KPI Root-Cause Identification Using Causal Process Variable Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Identifying the root-cause of Key Performance Indicator (KPI) deviations in industrial processes is challenging due to the complex interdependencies among numerous process variables and the dynamic, non-linear characteristics of these processes, making it difficult for traditional methods to accurately detect and diagnose the causal effects.
Innovation Solution
A system and method utilizing machine learning models to determine changed performance characteristics, cluster key process variables, perform correlation and causality analyses, and use a knowledge graph to identify the root-cause of KPI deviations by determining the impact and causal relations of substantial process variables.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional detect-and-diagnose methods are used to identify process variables causing KPI deviation, then domain expertise can manually detect faults, but the method is time-consuming and difficult to scale due to the large number of process variables and their intricate dependencies
Solution Approach 1:
The patent replaces manual mechanical analysis by domain experts with automated machine learning models and algorithms. The system uses ML models to analyze process data, identify deviations, and determine causal relationships automatically, eliminating the time-consuming manual detection and diagnosis process while maintaining or improving accuracy.
Solution Approach 2:
The patent introduces an intermediary automated analysis system that acts as a bridge between raw process data and root-cause identification. This intermediary system uses ML models, clustering algorithms, and causal inference methods to process the complex interdependencies among process variables, providing accurate and timely identification without requiring direct manual analysis of all variables.
2Reliability
If domain experts manually supervise and identify root-causes in complex industrial processes, then they can observe and maintain KPIs, but they may miss blind spots and causal effects due to the complexity and dynamic nature of the processes
Solution Approach 1:
The patent implements feedback mechanisms where the automated system continuously monitors process data, compares actual performance with expected performance, and identifies deviations. The system provides feedback loops that allow for continuous learning and improvement, ensuring that causal effects are systematically captured without missing blind spots that human experts might overlook.
Solution Approach 2:
The patent replaces human expert supervision with automated ML-based analysis systems that can process all process variables simultaneously without fatigue or oversight. These systems use causal inference algorithms to systematically identify causal relationships, eliminating the blind spots inherent in manual analysis while maintaining the reliability needed for industrial process management.
3Adaptability or versatility
If the number of process variables monitored is increased to cover all assets in the industrial process, then more comprehensive coverage is achieved, but the complexity of analyzing interdependencies and identifying root-causes increases significantly
Solution Approach 1:
The patent segments the large set of process variables into manageable groups or clusters based on their relationships and dependencies. By dividing the comprehensive set of variables into smaller subsets, the system can analyze interdependencies within each segment more efficiently while maintaining overall coverage, thus reducing analysis complexity without sacrificing comprehensiveness.
Solution Approach 2:
The patent extracts and focuses on the most relevant process variables that have the greatest impact on KPI deviations. By identifying and extracting only the critical variables from the large set of monitored parameters, the system reduces analysis complexity while maintaining comprehensive coverage of the factors that truly matter for root-cause identification.
Data Source
AI summary
The present disclosure discloses a system and a method for identifying root-cause in potential process variables causing KPI deviation in the industrial process. The system identifies the root-cause in the process variables causing the KPI deviation based on a knowledge graph, a causal effect, and a relation between the process variables. The disclosed system and method improve the overall performance of the industrial process and prevent future KPI deviations in the industrial process.