A Neuromorphic Electrothermal Processor for Near‐Sensor Computing

Abstract

The statistical processing of sensor data using conventional digital computers is inefficient in terms of time, energy usage, and communication bandwidth, among others. Therefore, new approaches are sought to create context and make sense of the sensor data using special‐purpose computers that excel in specific computation tasks. Herein, the requirements for physical systems to perform sophisticated nonlinear computations needed for real‐time pattern recognition in data, specifically sensor data, are discussed. The focus is on physical reservoir computing as a neuromorphic computing approach. Considering energy flow as the coupling mechanism between nonlinear dynamic systems, it is demonstrated that many physical systems satisfy the basic requirements for building reservoir computers. Using physical reservoir computers brings up exciting opportunities for near‐ or in‐sensor computing as to how new data are collected and processed. The concepts are demonstrated through a novel physical computation platform, where off‐the‐shelf, temperature‐sensitive resistors are used to perform various standard and specific computational tasks. This platform is used as a near‐sensor processor to detect particular events. How a similar platform may be used for in‐sensor neuromorphic computations is further discussed.

A 3D‐Printed Computer

“…Researchers recently demonstrated an RC that was built using off‐the‐shelf thermistors that provided the nonlinear response and fading memory requirements while the coupling between the elements was achieved by directional sharing of the electrical current. [ 26,27 ]…”

“…Researchers recently demonstrated an RC that was built using off-the-shelf thermistors that provided the nonlinear response and fading memory requirements while the coupling between the elements was achieved by directional sharing of the electrical current. [26,27] 3D-printed resistors exhibit interesting time-dependent, nonlinear responses. These resistors can be made, for instance, from a conductive material using fused deposition modeling (FDM) technology.…”

Emerging Iontronic Neural Devices for Neuromorphic Sensory Computing

“…To date, a series of neuromorphic sensing devices and systems have been reported with promising biomimetic sensing and signal processing functionalities. [12,15,101,102,[188][189][190][191][192][193][194][195] Among them, neuromorphic sensory computing based on iontronic neural devices has received extensive attention (Figure 3). This section summarizes recent developments in low-level neuromorphic sensing based on iontronic neural devices.…”

Near‐Sensor Reservoir Computing for Gait Recognition via a Multi‐Gate Electrolyte‐Gated Transistor

“…[ 25,27,28 ] The reservoir‐based signal processing is high speed, low cost, and energy efficient, therefore, serves as an ideal candidate to be integrated with the sensors for the implementation of near‐sensor computing paradigm. [ 27,29–31 ]…”

“…[25,27,28] The reservoir-based signal processing is high speed, low cost, and energy efficient, therefore, serves as an ideal candidate to be integrated with the sensors for the implementation of near-sensor computing paradigm. [27,[29][30][31] Various emerging nanodevices represented by memristors have been utilized for physical reservoirs. [29,32,33] By leveraging the internal ionic dynamics that offer nonlinearity and short-term memory effects, these reservoir devices have been successfully used for different task implementations including hand-written letter recognition, spoken-digit prediction, neural spike analysis, language learning, breast tumors classification, etc.…”