By Zoe KleinmanTechnology Editor
3 hours ago
picture description,
In this demo, the hand is pressing on the screen, but the student’s actual hand is not
A system that can convert human brain activity into actions without physical movement is being developed by a neurotech company called Cogitat.
Wearing a prototype headset, basic actions can be performed in virtual reality by thinking about it.
For example, in a game where a VR jet ski is controlled via handles, you move by thinking about it instead of squeezing your hands.
Elon Musk’s company Neuralink is developing a similar concept.
It’s called the brain-computer interface, and there are a lot of neurotech companies exploring it.
One goal is that it could eventually allow people who have suffered a stroke or other brain injuries to control phones or computers remotely.
Neuralink’s method requires a chip to be inserted into the brain itself. The company has only worked with animals and has been criticized for its treatment. It has released videos showing a monkey playing the video game Pong with its mind and the brain activity of a pig with a chip implanted in its brain.
Cogitat is one of the companies developing a system that works on the head, not in it.
It could one day take the form of a headband worn with a VR headset. Some companies are already developing their own hardware, but as a university spin-off, Cogitat only focuses on the technology behind it.
picture description,
BBC Tech Editor Zoe Kleinman tries out a prototype
It is led by NHS adviser Allan Ponniah and computer scientist Dimitrios Adamos from Imperial College London.
The technology is in the development phase but has already been tested on stroke patients with positive results. The goal is to encourage them to continue with rehabilitation exercises by making them more engaging.
“If a person has had a stroke and cannot move their arm, they are very discouraged to participate in rehabilitation. But our technology allows them to visualize moving their hand and see a hand moving on screen, which we believe will motivate them to start their course of physical therapy,” Mr Ponniah told the BBC Podcast Tech Tent.
Strange experience
I tried it and it’s a very strange experience. For starters, thinking about making a move without actually making it is harder than it sounds. And you also have to try not to think about other things, which increases your brain activity and creates more noise for the tech to decipher as it looks for the engine signal.
I had never before seen my own brain activity on a screen in real time, like an intricate, multi-layered cardiogram. That in itself was strange – seeing the essence of my thoughts in a graph. But hearing that VR jet ski engine roar just because you’ve thought about doing it is an amazing feeling.
Of course, the prototype device couldn’t fully read my mind. It didn’t mean translating my thoughts or looking deep into my soul. It only focused on motor signals.
“If you don’t choose to interact with the system, nothing happens,” says Mr. Adamos. “Nothing will be picked up from you if you stop using it.”
Other companies focus on different types of brain activity — for example, visual cues so you can focus on a number and press buttons on a screen. It’s also possible – but debatable – to focus on more personal answers like likes and dislikes.
Cogitat expects to have a working prototype of its technology within the next 12 months – but there are still some challenges neurotech still has to face.
Experts are still learning about brain activity. It is individual to each of us and not constant. It changes throughout the day and can be affected by factors such as fatigue and dehydration, as well as aging. This means that all brain activity reading systems require continuous recalibration.
Cogitat trains its technology against a database of hundreds of volunteers who have tested it, which speeds up the calibration process. I met some of the team – mainly students who were very enthusiastic, not to mention patient, as they guided me through the demonstration.
Mr. Adamos proudly tells me that in a recent global machine learning competition, Cogitat not only took first place, but also beat a US Army team.
He offered everyone time off to party — but no one took it.
“The next day they were all there,” he says. “It’s really fascinating for us and everyone who has joined this journey.”
You can follow Zoe Kleinman on Twitter @zsk.
