A biologically inspired tactile sensor array utilizing phase-based computation

We propose a biologically inspired tactile sensor array based on neural spike processing. Asynchronous data processing is integrated on-chip with the sensor array and converts the high-bandwidth raw input data into higher-level information with lower bandwidth requirements. Using phase-based computation primitives, as opposed to traditional rate-based neural spike codes, the array can compute the point-of-contact, force magnitude, force direction, and the presence or absence of slips, all in real time. Simulations demonstrate a reduction in the number of computations by up to three orders of a magnitude over a comparable synchronous approach.