At a time when there's widespread interest in self-improving AI, researchers here have tackled one of the central challenges affecting its development by using an approach adapted from evolutionary biology.

A new forest vision training simulator developed by a PhD student here is teaching algorithms how to see trees, to help AI tools to calculate how forests are growing and how much carbon they store.

We held Open Days on 9 – 10 July where we welcomed prospective undergraduate students to come and look round, try a taster lecture and find out about the Computer Science degree course here.

A new remote sensing AI tool, Tessera, has shown in trials that it can identify most types of crops on small farms more accurately than current methods do, while using just 8% of their computing power.