At Obeo, we are actively exploring how artificial intelligence can be combined with model-based systems engineering (MBSE) to enhance engineering efficiency and accessibility, while maintaining the rigor required for the development of complex and critical systems.
We currently see two main areas where AI brings clear value:
- AI as a modeling assistant: AI can support engineers in their daily modeling tasks by automating repetitive operations, suggesting improvements, generating model elements, or running intelligent queries. The goal is to make modeling more intuitive and efficient, especially for users who are less familiar with the tooling.
- AI as an interface to access the model: AI can serve as a natural-language interface to extract and communicate information from the model to stakeholders who are not trained in modeling tools. This significantly lowers the barrier to accessing valuable system knowledge captured in models.
We have already prototyped the technical feasibility of these AI-MBSE integrations with tools such as SysON or Capella, and are currently progressing on productizing it within our commercial offers.
The video below showcases a MBSE modeling assistant. It's an example of how we can integrate a LLM and SysON, with Mattermost used as a user interface.
That said, we clearly recognize the current limitations of AI: it cannot replace human validation and critical thinking. When developing complex systems, we cannot afford to overlook potential errors introduced by AI. It is a powerful assistant, not a decision-maker.
Our approach is both collaborative and iterative. We’ve structured our teams to support early-stage customer engagements and experimentation. Once our current developments reach sufficient maturity, we plan to showcase them through a webinar or integrated demonstrations.
