Neuro-Inspired Scalable Serial Sensor Network by Sequential Signal Propagation

Abstract

Electronic skin (e-skin) technology with high sensor distribution and versatility is required to develop robotic systems that exhibit the characteristics of living organisms. However, the scaling up of e-skin results in, issues such as the need for increased wiring, long readout time, and high power consumption. This study introduces a new circuit for a serial sensor network inspired by the signal propagation of neurons. The network comprises many sensor nodes chained together, wherein each node comprises only four logic elements for fast submicrosecond readout and low power consumption (microwatts). Similar to neurons, each sensor node communicates only with the adjacent node, thereby avoiding an increase in the bus line load or a reduction in the communication bandwidth during scaling up. In addition, various sensor products can be installed in a general-purpose manner.