In-Material Computation of High-Bandwidth Sensor Signals in Robotic Skin
Dana Hughes and Nikolaus Correll
Department of Computer Science
University of Colorado Boulder
Boulder, Colorado 80309–0430
Robotic skin possibly has the highest bandwidth requirements of any robotic sensor. Pressure signals must be sampled at high spatial resolution and texture recognition requires sampling in the kilohertz domain. Here, we summarize our recent works on pressure, gesture and texture sensing where information is processed within the skin itself. We present two tactile sensitive skins designed for full-body robotic applications—a texture sensitive skin capable of detecting, localizing and identifying textures rubbed against the skin, and a skin with joint force and proximity sensing capabilities for obstacle avoidance and gesture recognition.We describe construction of the skin, as well as architectural approaches to software, and provide experimental results for localization, gesture recognition, and collision avoidance.