During the first week of June 2026, we officially kicked off the AgriScienceFM project in Wageningen, bringing together an inspiring consortium of researchers, technology developers, agricultural scientists, data experts and partners from across Europe and beyond.
AgriScienceFM aims to unlock the potential of AI foundation models for agricultural sciences. By developing next-generation models across key dimensions of agriculture - environment, management and plant traits - the project will explore how AI can support scientific discovery and innovation in areas such as crop monitoring, soil health, resilient breeding, precision farming, and sustainable food systems.
The kick-off meeting was an important moment to align our shared vision and begin shaping the foundations of the project. Over the course of three days, partners exchanged ideas, introduced their roles and ambitions, and worked together to clarify how different areas of expertise will connect across the project.
The first day focused on setting the vision, getting to know the consortium, and introducing the main scientific and technical ambitions of the project. Through partner exchanges and plenary discussions, the consortium reflected on how AgriScienceFM can contribute to a new generation of open, trustworthy and agriculture-specific AI tools.
The second day moved into a deeper technical exchange. Partners discussed the relationships between foundation models, data, use cases, users, benchmarks, architectures and scaling. Breakout sessions explored themes such as environmental and management data, plant phenotypes, open data spaces, synthetic data, multimodal alignment, agentic architectures, and the connections between models and real-world agricultural use cases.
The third day focused on governance, communication, dissemination and the immediate next steps. This included discussions on how the project will engage with wider communities, share results, support collaboration, and prepare for upcoming activities.
What stood out most was the shared commitment to interdisciplinary collaboration. AgriScienceFM brings together expertise from AI, agricultural sciences, data infrastructures, responsible innovation, benchmarking, stakeholder engagement and open science. This combination is essential for developing AI systems that are not only technically advanced, but also useful, transparent, responsible and relevant to agricultural research and practice.
A big thank you to all partners for the energy, ideas and constructive discussions during the kick-off meeting. This is just the beginning of an exciting journey towards open and trustworthy AI foundation models for sustainable agriculture.
AgriScienceFM is funded by the European Union under Horizon Europe Grant Agreement No. 101293777.
