Students Use Artificial Intelligence to Advance Crop Research

Gina Garzón Martínez, Larissa van Vliet and Dr Kieran Atkins
30 June 2025
Aberystwyth University’s Institute of Biological, Environmental and Rural Sciences (IBERS), in collaboration with the Department of Computer Science, is training the next generation of crop scientists to use artificial intelligence to tackle global agricultural challenges.
Three research students—Kieran Atkins, Gina Garzón Martínez, and Larissa van Vliet—have recently published groundbreaking work that uses convolutional neural networks (CNNs) to understand how seed production in brassicas and related plants is influenced by the environment. The work, conducted at IBERS’ National Plant Phenomics Centre, used advanced imaging and modelling to study large, genetically diverse plant populations grown under controlled environmental conditions.
The team developed and trained AI models capable of detecting, counting, and measuring seed pods at a speed and accuracy far beyond what is possible manually. This innovation enables researchers to tackle previously intractable questions in crop science and supports Aberystwyth’s longstanding reputation in plant breeding and crop improvement research.
Professor John Doonan, Director of the National Plant Phenomics Centre, said:
“These students are not just advancing academic knowledge—they’re helping develop practical tools that could transform how crops are bred and grown around the world.”
Applications in Focus
- Genetics of Plant Cooperation: Two complementary studies explored how plants interact with their neighbours—whether they compete or cooperate—and how this is influenced by their genetic makeup. Gina Garzón Martínez led a paper published in Proceedings of the Royal Society B, while Kieran Atkins led a study in GigaScience. Together, the research identified several genes linked to cooperative and competitive behaviours, offering new strategies for breeding crops suited to high-density planting systems.
- Cold Response in Oilseed Rape (Canola): The team also developed DeepCanola, a tool to rapidly analyse seed pod traits under winter-like conditions. Used to study oilseed rape, this technology enables breeders to identify cultivars with improved resilience to cold—an increasingly important trait in the face of climate change. This research, by Kieran and Larissa, appears in Computers & Electronics in Agriculture.
Dr Chuan Lu, Senior Lecturer in Bioinformatics at the Department of Computer Science said:
“These are excellent examples of how interdisciplinary science at Aberystwyth is tackling real-world problems.”
These research students were funded by the UK Research and Innovation BBSRCand EPSRCprogrammes, the European Commission’s Horizon 2020programme, and Gina was supported by an AberDoc scholarshipand the president's award for international students.
Impact and Future Careers
The project highlights the value of cutting-edge training and its global impact:
- Gina Garzón Martínez is now a Project Leader at Colombia’s national agricultural research agency, Agrosavia.
- Larissa van Vliet is applying her expertise as a data engineer and project lead in agricultural statistics at the Netherlands’ National Statistics Institute.
- Dr Kieran Atkins continues his research at Aberystwyth, collaborating with our seed breeding partner companies on the even more complex challenges of multi-species cropping systems.
Access the Research
- Unlocking the power of AI for phenotyping fruit morphology in Arabidopsis (GigaScience, 2025): https://doi.org/10.1093/gigascience/giae123
- Uncovering the genetic basis of competitiveness and the potential for cooperation in plant groups (Proceedings of the Royal Society B, 2024): https://royalsocietypublishing.org/doi/10.1098/rspb.2024.1984
- Accelerating phenotyping of seed pod traits in oilseed rape using DeepCanola (Computers & Electronics in Agriculture, 2025): https://www.sciencedirect.com/science/article/pii/S0168169925005769