vis engine for natural language dashboards. A FastAPI backend delivers model outputs. Integration standards were defined with OpenAPI 2.1 and data formats. A conceptual AI-based policy modelling module was initiated.
Data Infrastructure and Tools (WP2)
A review of 135 European water datasets identified 61 with long-term records. An open-source tool with 17 API wrappers was created to unify access. Deliverable 2.1 presented the FARMWISE database design, storing internal data and outputs from innovation testing, and a prototype mapping tool for visualising water availability and quality.
Innovation Assessment and Modelling (WP3 & WP4)
WP3 analysed 11 agricultural innovations improving soil moisture, nutrients, and reducing runoff. Deliverable 3.1 defined monitoring variables. WP4 applied AI, including Random Forests, to 27 climate indices identifying stress thresholds for potato yields. A hydrological model (HydroGeoSphere) tested management and climate scenarios, showing Controlled Drainage with Subirrigation could reduce water use without yield loss. WP3 developed a pipeline for 4,000 nitrate leaching simulations and pesticide transfer scenarios.
Decision Support and Policy Modelling (WP5)
An AI framework was validated to improve water quality monitoring, producing reliable risk maps and transferable predictions. The Visual Decision Support System (VDSS) was advanced with React architecture, six visualisation components, and a Query-to-Vis engine for natural language dashboards. A FastAPI backend delivers model outputs. Integration standards were defined with OpenAPI 2.1 and data formats. A conceptual AI-based policy modelling module was initiated.
During the first part of the project (M1-M18), the FARMWISE project has operationalised the COSMIC platform, which has mapped complex stakeholder networks to reveal critical interdependencies, influence structures, and information flows. This achievement embeds social intelligence into the project’s technological design, ensuring that subsequent tools are developed with a deep understanding of the socio-technical landscape.
Building on this foundation, the project is currently advancing two additional innovations scheduled for completion by M36:
1. A Unified Data Ecosystem (In Progress): Development is underway for an open-source tool to harmonise 135 datasets. This will be supported by AI models, including Bayesian deep learning, to predict water quality in data-poor areas, enabling predictive rather than reactive management.
2. The VDSS (In Progress): The project is finalising the Query-to-Vis engine (Visual Decision Support System). Once launched, this will lower barriers to data use for non-experts, facilitating evidence-based decisions at both farm and policy levels.
Needs for Further Uptake
Further progress requires filling data gaps, demonstrating innovations at scale, developing business models and financing, balancing IPR with Open Science, adapting outputs to diverse contexts, and engaging in regulatory standardisation and GDPR compliance.
Building on this foundation, the project is currently advancing two additional innovations scheduled for completion by M36:
1. A Unified Data Ecosystem (In Progress): Development is underway for an open-source tool to harmonise 135 datasets. This will be supported by AI models, including Bayesian deep learning, to predict water quality in data-poor areas, enabling predictive rather than reactive management.
2. The VDSS (In Progress): The project is finalising the Query-to-Vis engine (Visual Decision Support System). Once launched, this will lower barriers to data use for non-experts, facilitating evidence-based decisions at both farm and policy levels.
Needs for Further Uptake
Further progress requires filling data gaps, demonstrating innovations at scale, developing business models and financing, balancing IPR with Open Science, adapting outputs to diverse contexts, and engaging in regulatory standardisation and GDPR compliance.