Neuromorphic MEMS Resonator for Energy-Efficient Nonlinear Sensing

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing smart sensor devices for IoT and non-linear systems lack the capability to efficiently produce control signals through local nonlinear transformations, which is crucial for neuromorphic computing, especially in compact and energy-efficient forms.

Innovation Solution

A micro-electro-mechanical system (MEMS) device with an inertial component and resonant components connected via a flexible link, where the resonant components are driven into high-frequency oscillations by an electrostatic pump, allowing for neuromorphic computing through mechanical non-linearity, with transduction units measuring oscillatory motion and deformation to generate output signals.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Use of energy by moving object

If smart sensor devices are made compact and energy efficient for IoT applications, then device size and energy consumption are reduced, but the capability to perform local nonlinear transformations for neuromorphic computing is compromised

Engineering Contradiction:
Improveenergy efficiencyVSAvoidnonlinear transformation capability
Core Design Contradiction:
Use of energy by moving objectVSAdaptability or versatility

Solution Approach 1:

The patent replaces traditional electronic computing systems with a mechanical MEMS-based neuromorphic system. The inertial component coupled to resonant components creates mechanical nonlinear oscillations that perform computations, substituting electronic processors with mechanical dynamics to achieve energy-efficient neuromorphic computing

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The system changes operational parameters by driving resonant components at different frequencies and amplitudes to achieve various nonlinear transformation behaviors. By adjusting the drive frequency relative to resonant frequencies and modifying coupling strength, the device can adapt its computational characteristics for different tasks

Inventive Principle:
Principle #35Parameter changes

2Productivity

If traditional electronic systems are used for neuromorphic computing, then computational capability is achieved, but device complexity and energy consumption increase

Engineering Contradiction:
Improvecomputational capabilityVSAvoidsystem complexity
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The MEMS device performs computations through its own inherent mechanical dynamics without requiring external electronic processors. The coupled inertial and resonant components naturally generate complex nonlinear oscillations that encode computational results, allowing the system to compute through its physical behavior rather than through separate computing hardware

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The single MEMS device with coupled inertial and resonant components can perform multiple computational tasks by adjusting drive parameters. The same physical structure can be configured for different neuromorphic computations by changing frequency, amplitude, and coupling conditions, providing universal computational capability

Inventive Principle:
Principle #6Universality (Multi-functionality)

3Productivity

If resonant components are driven into high-frequency oscillations, then neuromorphic computing performance is improved, but sensitivity to external stimuli decreases

Engineering Contradiction:
Improveneuromorphic computing performanceVSAvoidsensitivity to external stimuli
Core Design Contradiction:
ProductivityVSObject-affected harmful factors

Solution Approach 1:

The system exploits mechanical vibration of resonant components at high frequencies to generate complex oscillatory patterns for neuromorphic computing. The vibration-based computation occurs in a frequency range where the system is less susceptible to low-frequency environmental disturbances, naturally filtering out noise while maintaining computational performance

Inventive Principle:
Principle #18Mechanical vibration

Applied Scientific Principles

This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.

Function Achieved in This Case

The MEMS device effectively implements neuromorphic computing by linearly combining measured characteristics, achieving efficient and compact neuromorphic signal production with minimal sensitivity to external stimuli, enabling robust control signal generation for IoT and non-linear systems.

Implementation Method 1

the resonant components are driven into high-frequency oscillations by an electrostatic pump

Methodology Applied
Scientific EffectElectrostatic force: Electrostatics

Implementation Method 2

an inertial component and resonant components connected via a flexible link

Methodology Applied
Scientific EffectElastic deformation: Elasticity

Data Source

PatentEP3898502B1Neuromorphic micro-electro-mechanical-system device
Publication Date: 2024.07.31 SCOPRA SCI & GENIE SEC
  • EP3898502B1 patent drawingFigure 1
  • EP3898502B1 patent drawingFigure 2
  • EP3898502B1 patent drawingFigure 3

AI summary

A micro-electro-mechanical-system (MEMS) device comprises an inertial component configured for being connected to a structure by a flexible connection allowing the inertial component to deform or move relative to the structure in response to an external stimulus applied to the structure. One or more resonant components are connected to the structure or inertial component, the resonant component(s) having resonant mode(s). Transduction unit(s) measures an oscillatory motion of the resonant component relative to the inertial component and/or structure. An electronic control unit applies a pump of electrostatic force to induce an oscillatory motion of the resonant component(s) in the resonant mode, the oscillatory motion being a non-linear function of a strength of the electrostatic force. The resonant component is configured to be coupled to the inertial component and/or the structure such that a deformation and/or motion of the inertial component in response to an external stimulus changes the strength of the pump, the electronic control unit configured for producing and outputting an output signal being a mathematical function of the measured oscillatory motion. A system for producing a neuromorphic output for a MEMS device exposed to external stimuli is also provided.