Sensor Image Monitoring for Early Fault Classification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems lack effective methods to support users and administrators in monitoring and analyzing complex technical systems for impending malfunctions or defects, requiring them to manually analyze extensive data after failures occur, which is complex and time-consuming.
Innovation Solution
Implement a method using sensors to cyclically measure physical variables, transforming data into graphical images, and utilize a self-learning component (SLU) to classify system status, allowing users to confirm or reject classifications, with the SLU adapting to user feedback for improved future analysis.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If comprehensive tables containing measured values and status information or extensive log files are provided to system administrators for analysis, then the information completeness is improved, but the complexity of analysis increases
Solution Approach 1:
The patent segments the complex data analysis task into multiple visualization layers: raw data display, pattern recognition results, and anomaly detection alerts. This allows administrators to access comprehensive information while avoiding overwhelming complexity through structured presentation.
Solution Approach 2:
The system introduces an intermediary AI component that processes raw measured values and log files, transforming them into interpretable visual patterns and anomaly detections. This intermediary layer preserves information completeness while reducing analysis complexity for administrators.
2Extent of automation
If machine learning or artificial intelligence solutions are used for automated monitoring and analysis, then the automation level is improved, but the understandability of automated reactions decreases
Solution Approach 1:
The system provides feedback loops where AI detection results are displayed to administrators for verification. Administrators can confirm or correct anomaly detections, and this feedback is used to improve future AI analysis. This maintains high automation while ensuring understandability through human-in-the-loop validation.
Solution Approach 2:
The visualization interface acts as an intermediary between the AI system and administrators, translating automated reactions into understandable visual forms. Anomaly detections are presented with contextual information about measured values and patterns, making automated decisions transparent and comprehensible.
3Loss of information
If extensive log files and measured values are provided for post-failure analysis, then the diagnostic information is improved, but the time required for analysis increases
Solution Approach 1:
The system performs preliminary analysis continuously during normal operation, detecting patterns and anomalies before failures occur. This preliminary action prepares diagnostic information in advance, so when failures happen, administrators receive pre-processed insights rather than raw data requiring extensive post-failure analysis.
Solution Approach 2:
The patent replaces manual mechanical analysis of log files with automated computational analysis using AI and visualization algorithms. This substitution dramatically reduces analysis time while preserving diagnostic information through intelligent pattern recognition and anomaly detection.
4Adaptability or versatility
If decentralized monitoring of distributed systems is used, then the system scalability is improved, but the holistic system view decreases
Solution Approach 1:
The system merges decentralized monitoring data from multiple system components into a unified visualization interface. Individual component statuses, measured values, and anomaly detections are combined to provide a holistic system view, maintaining scalability while preventing information loss through integrated presentation.
Data Source
Figure 1
Figure 2
AI summary
The invention relates to a method for assisting a person in system analysis and monitoring a technical system for impending malfunctions or technical defects. A plurality of sensors cyclically record measured values quantifying different physical quantities at one or more components of the system and feed them as a data set to an evaluation device. Each data set is transformed by the evaluation device into a measured value image in which each measured value is represented by a pixel cluster of a uniform color or gray value. This is transferred to a self-learning, initially trained unit (SLU) of the evaluation device and evaluated by it. If the SLU classifies the system status as critical based on the current measured value image, an alarm is triggered.The SLU classification can be categorized by the supported person, namely by activating a control element to the SLU, either by confirming or rejecting it. The category assigned to the SLU classification by the person can then be applied by the SLU to a new data set of measurements recorded in the meantime.