New software tries to expand the "touch" in "touchscreen."
Touchscreens were a pivotal innovation for computing: Much like the graphical user interface, much like the trackpad, much like the mouse, they gave us a brand-new way to put human bodies to the service of human technology. But the touchscreens we have today are also -- just like graphical interfaces and trackpads and mice once were -- a first step toward something more intricate and sophisticated and better. Mice were, in short order, improved with scrolling and right-click capabilities. Trackpads got adapted to distinguish between the gestures of one finger, or two, or three. Graphical interfaces became more intuitive and more attractive.
And now touchscreens are approaching the moment that all good innovations do: the end of the beginning. Computer scientists are developing software that expands the capability of the touchscreen by expanding the idea of touch itself.
So researchers at Microsoft are working on 3-D gestural control that would transfer users' hand gestures to a screen. Scientists at Purdue are experimenting with haptic feedback (pdf) that approximates the feel of discrete buttons -- even on a piece of flat glass. There's the bendable touchscreen. There's the capacitive fingerprinting that allows screens to distinguish among different users. And over at Carnegie Mellon, the computer scientist Chris Harrison has developed software that is able to distinguish between touches delivered from the fingertip, the knuckle, and even (ew?) the fingernail. The whole of one hand, communicating with the machine held by the other.
More on the workings of Harrison's software, FingerSense, in the video above. But the most intriguing operating feature is that the system works not so much by detecting touch as by detecting sound. In Harrison's prototype, a modified Samsung Galaxy S3 smartphone, the device is fitted with a vibration sensor that listens for the acoustic and vibrational differences that occur when a screen is swiped by a fingerpad or a fingernail. So a fingertip could select an object, say, while a knuckle tap could do the kind of next-step and deep-cuts work that a right-click does on a computer. One gesture could get a user deep into his or her own device. And one appendage could operate even more effectively as a medium between mind and machine.