Translating Ecological Evidence into Local Action

A Proof-of-Concept AI Tool for Pollination Management

This fellowship will explore whether modern artificial intelligence, and large language models like ChatGPT in particular, can provide practical, easy-to-follow farming advice tailored to local conditions. The project will focus on pollination, which is essential for growing many crops and for maintaining healthy wildlife. Because pollination needs vary between different crops, landscapes and farming methods, one-size-fits-all advice rarely works.

Aims and Objectives

This project will develop a pilot AI tool to see whether we can use it to transform complex ecological knowledge into clear, farm-specific guidance that helps improve pollination on UK farmland.

The fellowship will spark a new collaboration at the University of York, bringing together ecologists who study pollination and sustainable farming with computer scientists working on artificial intelligence and data analysis.

In line with YESI’s ambition to support bold, cross-disciplinary ideas, the project will explore ethical, user-focused ways of using AI to turn environmental research into practical tools that farmers can use.

Working across these fields is vital to make sure AI is developed responsibly—grounded in real ecological knowledge and shaped by the needs and values of the people it is meant to support.

Principal and Co-Investigators

Principal Investigator

Dimitar Kazakov, Computer Science

Co-Investigator

Thomas Timberlake, Leverhulme Centre for Anthropocene Biodiversity

Project Outputs

A core focus of this Discipline Hopping fellowship has been to investigate whether artificial intelligence can translate complex ecological evidence into locally relevant, scientifically sound advice for farming communities. Led by Tom Timberlake alongside Dimitar Kazakov (Computer Science) and Jamie Carr, the project uses pollination management as a key case study to assess both the practical benefits and potential risks of AI tools in supporting sustainable agriculture.

Close-up of a bumblebee collecting pollen from a thistle flower. — Credit: Tom Timberlake

Can AI help farmers improve crop pollination?

Workshop participants reviewing sticky notes around a discussion table. — Participants working through small-group discussions during the project workshop to map out the practical risks and opportunities of AI in decision-making.

Effective management relies on balancing crop types, landscape features, and local farming practices.

To ensure the research reflects real-world perspectives, the team conducted targeted questionnaires, interviews, and a collaborative workshop with farmers, ecologists, and agricultural advisers. These activities explored current AI usage, concerns around data security and farmer autonomy, and key opportunities to steer AI tools towards supporting biodiversity and resilient food systems.

Alongside this engagement, the team is developing and testing a structured approach to AI-generated pollination advice, designed to complement human judgement by bridging the gap between ecological research and local farming expertise.

Farmers, ecologists and agricultural advisers unite at the 'Workshop on Artificial Intelligence (AI) in Sustainable Agriculture' to explore collaborative solutions for resilient food systems.

Workshop attendees stood in front of a presentation screen.

Credit: Jamie Carr

News article: Timberlake, T., Kazakov, D. and Carr, J. (2026) "Exploring the role of AI in sustainable farming."