Causal Graph Extraction From PLC Code for Faster Root Cause Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Identifying causal relationships in industrial automation systems is challenging due to confounding variables, complex causal relationships, lack of experimental control, complexity, non-linearity, and hidden variables, which complicates determining the root cause of anomalies such as security breaches and unauthorized control modifications.
Innovation Solution
A system that extracts causal graph relationships directly from design artifacts like PLC code and P&ID digital designs, generating dynamic and real-time causal graphs to provide a custom view of the industrial automation system, enabling root cause analysis and detection of security breaches.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If historical data is used to determine probable causes of anomalies, then analysis can be performed without certainty, but multiple remedial actions must be tried before resolving the issue
Solution Approach 1:
The system performs preliminary parsing of design artifacts during the design phase to extract causal relationships before anomalies occur. This pre-processing creates a ready-to-use causal model that enables immediate root cause identification when anomalies are detected, eliminating the need for multiple remedial action attempts and reducing resolution time while maintaining high certainty.
Solution Approach 2:
The system creates a causal graph copy of the design artifact that mirrors the system's causal structure. This graphical representation allows operators to visually trace causal relationships and identify root causes without needing to interpret complex historical data, providing certain identification while reducing analysis time.
2Measurement precision
If design artifacts are parsed to extract causal relationships, then certain root cause identification is achieved, but complex processing of design artifacts is required
Solution Approach 1:
The system segments the design artifact parsing process into distinct modules: parsing the design artifact to identify components, extracting causal relationships between components, and generating a causal graph representation. This segmentation simplifies the overall processing complexity by breaking down the complex task into manageable, independent steps that can be executed sequentially.
Solution Approach 2:
The system introduces a causal graph as an intermediary representation between the design artifact and the anomaly analysis process. This graphical intermediate form simplifies the complex causal relationships into visual nodes and edges, making it easier to process and query without requiring direct interpretation of the complex design artifact structure.
3Productivity
If causal relationships are extracted from design artifacts, then root cause analysis capability is improved, but additional processing steps are added to the system
Solution Approach 1:
The system performs causal relationship extraction as a preliminary action during the design phase, creating a ready-to-use causal graph before operational anomalies occur. This advance preparation eliminates the need for complex real-time processing during anomaly resolution, improving productivity while the additional processing steps are confined to the initial setup phase.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A system for providing causal graph relationships for industrial automation systems. The system receives a first input indicative of a selection of a design artifact associated with an industrial automation system having one or more industrial automation devices, where the design artifact defines an industrial automation project configured to control operations of the one or more industrial automation devices. In response to the first input, the system parses the design artifact to identify causal relationships between the one or more industrial automation devices, stores data indicative of the identified causal relationships in a database, and generates causal graphs based on the stored data. The system receives a second input indicative of a request to generate a custom view of the industrial automation system, and generates a custom view of the industrial automation system in response to the second input, where the custom view includes the causal graphs.