Two people have successfully steered a virtual spacecraft by combining the power of their thoughts – and their efforts were far more accurate than one person acting alone. One day groups of people hooked up to brain-computer interfaces (BCIs) might work together to control complex robotic and telepresence systems, maybe even in space.
A BCI system records the brain’s electrical activity using EEG signals, which are detected with electrodes attached to the scalp. Machine-learning software learns to recognise the patterns generated by each user as they think of a certain concept, such as “left” or “right”. BCIs have helped people with disabilities to steer a wheelchair, for example.
Researchers are discovering, however, that they get better results in some tasks by combining the signals from multiple BCI users. Until now, this “collaborative BCI” technique has been used in simple pattern-recognition tasks, but a team at the University of Essex in the UK wanted to test it more rigorously.
So they developed a simulator in which pairs of BCI users had to steer a craft towards the dead centre of a planet by thinking about one of eight directions that they could fly in, like using compass points. Brain signals representing the users’ chosen direction, as interpreted by the machine-learning system, were merged in real time and the spacecraft followed that path.
The results, to be presented at an Intelligent User Interfaces conference in California in March, strongly favoured two-brain navigation. Simulation flights were 67 per cent accurate for a single user, but 90 per cent on target for two users. And when coping with sudden changes in the simulated planet’s position, reaction times were halved, too. Combining signals eradicates the random noise that dogs EEG signals. “When you average signals from two people’s brains, the noise cancels out a bit,” says team member Riccardo Poli.
The technique can also compensate for a lapse in attention. “It is difficult to stay focused on the task at all times. So when a single user has momentary attention lapses, it matters. But when there are two users, a lapse by one will not have much effect, so you stay on target,” Poli says.