Automated Diagnosis System for Root Cause Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current system maintenance approaches, whether manual or automatic, face inefficiencies in diagnosing and resolving system problems due to reliance on initial symptom information, which often fails to determine the exact root cause, leading to waste in manpower, resources, and time.
Innovation Solution
A monitoring apparatus and diagnosis apparatus system that collects and analyzes symptom information, with the diagnosis apparatus interacting with the monitoring apparatus to gather additional information when necessary, to enhance the accuracy and efficiency of problem diagnosis and recovery.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If automatic diagnosis is performed based on initial symptom information, then diagnosis speed is improved, but diagnosis accuracy deteriorates because initial symptoms may not suffice to determine the real root cause
Solution Approach 1:
The system performs preliminary diagnosis actions using initial symptom information to quickly identify potential root causes. The diagnosis apparatus initially processes available symptoms to generate preliminary diagnosis results, which then trigger subsequent information collection actions to refine the diagnosis accuracy without sacrificing initial response speed
Solution Approach 2:
The system implements a feedback mechanism where the diagnosis apparatus evaluates the confidence level of preliminary diagnosis results. When confidence is insufficient, the system automatically requests additional symptom information from the monitoring apparatus, creating a closed-loop feedback process that iteratively improves diagnosis accuracy while maintaining efficient initial response
2Measurement precision
If manual approach is used with professional technicians, then diagnosis accuracy is improved, but time consumption and cost increase significantly
Solution Approach 1:
The system enables self-service diagnosis by automatically collecting symptom information, analyzing it to determine root causes, and generating recovery packages without requiring professional technicians. The monitoring apparatus autonomously performs information collection and the diagnosis apparatus autonomously performs analysis, eliminating the need for human intervention in routine diagnosis tasks
Solution Approach 2:
The system replaces the mechanical system of manual technician intervention with an automated information processing system. The diagnosis apparatus uses knowledge repositories and automated reasoning to substitute human expert analysis, while the monitoring apparatus uses automated agents to substitute manual information gathering, thereby reducing both time consumption and labor costs
3Difficulty of detecting and measuring
If manual approach is used with professional technicians, then comprehensive problem analysis is improved, but resource waste increases due to 60% time spent on problem identification and 95% of problems being previously encountered
Solution Approach 1:
The monitoring apparatus continuously and preliminarily collects symptom information and system state data before problems occur or as they emerge. This preliminary information gathering eliminates the need for technicians to spend time on initial problem identification, as the system has already accumulated relevant data automatically
Solution Approach 2:
The system creates a digital copy of the problem diagnosis process by storing symptom patterns, root causes, and recovery solutions in knowledge repositories. When similar problems occur, the system copies and applies previously stored diagnostic patterns and solutions, eliminating redundant analysis work and reducing resource consumption
4Speed
If knowledge repository is queried based on initial symptoms, then quick response is achieved, but root cause determination accuracy deteriorates due to similar symptoms of different problems
Solution Approach 1:
The system dynamically adjusts the diagnosis process based on confidence levels. Initially, the system quickly queries the knowledge repository using available symptoms. When the confidence level of the result is insufficient, the system dynamically transitions to a second phase, automatically collecting additional symptom information to refine the root cause determination, thus adapting the diagnosis depth to the specific problem complexity
Data Source
AI summary
Embodiments of the present invention relate to method and apparatus for system problem diagnosis and recovery. According to embodiments of the present invention, problem symptom information in a system can be automatically monitored and collected BY a monitoring apparatus (or referred as to “agent”) deployed at the system side. Upon after receiving such information, the diagnosis apparatus, for example, may automatically determine a root cause of the problem by querying a backend knowledge repository, and possibly generate an executable software package for recovering the problem. If the diagnosis apparatus determines that the currently available information is insufficient to determine a creditable enough root cause and/or is insufficient to generate the software package for recovering the problem, the diagnosis apparatus may interactively control the monitoring apparatus to collect desired additional information. In this way, the efficiency and accuracy of problem diagnosis and recovery may be improved.