Multi-Sensor Fusion for Non-Stationary Machine Fault Analysis

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional machine monitoring and diagnostic techniques are poorly suited for analyzing non-stationary machines due to the time-varying nature of their operational parameters, leading to inaccuracies in extracting features related to machine condition and faults.

Innovation Solution

A system and method that utilizes multiple sensors to monitor non-stationary machines, synchronously sampling their outputs, and fuses these signals using wavelet transforms or machine learning to extract features insensitive to the machine's non-stationary operation, enabling accurate analysis through conventional signal processing techniques.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If conventional machine monitoring techniques are used on non-stationary machines, then the system complexity remains low, but the measurement precision and reliability of fault detection deteriorate due to time-varying operational parameters

Engineering Contradiction:
Improveaccuracy of feature extractionVSAvoidsystem complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent introduces signal processing intermediaries (wavelet transforms, envelope analysis, synchronous averaging) that act as mediators between the non-stationary machine signals and the analysis system. These intermediaries transform the time-varying signals into a form that can be accurately analyzed, resolving the contradiction by adding processing complexity rather than fundamental system complexity

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

The patent changes the parameters of the signal representation by transforming from time domain to frequency domain, and further to time-frequency domain using wavelets. This parameter transformation allows accurate feature extraction from non-stationary signals without requiring fundamentally complex monitoring infrastructure

Inventive Principle:
Principle #35Parameter changes

2Reliability

If multiple sensors are used to monitor non-stationary machines, then the measurement precision improves, but the device complexity increases due to signal synchronization and fusion requirements

Engineering Contradiction:
Improvefault detection accuracyVSAvoidsignal processing complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent segments the multi-sensor signal processing into distinct functional stages: synchronization of sensor outputs, wavelet transform of synchronized signals, extraction of instantaneous features, and fusion of features from multiple sensors. This segmentation manages complexity by making each stage independent and well-defined

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent adds the time-frequency dimension through wavelet transforms, moving analysis from单纯的 time domain to a two-dimensional time-frequency representation. This dimensional change allows simultaneous capture of transient features and frequency content from multiple sensors without linearly increasing processing complexity

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

3Measurement precision

If wavelet transforms and machine learning are used to fuse sensor signals, then the measurement precision and reliability improve, but the device complexity and computational requirements increase

Engineering Contradiction:
Improvefeature extraction accuracyVSAvoidsignal processing complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent performs preliminary signal processing (synchronization, wavelet transform, instantaneous feature extraction) before fusion with machine learning algorithms. This preliminary action prepares the data in an optimized format, reducing the computational burden on subsequent machine learning components and managing overall system complexity

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP3877819B1Automated analysis of non-stationary machine performance
Publication Date: 2026.01.28 AUGURY SYSTEMS LTD
  • EP3877819B1 patent drawingFigure 1
  • EP3877819B1 patent drawingFigure 2A
  • EP3877819B1 patent drawingFigure 2B

AI summary

A method for monitoring at least one machine including causing at least a first sensor to acquire at least a first non-stationary signal from at least one machine operating in a non-stationary manner during at least one operational time frame, the at least first sensor providing at least a first non-stationary output, causing at least a second sensor to acquire at least a second non-stationary signal from the at least one machine during the operational time frame, the at least second sensor providing at least a second non-stationary output, fusing the at least first non-stationary output with the at least second non-stationary output to produce a fused output, extracting at least one feature of at least one of the first and second non-stationary signals based on the fused output, analyzing the at least one feature to ascertain a state of health of the at least one machine and performing at least one of a repair operation, maintenance operation and modification of operating parameters of the at least one machine based on the state of health as found by the analyzing.