Automating Model Based Systems Engineering with AI

Model Based Systems Engineering (MBSE) offers a robust approach to designing complex systems, but often involves significant manual effort and specialized expertise. The increasing complexity of engineering projects makes this challenge even more acute. Imagine having complete visibility into your systems architecture, with the ability to automatically generate and validate engineering models.

Today we'll discuss a new AI-driven approach that automates the generation and validation of engineering models. Through the use of AI, organizations can now model their systems in detail and automatically verify design consistency. This allows them to identify inconsistencies before they become problems and implement targeted corrections. The solution enables organizations to improve collaboration across disciplines while making data-driven decisions about systems architecture. This solution leverages AI to deliver insights into system design, enabling both efficiency improvements and more robust traceability.

Understanding AI-driven MBSE

Traditional systems engineering approaches often struggle with the complexity of modern engineering projects. The manual creation of Systems Modeling Language (SysML) models can be a daunting task, potentially requiring multiple man-years of effort. Furthermore, ensuring traceability, consistency, and adherence to SysML standards throughout the model is a complex and time-consuming process.

An AI-driven MBSE software offers a powerful paradigm to automate model generation and validation. By combining generative AI models, graph algorithms, and AI-powered verification, the software can automatically generate SysML graphs from technical documentation, stakeholder meeting notes, and regulatory documents. This approach captures intricate relationships and potential inconsistencies often missed by manual methods.

As engineering projects become more complex, distributed, and impacted by interconnected requirements, the need for automation and intelligent tools increases. In uncertain, complex environments, AI-driven MBSE provides an intuitive and efficient technique to accelerate model creation and ensure design consistency. By explicitly modeling granular system elements and their interactions, AI reveals how localized inconsistencies can cascade through the model. This granularity in simulating multi-agent adaptive behaviors is pivotal for enhancing design robustness and accelerating development.

The unique advantages of AI for SysML model generation and verification

The strengths of AI-driven MBSE hinge on its ability to encapsulate the complexity and dynamism of engineering projects. Unlike traditional frameworks, AI thrives on complexity and interdependencies. It uniquely facilitates the automated generation of SysML models and the exploration of potential inconsistencies. This characteristic is especially valuable in understanding how minor discrepancies can cascade through a system design, potentially unveiling vulnerabilities and inefficiencies not easily captured under more aggregate analysis methods.

AI-driven MBSE excels in:

• Automated Model Generation with Traceability: Automatically generating SysML graphs from technical documentation and other sources.
• AI-Powered Verification: Validating design consistency against technical documentation and SysML standards.
• AI CoPilot: Automating engineers' interactions with systems architecture modeling software.

This enables assessing system design resilience in a controlled, virtual environment under various conditions. By enabling a detailed examination of the ripple effects from such discrepancies, as well as facilitating model exploration, AI empowers decision-makers with the insights needed to focus their efforts and fortify system designs against future issues while optimizing performance under normal operations.

Moreover, the adaptability of AI-driven MBSE to incorporate new data and adapt to changing project requirements stands as a significant advantage, ensuring analyses remain timely, relevant and actionable.

Key Components of the AI-Driven MBSE System

AI Graph Agent for SysML Model Generation

SysML v2 models are represented as Directed Acyclic Graphs (DAGs). The AI Graph Agent overcomes the limitations of Large Language Models (LLMs) by building the graph incrementally, one element or relationship at a time. A dynamic prompting system provides instructions based on the local structure of the graph and validates the AI's SysML output.

SysML Model Verification Using AI Graph Traversal

The AI system traverses the SysML graph using breadth-first search to identify and correct inconsistencies. This process includes:

• Checking each Part for issues within its aggregated elements and relationships
• Verifying that fields trace back to source documentation and parent elements
• Ensuring self-consistency between element fields and neighboring elements
• Validating adherence to specific requirements related to part relationships, port connections, requirement traceability, constraint satisfaction, and action/state associations

Corrections are made using a dynamic prompt based on the element's location in the graph, incorporating graph and vector search.

SysML CoPilot for Enhanced User Interaction

The AI CoPilot simplifies user interaction with the SysML model by enabling natural language commands for modifications. The system identifies the relevant subgraph, performs the changes autonomously, and verifies the changes for consistency. Users can review the changes and accept or reject them. The AI CoPilot also assists engineers from different disciplines in understanding the SysML model and its implications.

Benefits of AI-Driven MBSE

The AI-driven MBSE approach offers several advantages over traditional methods:

• Automation: Automates the tedious effort of building SysML models.
• Robust Traceability: Provides more robust traceability than manual methods.
• Automated Verification: Automatically verifies model adherence to SysML standards and consistency with source documentation.
• Simplified Training: Simplifies and accelerates the training needed for engineers to utilize SysML.
• Reduced Costs: Lowers the cost for companies to adopt a robust MBSE approach.
• Improved Collaboration: Provides a more intuitive experience for engineers of different disciplines to collaborate.

Conclusion

By leveraging AI to automate and enhance the MBSE process, ThunderGraph aims to make systems engineering more efficient, accessible, and accurate. The innovations described in this blog post have the potential to revolutionize the application of MBSE across various industries.

Call to Action

We invite you to learn more about this project and explore how AI-driven MBSE can transform your systems engineering practices. Contact us to schedule a demo and discuss your specific needs.