Predicting Malfunction Recurrence via Log Anomaly Analysis

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Complexity and increased data processing in digital systems make error diagnosis and prevention difficult due to sporadic malfunctions, which can have far-reaching consequences.

Innovation Solution

A method using logged log data to identify anomalies in operating parameters by analyzing probability differences within specific time intervals, predicting imminent malfunctions through monitoring these parameters, and initiating countermeasures.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If digitalization, automation, and networking are increased to improve system functionality and data processing capability, then system performance and productivity are improved, but system complexity increases and the system becomes more prone to errors and malfunctions

Engineering Contradiction:
Improvedata processing capabilityVSAvoidsystem complexity
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The system performs preliminary anomaly analysis on log data to identify combinations of operating parameter values that precede malfunctions. By detecting these anomalies in advance and monitoring them continuously, the system can predict impending malfunctions before they occur, allowing preventive actions to be taken. This resolves the contradiction by adding predictive capability that increases with system complexity.

Inventive Principle:
Principle #10Preliminary action

2Measurement precision

If comprehensive log data is collected and analyzed to improve malfunction detection accuracy, then measurement precision and reliability are improved, but device complexity and data processing requirements increase

Engineering Contradiction:
Improvemalfunction detection accuracyVSAvoiddata processing complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The system extracts only the relevant information needed for anomaly detection - specifically, combinations of operating parameter values that show significant probability differences between normal and malfunction periods. Instead of analyzing all log data comprehensively, the system focuses on extracting these specific anomaly patterns, thereby improving detection accuracy while managing data processing complexity.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The system changes the approach from analyzing individual parameter values to analyzing combinations of parameter values. By considering joint probability distributions of multiple operating parameters simultaneously, the system can detect subtle anomalies that individual parameter analysis would miss, thereby improving detection precision without proportionally increasing complexity.

Inventive Principle:
Principle #35Parameter changes

3Reliability

If anomaly analysis is performed on multiple operating parameters to improve prediction accuracy, then reliability is improved, but loss of time for data processing and analysis increases

Engineering Contradiction:
Improvemalfunction prediction accuracyVSAvoiddata processing time
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The system performs anomaly analysis on a selected subset of operating parameters that are most relevant to the specific malfunction type being predicted. Instead of analyzing all possible operating parameters, the system focuses on the critical subset that shows significant probability differences, thereby maintaining high prediction accuracy while reducing the time required for data processing.

Inventive Principle:
Principle #16Partial or excessive action

4Reliability

If continuous monitoring of operating parameters is implemented to predict impending malfunctions, then reliability and early detection capability are improved, but device complexity and computational requirements increase

Engineering Contradiction:
Improvesystem availabilityVSAvoidmonitoring system complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The system uses the existing log data infrastructure and operating parameter collection mechanisms already in place, rather than requiring separate dedicated monitoring hardware or systems. By leveraging existing system resources and data collection capabilities, the anomaly detection and prediction functionality is added with minimal additional complexity, maintaining high reliability while avoiding excessive system complexity.

Inventive Principle:
Principle #25Self-service

Data Source

PatentEP4350517A1Predicting a recurrence of a malfunction
Publication Date: 2024.04.10 BUNDESDRUCKEREI GMBH
  • EP4350517A1 patent drawingFigure 1
  • EP4350517A1 patent drawingFigure 2
  • EP4350517A1 patent drawingFigure 3

AI summary

The invention relates to a method for predicting a repeated occurrence of a malfunction (110) of a computer system (100, 130, 160, 198) using logged log data (122, 152, 182, 196) of the computer system (100, 130, 160, 198). The procedure comprises logging data (122, 152, 182, 196), performing an anomaly analysis for the occurrence of a malfunction (110) of the computer system (100, 130, 160, 198), identifying operating parameters (112) whose values ​​are encompassed by one or more specific anomalies, and monitoring the values ​​recorded for the identified operating parameters (112) in the log data records, wherein the monitoring includes predicting an impending recurrence of the malfunction (110).