How you choose to interface with a computer is important, and the options vary wildly. You can use the always-efficient keyboard and mouse combo, track your body movement with a Kinect sensor, or simply speak your commands. The rising popularity of wearables is generating a need for tactile interfaces that can be subtly manipulated when you’re out and about. Serpentine takes a novel approach towards providing that by sensing how you squeeze, twist, and pull a cable.
Serpentine is a thin, flexible multi-material cable that’s ideal for wearables like necklaces and bracelets. The cable is made up of silicone embedded with coiled copper wire and conductive nylon thread. That unique construction makes it possible to measure changes in resistance as the cord is deformed. Because the resistance changes differently based on how the cable is deformed, Serpentine can register unique deformation “gestures.”
Classifying those resistance changes as a specific gestures, however, is a complex challenge. The analog nature of the setup means that the measurements are somewhat inconsistent, and depend on where you deform it and by how much. The solution was to use machine learning to train the system on the basic gestures. After training, Serpentine can recognize tapping, twisting, pressing, stretching, and sliding. Those basic gestures can then be combined for more complex input, yielding a wearable that could potentially be a practical way to interact with your computer or smartphone.