Brain Waves Guide Walking Robot
From: Discovery Channel - 01/10/2007
By: Tracy Staedter

Researchers at the University of Washington in Seattle are developing a way
for people to control a robot using only their thoughts. Associate professor
of computer science and engineering Rajesh Rao says, "We're using a
well-known, well-characterized response that occurs in the brain to control a
physical device in the world." This reliable response is known as a P300,
which occurs when a person sees something they have been looking for, like a
missing set of keys. The system works by flashing images in front of a user,
who wears an electrode cap that picks up brain waves, and when the image of
the object the user is focusing on shows up, the P300 signal is recognized
and the robot is commanded to go to this object or pick it up. In order for
interaction between the user and the robot to occur, 10 minutes of
calibration exercise was needed for the computer to recognize a user's unique
P300. The computer then takes about five to 10 seconds to confirm the
specific image as the one the user is thinking of. The robot can only respond
to a small number of thought instructions, but does so with 94 percent
accuracy. However, Columbia University associate professor of biomedical
engineering Paul Sajda believes that while P300 response recognition is a
valuable area of research, as it could allow a better understanding of other
brain signals, it is not the best way to command robots. He says, "The
signals related to eye movements are 1,000 times stronger than scalp ECG,
you're better off using an eye tracker, which could be mounted on a pair of
glasses."  

Read the entire article at:
http://dsc.discovery.com/news/2007/01/10/thoughtrobot_tec.html?category=technology&guid=20070110120030&dcitc=w19-502-ak-0000

Links:
Rajesh P. N. Rao
http://www.cs.washington.edu/homes/rao/

Brain-Controlled Humanoid Robot
http://www.cs.washington.edu/homes/pshenoy/BrainControlledRobot.html

Researchers demonstrate direct brain control of humanoid robot
http://uwnews.washington.edu/ni/article.asp?articleID=28819

Neural Systems Group
http://neural.cs.washington.edu/

Paul Sajda
http://liinc.bme.columbia.edu/mainTemplate.htm?liinc_people_sajda.htm
