Exploration of a Generative AI Assistant for Model-Based System Engineering (eGAIA4MBSE)

Open Access
Article
Conference Proceedings
Authors: Bryan Croft

Abstract: Model-Based System Engineering (MBSE) has served as a formal model of systems engineering in terms of requirements, design, analysis, verification, and validation. This process is applied and updated throughout the lifecycle of a project. Such creative models provide the means to exchange information about the system. A consensus exists that the MBSE process has improved the systems engineering process especially when compared to the document-based approach. MBSE uses a form of XML called SysML to represent the MBSE model and a set of diagrams like UML diagrams used in the software development arena. MBSE has grown in use and expanded into other areas including simulation. This poster represents the ongoing work to incorporate and integrate avenues of Generative Artificial Intelligence (GenAI) into MBSE. GenAI has quickly reached a level of capability, maturity, and widespread use in recent years. The generation of content as well as supporting MBSE users in Development including the supporting SysML, positions GenAI to play an important role in the modeling with MBSE. Work is underway to explore how GenAI can generate the MBSE content and support the end user in providing crucial feedback on the rules and processes involved with MBSE while generating MBSE content. This work tends to show how GenAI along with RAG-based MBSE information can be incorporated into GenAI to serve as an AI agent which supports the development, validation, and documentation of said models. GenAI has been poised to be such an agent for this creative MBSE generation and user support in the systems engineering process. The application of GenAI with MBSE is still in the infant stages and this work seeks to explore the effectiveness of that integration.

Keywords: Generative AI, Model-Based Systems Engineering, Assisted Agent

DOI: 10.54941/ahfe1006393

Cite this paper:

Downloads
13
Visits
57
Download