Enriching standards-based digital thread by fusing as-designed and as-inspected data using knowledge graphs

Area of Science:

• Manufacturing Engineering
• Data Science
• Ontology Engineering

Background:

• The digital thread is crucial for smart manufacturing, requiring linked data across the product lifecycle.
• Current methods struggle to unify design and inspection data, hindering automated quality assurance.
• Ontology models and knowledge graphs have shown success in integrating engineering data.

Purpose of the Study:

• To develop a standards-based digital thread by fusing as-designed (STEP) and as-inspected (QIF) data.
• To address the lack of unified information models for automated product quality assurance.
• To leverage ontology and knowledge graphs for integrating heterogeneous lifecycle data.

Main Methods:

• Developed an automated pipeline for generating knowledge graphs from STEP and QIF data.
• Implemented a mapping strategy to integrate separate STEP and QIF knowledge graphs.
• Utilized rules and queries to demonstrate the integrated data's potential for decision-making.

Main Results:

• Successfully generated knowledge graphs representing both STEP and QIF data.
• Established a method for mapping and fusing these knowledge graphs into a cohesive digital thread.
• Demonstrated the potential for enhanced product quality assurance through integrated data analysis.

Conclusions:

• The proposed approach effectively fuses STEP and QIF data within a knowledge graph-based digital thread.
• This integration provides a foundation for more automated and informed product quality assurance in smart manufacturing.
• The use of ontology and knowledge graphs facilitates data associativity and traceability throughout the product lifecycle.