TL;DR

University of Essex researchers have been granted 10,000 hours on Isambard-AI, one of the UK’s most powerful AI supercomputers, for a £1.2 million study investigating whether mental fatigue and concentration lapses can be detected through eye behaviour.

The EyeWarn Project

The project, dubbed EyeWarn, will collect and analyse eye movements in natural settings. Researchers aim to gain detailed understanding of how the eye behaves, reacts and is influenced by cognition—potentially offering new methods to predict fatigue levels.

“Thanks to the AI supercomputer, we will be able to quickly assess and predict fatigue levels and offer new insights into how cognitive strain affects behaviour,” said Dr Javier Andreu-Perez, the research lead and computer science lecturer at Essex.

Supercomputing Power

The Isambard-AI system will power exploration of large-scale models integrating data from people and their surroundings. This processing capability enables analysis that would be impractical with conventional computing resources.

“This project represents a significant advancement in the UK’s scientific efforts in human-centred artificial intelligence,” Dr Andreu-Perez explained. “EyeWarn places a strong emphasis on enhancing our understanding and monitoring of human factors, such as fatigue and its effects on the body through the eye.”

Collaborative Research

Academics from across the UK are collaborating with AI company the Solvemed Group on the research. The team wants to bring “innovative research into real-world settings,” according to the university, which has campuses in Colchester, Southend-on-Sea and Loughton.

Essex’s AI Ambitions

The region is positioning itself at the forefront of UK AI innovation. A £2 billion data centre is planned for Loughton, potentially establishing Essex as a significant hub for AI research and development.

Looking Forward

If successful, EyeWarn could have applications in workplace safety, driving, healthcare and any context where detecting mental fatigue before it affects performance could prevent accidents or errors. The research represents a growing focus on human-centred AI that monitors wellbeing rather than simply automating tasks.

Source: BBC News