Dr Talitha Bromwich, Dr Alice Morrell, and Dr Jonathan Rutter have joined the Eric and Wendy Schmidt AI in Science Postdoctoral Fellowship. A programme of Schmidt Sciences, the interdisciplinary programme aims to accelerate the next scientific revolution by supporting talented postdoctoral researchers to apply AI techniques across the natural sciences, engineering, and mathematical sciences.
Environmental footprints for food products
The way we produce and consume food has enormous impacts on climate and biodiversity. But if you’re looking at products on shelf, how can you tell which is less environmentally damaging? Despite seeming like a simple question, it’s one that we still cannot answer for many food products.
Current environmental impact datasets operate at the level of around 120 food categories, but the UK market has more than 100,000 distinct products. At this scale, category averages conceal huge product-specific variation and make manual assessment impossible.
Talitha’s fellowship will apply machine learning approaches to improve our ability to estimate environmental footprints at the product level. To do this, she’ll focus on three areas: linking food products across different large diverse datasets, applying machine learning methods to estimate missing ingredient and environmental information where data are incomplete, and developing transparent, easy-to-use tools to support sustainable food and health research, including through the THRIVING Food Futures project.
Seismic sensing for non-invasive wildlife monitoring
Animal movements reveal how species find resources, avoid predators, and respond to environmental change, yet studying them in the wild is challenging. GPS collars provide valuable data but are invasive, expensive, and limited to a few individuals, while camera traps have a narrow field of view and spatial constraints.
So is there another option for tracking wildlife? When animals move, they generate ground vibrations which can be detected by seismic sensors. Typically used for studying earthquakes, these sensors offer a covert, non-invasive, and potentially real-time method of detecting subtle animal movements.
Using data collected in the Kenyan savanna, Alice’s fellowship will redefine wildlife monitoring by developing AI-driven methods that allow researchers and practitioners to detect and localise animal seismic signals. The project will provide fresh insights into movement ecology and deliver scalable tools for conservation.
Saving seabirds from fishing vessel bycatch
From sharks to turtles to albatrosses, many of the world’s most iconic marine animals face a common threat to their survival – they are frequently caught and killed in fishing gear.
Boat-mounted cameras can use AI to detect if a dead seabird is brought on board. But what about the birds that have not yet been caught – can their movements give us clues about how to reduce their risk of capture?
Jonathan’s fellowship will explore how large flocks of seabirds interact with fishing gear and how well different conservation strategies – like streamer lines to scare birds – reduce risk of seabird death. By examining bird movements rather than counting dead birds, this work will also shed light on why some strategies work better than others and support smarter, more scalable fisheries monitoring to help protect threatened seabird species worldwide.
Find out more about the Schmidt AI in Science Fellowship at Oxford: https://saiis.site.ox.ac.uk/
