Imagine about a motorized wheelchair that a completely paralyzed person could control simply by thinking about it. Such wheelchair will transform the way people with severe disabilities spend their life. No this is not just an imagination. Scientists at University of Zaragoza, Spain, have actually developed such a thought-controlled wheelchair.
The thought-controlled wheelchair has a laser sensor and a screen that displays a real-time, three-dimensional virtual reconstruction of the wheelchair's surroundings. The individual thinks of a particular word and a sensor collar around the user's neck detects the electrical signals sent from their brain to the larynx muscles. To steer the chair, a user concentrates on the part of the display where he or she wants to go, and electrodes in a skullcap detect the user's brain activity and work out the destination, the researchers said.
Volunteers took just 45 minutes to learn how to use a prototype chair safely and accurately, said University of Zaragoza associate professor Javier Minguez. The wheelchair prototype can handle only two thought commands a minute and can be used for only about two hours since the wet gel used to fix the electrodes to a user's head dries and loses its effectiveness.
But Dr Minguez hopes that a faster and more sophisticated version of the mind-controlled wheelchair could be developed and soon will be released for commercial production. So, for the arrival of the amazing wheelchair we will have to wait some more years.
1 comments:
Another interesting result of this excellent group. However, as shown in http://www.electrical-neuroimaging.ch/download.html (as well as on the page of Javier MÃnguez) some wheelchairs does not really need a subject to guide them. Thus the question: who is really controlling the wheelchair, the BCI or the Artificial Intelligence? The only way to answer that is making a demo without AI.
Another question is whether a P300 might yield the response speed and precision needed for a control without AI?.
For alternatives addressing these issues see: http://www.youtube.com/RolandoGrave
where the first (06-03-2009) control of a robot via internet was illustrated (Presented also at ESANN09).
Post a Comment