Neural network learns the recorded signal patterns of gestures
Nasa scientists have succeeded in an astonishing experiment. A pilot succeeded for the first time to steer and land a passing airplane without operation of the knob and only with the decrease of the signals of muscle nerves. However, it was not a real aircraft, but only a simulated one.
The pilot wore a bracelet with 8 electrodes to record the nerve signals. The sensors recorded the signals that the muscle neurons produced when the pilot made the appropriate gestures of steering: "In the experiment, a pilot clenches his hand in empty air, makes movements, and thus generates nerve signals that are picked up by dry electrodes on his arm", explains Charles Jorgensen of the Ames Research Center The signals are analyzed and then sent through a computer, allowing the pilot to control the simulated aircraft."