Graph Model for Machinery Health Monitoring
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems for creating connected models for diagnostics and prognostics are inefficient, requiring significant manual effort, being slow, and not scalable, as they rely on structured databases that struggle with complex relationships and data from multiple sources.
Innovation Solution
A system that generates a comprehensive connected graph model from multiple input sources using a graph database, allowing for automatic creation and updating, and enabling efficient querying and diagnostics by representing data as nodes and edges, rather than relying on predefined schemas.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Stability of the object's composition
If a structured database with predefined schemas is used to store system data, then data organization is standardized and manual mapping is straightforward, but querying complex relationships becomes slow and system crashes occur with multiple simultaneous queries
Solution Approach 1:
The patent replaces the traditional structured database (relational model with tables and joins) with a graph database model. This substitution enables efficient traversal of complex relationships through native graph operations, achieving orders of magnitude improvement in query performance while maintaining data organization stability through the graph schema.
2Stability of the object's composition
If manual mapping of parsed data to predefined tables is performed, then data placement follows a fixed schema, but the amount of manual effort required limits automation and scalability
Solution Approach 1:
The patent implements automatic data mapping capabilities where the system self-configures the graph schema and automatically maps incoming data from various sources without requiring manual intervention. The graph database's flexible schema allows automatic inference of relationships and data placement, enabling full automation of the data ingestion process.
3Reliability
If a comprehensive connected model is created from multiple input sources, then diagnostics and prognostics capabilities are enhanced, but the model becomes difficult to maintain and update manually
Solution Approach 1:
The patent creates a dynamic graph model that automatically adapts to changes in the system architecture and data sources. The graph structure dynamically incorporates new components and relationships as they are added to the monitored system, eliminating the need for manual model updates and maintaining comprehensive connectivity automatically.
4Measurement precision
If detailed engineering documentation is used to create connected models, then diagnostic precision is improved, but the documentation does not necessarily connect and requires system engineering insights to resolve issues
Solution Approach 1:
The patent creates a universal graph model that can ingest and integrate data from multiple different documentation formats and sources simultaneously. The graph database serves as a common framework that universally represents relationships across different engineering domains and documentation types, automatically connecting previously siloed information sources.
Data Source
AI summary
Systems and methods are described to create a comprehensive model from multiple input data sources in distributed manner for performing diagnostics and/or prognostics of a complex system/platform having its modules/parts implemented independent of each other. In an embodiment, the proposed system for creating a comprehensive connected model of a complex platform includes a input data receive module configured to receive data/content/information from one or more input data sources associated with different modules of the complex platform in a distributed manner, a dependency determination module configured to analyze a plurality of entities/variables retrieved from the data and determine dependency relationships between the plurality of entities by creating a edge list, and a connected model creation module configured to create a comprehensive connected model (connected graph) from the edge list.