UML Architecture Modeling via XMI for Automated AI Workflows

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Current generative AI-based tools generate UML diagrams in an image format that lack model elements and metadata, preventing automation processes and requiring manual updates for changes, leading to increased effort and costs in architecture modeling.

Innovation Solution

An architecture modeling assistant utilizing generative AI and robotic process automation generates XMI representations of UML models, including model elements and relationships, and automates tasks like validation and alignment, enabling seamless integration with architecture modeling tools.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If generative AI-based tools generate UML diagrams in image format, then generation speed is improved, but model elements and metadata are lost

Engineering Contradiction:

Improvegeneration speedVSAvoidmodel elements and metadata

Core Design Contradiction:

ProductivityVSLoss of information

Solution Approach 1:

The patent uses XMI (XML Metadata Interchange) format to create a structured copy of UML model data that preserves both the visual diagram information and the underlying model elements and metadata. This allows the system to generate diagrams quickly while maintaining complete information about the modeled components and their relationships.

Inventive Principle:

Principle #26Copying

Solution Approach 2:

The patent changes the output format parameter from traditional image formats to XMI/XML format, which enables the system to represent UML models with both visual and structural information. This parameter change allows automated processing while preserving all necessary model elements and metadata.

Inventive Principle:

Principle #35Parameter changes

2Shape

If UML diagrams are generated as images, then visual representation is improved, but automation processes cannot be performed

Engineering Contradiction:

Improvevisual representationVSAvoidautomation processes

Core Design Contradiction:

ShapeVSExtent of automation

Solution Approach 1:

The patent introduces XMI as an intermediary format between the visual UML diagram representation and automated processing systems. This intermediary structure allows the system to maintain visual representational capabilities while enabling automated tasks such as validation, transformation, and integration with other tools through standardized XML parsing.

Inventive Principle:

Principle #24Intermediary (Mediator)

3Adaptability or versatility

If manual updates are required for changes, then flexibility is improved, but effort and costs increase

Engineering Contradiction:

ImproveflexibilityVSAvoidmanual effort

Core Design Contradiction:

Adaptability or versatilityVSLoss of time

Solution Approach 1:

The patent implements automated feedback mechanisms where changes to the XMI representation can be automatically detected and propagated through the model. This allows the system to maintain flexibility in accommodating changes while automatically updating related elements, reducing manual intervention and time investment for architects.

Inventive Principle:

Principle #23Feedback

Data Source

PatentUS20250378340A1Generating Architecture Solutions

Publication Date: 2025.12.11 INTERNATIONAL BUSINESS MACHINE CORPORATION

AI summary

Generating architecture solutions is provided. An Extensible Markup Language Metadata Interchange (XMI) representation of a Unified Modeling Language (UML) model corresponding to an architecture solution is generated based on a set of instructions in a predefined format. The UML model corresponding to the architecture solution is generated based on the XMI representation of the UML model.