Monday, September 15, 2008


Time - Improvements in computing power and a better understanding of how the brain works have scientists busy hunting for the distinctive neural fingerprints that flash through a brain when a person is talking to himself. The Army's initial goal is to capture those brain waves with incredibly sophisticated software that then translates the waves into audible radio messages for other troops in the field. "It'd be radio without a microphone, " says Dr. Elmar Schmeisser, the Army neuroscientist overseeing the program. "Because soldiers are already trained to talk in clean, clear and formulaic ways, it would be a very small step to have them think that way.". . .

The five-year contract it awarded last month to a coalition of sceintists from the University of California at Irvine, Carnegie Mellon University, and the University of Maryland, seeks to "decode the activity in brain networks" so that a soldier could radio commands to one or many comrades by thinking of the message he wanted to relay and who should get it. Initially, the recipients would most likely hear transmissions rendered by a robotic voice via earphones. But scientists eventually hope to deliver a version in which commands are rendered in the speaker's voice and indicate the speaker's distance and direction from the listener.

"Having a soldier gain the ability to communicate without any overt movement would be invaluable both in the battlefield as well as in combat casualty care," the Army said in last year's contract solicitation. "It would provide a revolutionary technology for silent communication and orientation that is inherently immune to external environmental sound and light.". . .

Dr. Mike D'Zmura of UC-Irvine, the lead scientist on the project, says his task is akin to finding the right strands on a plate full of pasta. "You need to pick out the relevant pieces of spaghetti," he says, "and sometimes they have to be torn apart and re-attached to others." But with ever-increasing computing power the task can be done in real time, he says. Users also will have to be trained to think loudly. "


Post a Comment

<< Home