Machine Health Monitoring with Sensor Fusion Heat Maps
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing machine health monitoring systems rely on limited spatio-temporal information, making them less effective in detecting anomalies and requiring costly manual checks, which can be hazardous and disrupt production.
Innovation Solution
A system that integrates sensors such as microphones, cameras, radio transceivers, and inertial movement units to collect and process multi-layer spatial data, using machine learning to classify features and generate heat maps for automated anomaly detection and maintenance planning.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If manual health monitoring is used, then machine health can be monitored periodically, but it is costly in terms of labor and can be potentially hazardous
Solution Approach 1:
The machine monitoring system enables self-service by automatically detecting and diagnosing its own health status through multiple sensors and machine learning algorithms, eliminating the need for manual inspection while providing continuous reliability assessment
Solution Approach 2:
The patent replaces manual mechanical inspection with an automated electronic monitoring system that uses sensors, signal processing, and machine learning to detect machine health anomalies, thereby eliminating labor costs and safety hazards associated with manual checks
2Reliability
If manual health monitoring is used, then machine health can be monitored, but the machine may not be used for actual production during monitoring
Solution Approach 1:
The monitoring system operates continuously in the background without interrupting machine operation, allowing health assessment to occur while the machine remains productive. The system processes sensor data in real-time during normal operation, ensuring both continuous monitoring and uninterrupted production
3Device complexity
If limited spatio-temporal information is used, then monitoring system is simpler, but anomaly detection effectiveness is reduced
Solution Approach 1:
The patent merges multiple sensor types (acoustic, vibration, thermal, visual) to collect comprehensive spatio-temporal data from different modalities. This fusion of diverse sensor inputs creates a rich dataset that significantly improves anomaly detection precision while maintaining system manageability through integrated processing
4Productivity
If unexpected machine fault occurs, then production can be halted, but maintenance service on short notice is costly
Solution Approach 1:
The system performs preliminary detection and diagnosis of potential failures before they occur by analyzing sensor data patterns and predicting future machine states. This advance warning enables planned maintenance scheduling, avoiding both production halts and emergency maintenance costs
Data Source
AI summary
A system that includes one or more sensors installed in proximity to a machine configured to collect raw signals associated with an environment of the machine, are multi-layer spatial data that include time-stamp data. The system may include a processor in communication with the sensors and programmed to receive one or more raw signals, denoise the one or more raw signals to obtain a pre-processed signal, extract one or more features from the pre-processed signals, classify the one or more features to an associated class, wherein the associated class includes one or more of a normal class, abnormal class, or a potential-abnormal class, create fusion data by fusing the one or more features utilizing the associated class and the time-stamp data, and output a heat map on an overlaid image of the environment.