An international competition where cybathletes race against each other using their brainwaves alone has been won by a team led by the University of Bath.

The victory is shared between the team ‘pilot’ and the scientists who developed the technology that enabled him to compete without moving any part of his body.

The brain-computer interface (BCI) race was one of six races held last week as part of the global CYBATHLON Challenge. It involved pilot Owen Collumb – a spinal-injury athlete based in Dublin, Ireland – using ‘movement imagination’ and imagery, along with deep mental focus, to complete a number of computer game-like challenges.

The technology that enables this brainwave control was invented by University of Bath scientists who research and develop headsets that can act as interfaces between the human brain and computers.

The winning team – named NeuroCONCISE – is a collaboration between the Bath Institute of the Augmented Human (IAH) at the University of Bath and Ulster University's Spatial Computing and Neurotechnology Innovation Hub.

It is led by Professor Damien Coyle, head of IAH and Turing AI Fellow, who said: “This victory has been possible thanks to the great determination and of course skills of our team. Our cutting-edge research, combined with intense training of our pilot, gave Owen the chance to showcase remarkably accurate brain-activated commands and control, without physical movement.”

NeuroCONCISE has entered the CYBATHLON Challenge four times since 2016. “This year’s achievement marks a significant milestone for the team, having shown steady progress from their third place finishes in previous years to emerge as champions in 2024,” said Professor Coyle.

The team is affiliated with the company NeuroCONCISE, which was established in 2016 by Professor Coyle to develop high-precision neurotechnology. The company creates wearable electroencephalography (EEG) headsets (similar to the headsets used in medicine to record the electrical activity of the brain) that can be used to assist people in everyday life, sport and medicine.

It is hoped that, in time, this technology will enable a person with tetraplegia or locked-in syndrome to autonomously steer a wheelchair, control a robotic manipulator or use a smartphone, improving the autonomy and social participation of people with severe physical disabilities. The tech also shows promise for helping people with their cognitive and physical rehabilitation, for instance, following a stroke.

Winning team

For the CYBATHLON challenge, Mr Collumb wore an EEG headset linked to a sophisticated AI/signal-processing computer that translated his imagined hand and arm movements into a range of control commands. These commands were then sent to a virtual avatar in order to complete a sequence of complex tasks in record time (4.44 minutes).

One task involved Mr Collumb guiding a wheelchair through a room where it had to avoid obstacles and then leave through the exit. Another involved him steering a mechanical arm holding a cup towards an ice dispensing machine. The task was complete when the cup sat directly beneath the water outlet and all ice was collected by the cup. A third task and final task involved selecting an option out of several in a menu system. Each task had to be completed twice in 6 minutes.

Succeeding in these tasks was a truly remarkable feat for Mr Collumb, who suffered a spinal injury 20 years ago that left him almost completely paralysed from the chest down.

He said: “CYBATHLON has opened many opportunities for me. I was a keen sportsman before my injury and am delighted to have been able to compete and win in this fantastic competition.”

He added: “The research by Professor Coyle and the rest of the team has had a major impact on my life and wellbeing.”

To be close to Mr Collumb’s home, NeuroCONCISE joined the CYBATHLON Challenge remotely, from the Movement Analysis Laboratory at University College Dublin, and was livestreamed to the main arena in Zurich, Switzerland.

As part of his preparation for the competition, Mr Collumb spent many hours training with Professor Coyle and Dr Attila Korik, the team’s race engineer and a researcher from the Department of Computer Science at Bath. He also received expert advice from Dr Lee Moore, a sports-performance psychologist and researcher from the Department for Health, also at Bath.

Dr Korik said: “This year’s competition was particularly challenging as the tasks were all new and involved complex control scenarios that required Owen to learn new approaches. We also had to develop strategy and innovate new controllers to ensure we could first complete the commands to score maximum and then accelerate the speed of control to achieve winning times.

“The atmosphere during training sessions was highly focused and reminiscent of a Formula One race.”

Dr Moore, who taught Mr Collumb to moderate his breathing to optimise his performance, said: “We focused on ensuring Owen had mechanisms and tricks to get into the zone in each race.

“For everyone, brain activity is variable, and anything can influence its dynamics, but this variability can cause problems for the AI that tries to make sense of the brain activity. That’s why it was important for Owen to stabilise his mind and maintain focus even under intense pressure on the day of competition. Our approach to regulating his breathing helped him to relax and perform to his best.”

The main event

Following their victory, the team will immediately begin training for CYBATHLON 2024 – the main CYBATHLON event, scheduled to take place in October in an arena in Zurich, where there will be more pressure from additional teams and a live audience of over 2,000 audience members, which is expected to create an amazing atmosphere.

This non-profit event – hosted by the university ETH Zurich – challenges universities, companies and NGOs to develop assistive technologies (systems that support and help people with disabilities) to accomplish everyday tasks using only their brain activity.

Professor Damien Coyle said: “Our plan now is to capitalise on everything we’ve learned over the past few months to further innovate and improve our brain-computer interface and develop a training regime that will give us another victory in October.

“Owen and the team are fully aware of the effort, sacrifices and commitment required but now, having tasted victory and achieved substantial advances in our neurotechnology framework, we’re more eager to compete and win than ever."

Other members of the NeuroCONCISE team are Dr Naomi du Bois, , Mr Hardeep Sidhu, Mr Robert Clarke and Ms Eesha Shah from the University of Bath and Dr Karl McCreadie and Mr Niall Shane from Ulster University.

The research developed by NeuroCONCISE is supported by the Alan Turing Institute and the EPSRC.