Not a member yet? Register for full benefits!

VR Interfaces: BrainPort


Overview of BrainPort
First created back in 2006, interfaces which use the tongue as a sensory pathway to the brain, are just becoming commercially viable as of this writing.

BrainPort as the device is known, uses the sensitive surface of the tongue to send pseudo-visual information along a series of electrode prongs arranged in an array. It seems a novel concept; replacing one sense with another, but it does work, quite well. Similar techniques using the sensitive forehead and heat-maps to 'see'. Still, this is the first time an attempt has gone commercial.

Other than the tongue interface component, BrainPort functions just like any other virtual sight interface currently around. A pair of cameras mounted into a blacked out HUD are used to record low resolution real-time visual data of where the person is looking. For this application, a high resolution just makes more work for the processor, and since it has to be converted into an array of electrodes, a simple 100 pixel x 100 pixel display is more than enough.

This is then sent to a processing unit the size of a mobile phone, via a cable. The unit performs spatial calculations, and converts the two display signals into a single image. It might seem counter-intuitive to use more than one camera at this point, but the extra stereoscopic data has given the processing unit a sense of depth to work with in reconstructing the scene.

The until now purely visual scene is re-encoded as a set of electrical pulses, with differences in frequency, amplitude and duration being used to separate out different objects in the person's view. It works a little like echo location, in that you cannot tell what something is in your tongue-vision, you can only tell how near it is, and roughly where it is relative to the user. As they turn their head and the camera tracks, so the pulses change to match, in real-time.

This ability is borne out of continued practice with the device. The sensations are so alien that it takes a good deal of training to be able to make sense of the data - it is an entirely new spatial sense. However, an experienced user can differentiate many multiples of these slight changes in amplitude, width and frequency of the pulses. This is helped in no small part, because the part that is placed in contact with the tongue, is an array of 400 electrodes, each transmitting a potentially different signal, and arranged in a 20 x 20 array.

Indeed, a few of the more long term clinical trial patients are able to perceive the colour of an object in the vibrator field, by subtle differences in the encoded signal. Ultimately, it may then be possible for a trained individual to see equal to a 'normal' individual, with a sufficiently complex electrode array.

BrainPort is initially launching in the USA. It is awaiting FDA approval, but if all goes well, the BrainPort should be available early 2010, at around 10,000 usd.

Untitled Document