Diagnostic Message Pattern Analysis for Component Root Cause Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for analyzing diagnostic messages are time-consuming and inefficient, often requiring complex documentation or tracing software code, and are not effective for identifying the cause of diagnostic reports across various technical systems.
Innovation Solution
A method that compares each status description in diagnostic messages with a predefinable reference to determine if components meet specified states, calculating a frequency for each component to identify those exceeding a threshold, thereby classifying them as the cause of diagnostic messages, independent of documentation availability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If documentation is used to analyze diagnostic messages, then analysis accuracy is improved, but analysis time increases significantly
Solution Approach 1:
The patent pre-processes and stores diagnostic message patterns and their associated causes in a database before actual diagnostic needs arise. This preliminary action allows the system to quickly retrieve and match patterns during runtime without requiring time-consuming manual analysis of documentation, thus improving analysis speed while maintaining accuracy through pre-validated pattern matching
Solution Approach 2:
The patent creates simplified copies of complex diagnostic documentation in the form of structured pattern templates. These templates capture essential diagnostic patterns and causes in a standardized format that can be quickly queried and matched, replacing the need to review original complex documentation while preserving the essential diagnostic information
2Measurement precision
If software program code is traced to analyze diagnostic messages, then root cause identification is improved, but analysis time increases significantly
Solution Approach 1:
The patent extracts essential diagnostic patterns and cause relationships from software program code during a preliminary analysis phase. By separating the extraction of diagnostic logic from the actual diagnostic execution, the system stores these extracted patterns in a searchable database, allowing rapid root cause identification without requiring time-consuming code tracing during operational diagnostics
Solution Approach 2:
The patent performs preliminary extraction and structuring of diagnostic logic from software code before actual diagnostic needs arise. This advance preparation creates a ready-to-query pattern database that enables rapid root cause identification during runtime without requiring actual code tracing, thus maintaining accuracy while dramatically reducing analysis time
3Loss of information
If complex documentation is reviewed to analyze diagnostic messages, then comprehensive analysis is improved, but time consumption increases
Solution Approach 1:
The patent segments complex diagnostic documentation into discrete, structured pattern templates that can be independently queried and matched. Each template represents a specific diagnostic pattern with its associated causes, allowing the system to retrieve only relevant segments rather than reviewing entire documentation sets, thus maintaining comprehensive coverage while reducing time consumption through targeted pattern matching
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The invention relates to a method for analyzing multiple diagnostic messages (d) of a technical system with multiple components, wherein each diagnostic message (d) describes a respective state of the technical system using multiple state descriptions (z) for the components, and wherein diagnostic messages (d) of the same type (a) each comprise identical state descriptions (z). To achieve an improved method, it is proposed that for each of the diagnostic messages (d), each state description (z) is compared with a predefinable reference (r) (4), and using the comparison (4), it is determined whether the respective component meets a respective predefinable state or not. It is further proposed that for diagnostic messages (d) of the same type (a), a frequency (6) is determined for each component, which indicates how frequently the respective component does not meet the respective predefinite state.Furthermore, it is proposed that the state description(s) (z) of the component(s) whose respective frequency (6) exceeds a specified threshold be classified as the cause of the diagnostic messages (d) of the respective type (a).