Man/Machine Interface Makes Muscles Obsolete
Scientists use brainwaves detected by a skull cap to control a robot over the internet.
Why bother attending all those boring family functions across state lines when you can send your robot proxy instead? A biomedical engineer named José del R. Millán from the École Polytechnique Fédérale de Lausanne has built a semi-autonomous robot that moves around based on the brain activity of a human. The best part is that this human doesn't have to be in the same room - or even have arms or legs anymore - and this kind of "telepresence" opens up the possibility for all kinds of applications from empowering the disabled to aforementioned family reunion problem.
The robot looks a bit like a hollowed out R2-D2, with a laptop and webcam on top running Skype so that the controller can see and hear what's in front of the automaton. Getting around normal obstacles is part of the machine's AI, but the robot can interpret brain activity as detected by EEG electrodes to receive movement instructions. All the subject has to do is think about moving in a direction and the robot would do just that.
Sounds complicated, especially for someone who has lost limbs or who is bed-ridden. Apparently Millán was able to teach two patients more than 100 kilometers away from his facility in Switzerland to operate the robot via the internet with only six hours of instruction over as many weeks. The patients were able to drive the robot around the lab for up to 12 minutes at a time, all while being able to listen to the Millán's guidance.
The applications for this technology are almost endless. Don't want to miss your daughter's graduation while you are orbiting the Earth and conducting experiments on the International Space Station? Send a robot! Want to be present at the birth of your first child but blood makes you squeamish? Send a robot! Feel bad about shooting and killing people in person? ROBOT!
Wait a second ...
Source: Science Mag [http://news.sciencemag.org/sciencenow/2011/09/disabled-patients-mind-meld-with.html]
Permalink
Scientists use brainwaves detected by a skull cap to control a robot over the internet.
Why bother attending all those boring family functions across state lines when you can send your robot proxy instead? A biomedical engineer named José del R. Millán from the École Polytechnique Fédérale de Lausanne has built a semi-autonomous robot that moves around based on the brain activity of a human. The best part is that this human doesn't have to be in the same room - or even have arms or legs anymore - and this kind of "telepresence" opens up the possibility for all kinds of applications from empowering the disabled to aforementioned family reunion problem.
The robot looks a bit like a hollowed out R2-D2, with a laptop and webcam on top running Skype so that the controller can see and hear what's in front of the automaton. Getting around normal obstacles is part of the machine's AI, but the robot can interpret brain activity as detected by EEG electrodes to receive movement instructions. All the subject has to do is think about moving in a direction and the robot would do just that.
Sounds complicated, especially for someone who has lost limbs or who is bed-ridden. Apparently Millán was able to teach two patients more than 100 kilometers away from his facility in Switzerland to operate the robot via the internet with only six hours of instruction over as many weeks. The patients were able to drive the robot around the lab for up to 12 minutes at a time, all while being able to listen to the Millán's guidance.
The applications for this technology are almost endless. Don't want to miss your daughter's graduation while you are orbiting the Earth and conducting experiments on the International Space Station? Send a robot! Want to be present at the birth of your first child but blood makes you squeamish? Send a robot! Feel bad about shooting and killing people in person? ROBOT!
Wait a second ...
Source: Science Mag [http://news.sciencemag.org/sciencenow/2011/09/disabled-patients-mind-meld-with.html]
Permalink
