Automated Drawing Information Mapping for Semantic Digital Twins

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

The process of aligning drawing data from process plants to industrial standards, such as DEXPI, is manual and time-consuming, requiring domain expert involvement and lacking automated semantic enrichment.

Innovation Solution

A method and device for automatically extracting drawing information from process plant diagrams, using named entity recognition models to map this information to industrial standards, and generating instances conforming to these standards, thereby reducing manual effort and increasing efficiency.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If manual mapping of drawing information to industrial standards is performed, then mapping accuracy and semantic enrichment are improved, but time consumption and manual effort increase significantly

Engineering Contradiction:
Improvemapping accuracyVSAvoidtime consumption
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The system enables automatic self-mapping of drawing information to industrial standards through a named entity recognition model that processes extracted entities and autonomously assigns them to appropriate standard classes without requiring manual intervention from domain experts

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The manual mechanical process of expert review and classification is replaced with an automated computational system using named entity recognition models that perform semantic enrichment and mapping through algorithmic processing

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

2Productivity

If automated information extraction is implemented, then productivity is improved, but the complexity of the extraction and mapping system increases

Engineering Contradiction:
Improveextraction speedVSAvoidsystem complexity
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The complex mapping system is segmented into distinct functional modules: an extraction module for obtaining drawing information, a named entity recognition module for identifying and classifying entities, and a mapping module for assigning standard classes, making each component more manageable and maintainable

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

A named entity recognition model serves as an intermediary component between the extraction module and the mapping module, processing extracted entities through standardized classification layers and facilitating smooth data flow between system components

Inventive Principle:
Principle #24Intermediary (Mediator)

3Reliability

If domain expert involvement is required for semantic enrichment, then mapping reliability is improved, but the process becomes less scalable and more time-consuming

Engineering Contradiction:
Improvemapping reliabilityVSAvoidprocessing efficiency
Core Design Contradiction:
ReliabilityVSProductivity

Solution Approach 1:

The system performs autonomous semantic enrichment by using a pre-trained named entity recognition model that automatically identifies and classifies entities according to industrial standards without requiring domain experts to manually review and annotate each entity

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The named entity recognition model is pre-trained with knowledge of industrial standards and entity classifications, enabling it to perform reliable mapping operations independently before human review is needed, thus maintaining quality while improving efficiency

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP4553694A1Automatic mapping of extracted drawing information
Publication Date: 2025.05.14 SIEMENS AG
  • EP4553694A1 patent drawingFigure 1~2
  • EP4553694A1 patent drawingFigure 3~4
  • EP4553694A1 patent drawingFigure 5

AI summary

Semantic digital twins of process plants shall be created faster on the basis of piping and instrumentation diagrams. Therefore, there is provided a method of mapping drawing information representing a design of a process plant to an industrial standard including the steps of automatically extracting (13) the drawing information from a drawing (12) representing the process plant, automatically mapping (17) the drawing information to classes (25, 28) of the industrial standard by using a named entity recognition model and generating an instance of the class of the industrial standard the drawing information is mapped to.