AI Federated Data Layer for Digital Twin Query Integration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The process of creating a digital twin for product lifecycle management is time-consuming and costly, requiring skilled consultants, and is not affordable by smaller enterprises due to the high cost and time required for data review and integration across disparate data sources with varying formats and properties.
Innovation Solution
An AI-assisted system uses machine-learning models to identify, map, and recognize data types, relationships, and patterns across disparate data sources, creating a federated data model that enables seamless querying and integration without physically moving data, leveraging industry standards and crowd-sourced best practices.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional manual data review and integration methods are used, then data accuracy and coherence are improved, but time consumption and cost increase significantly
Solution Approach 1:
The patent replaces the mechanical manual review process with an AI-based automated system that uses machine learning models to identify, classify, and integrate data from disparate sources. The AI system automatically detects data types, relationships, and patterns without human intervention, thereby reducing time consumption while maintaining data coherence through intelligent algorithms.
Solution Approach 2:
The patent introduces a federated data layer as an intermediary between disparate data sources and the digital twin creation process. This layer standardizes data from multiple sources using industry standards (ISO 15926, ISA-95, ISA-88) before processing, enabling automated AI processing while ensuring data coherence through standardized intermediate representation.
2Reliability
If skilled consultants are engaged for digital twin creation, then data integration quality is improved, but cost increases making it unaffordable for smaller enterprises
Solution Approach 1:
The patent enables enterprises to create their own digital twins without engaging external consultants by providing an automated AI-driven platform. The system performs data review, classification, and integration automatically using machine learning models, allowing smaller enterprises to access digital twin technology at a fraction of the traditional cost while maintaining integration quality through algorithmic consistency.
Solution Approach 2:
The patent replaces expensive human consultant services with a cost-effective automated AI system. The AI models, once trained, can be repeatedly applied to multiple data sources and enterprises at minimal marginal cost, making the service economically viable for smaller enterprises while maintaining consistent integration quality.
3Ease of operation
If data is physically moved and integrated from disparate sources, then data accessibility is improved, but data movement cost and complexity increase
Solution Approach 1:
The patent introduces a federated data layer as an intermediary that standardizes data from disparate sources using industry standards before processing. This intermediary layer enables the AI system to access and process data from multiple sources without physically moving or centralizing the data, thereby maintaining data accessibility while reducing integration complexity through standardized interfaces.
Solution Approach 2:
The patent segments the data integration process into distinct layers: the federated data layer handles standardization and interface management, while the AI processing layer handles analysis and digital twin creation. This segmentation allows data to remain in its original locations (maintaining accessibility) while reducing overall integration complexity through divided responsibilities.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Artificial Intelligence-assisted building/execution of federated data layer for enterprise engineering: A system trains at least one machine-learning model to identify information about industrial assets from training data, then map the information to a federated data model. The system retrieves information about data from an application in an industrial asset. The at least one machine-learning model identifies types of the data, relationships between the data, and patterns of the data, from the information and based on data types, data relationships, and data patterns in the federated data model. The at least one machine-learning model maps the types of the data, the relationships between the data, and the patterns of the data to the federated data model. The system identifies knowledge about the types of the data, the relationships between the data, and/or the patterns of the data in the federated data model, in response to a query about data.