talk-icon
Description
Rachel Miller1 Charles Dhong1 Darren Lipomi1

1, University of California, San Diego, San Diego, California, United States

We present a low-cost glove apparatus that can wirelessly transmit gestures (fingerspelling) using an array of hybrid, piezoresistive sensors developed from a carbon-fluoroelastomer which act as a test bed for haptic actuation. The mechanically compliant strain sensors are fabricated from a composite of carbon black nanoparticles in a fluoroelastomer matrix, supported by poly(dimethylsiloxane) and encapsulated in polyurethane. Sensors were resistant to repeated cycles of 5.5% strain (as experienced at the knuckles) for over 1000 cycles. The sensors were designed to be low-cost and easy to process, have an easily measureable change in resistance (ΔR) when bent on the knuckles, and to output a consistent electrical signal over many strain cycles without interfering with the movement of the glove. Integration of these low-cost, flexible strain sensors with Bluetooth and accelerometer allowed us to translate the twenty-six letters in American Sign Language via fingerspelling. This low-cost, robust and simple platform is an ideal testbed for methods of communicating reversing the line of communication in human-computer interaction and testing actuation of haptic devices. We showcase this platform from insights gathered from studies on human perception of “softness”, which is most closely related to compliance. We studied fine grades of compliance by modulating contact area and indentation depth independently by manipulating Young’s modulus and surface topography. By performing and referencing concurrent studies on human perceptions of “softness”, we can leverage the glove platform to actuate fine gradations in compliance by changing contact area and indentation depth independently.

Tags