GENAI WORKBENCH: AI-Assisted Analysis and Synthesis of Engineering Systems from Multimodal Engineering Data
Researchers propose an AI system that reads engineering docs to auto-generate system architectures and requirements.
Researchers H. Sinan Bank and Daniel R. Herber have proposed the conceptual framework for the GenAI Workbench, a novel AI-assisted Model-Based Systems Engineering (MBSE) environment. The system aims to solve a critical disconnect in modern engineering: while platforms like CAD, CAM, and CAE excel at detailed component design, they often operate in isolation from high-level systems engineering frameworks that manage requirements and architecture. The GenAI Workbench seeks to integrate these worlds by establishing a unified digital thread, linking semantic data from documents, physical 3D geometry (B-rep), and relational system graphs into a single open-source Product Lifecycle Management (PLM) platform.
At its core, the workbench facilitates an AI-driven workflow where a designer can feed in source documents. The system then uses vision-language models to automatically parse these documents, extracting key requirements and using that information to synthesize an initial system architecture, such as a Design Structure Matrix (DSM). This proposed methodology represents a significant shift towards a more integrated, data-driven design process, potentially reducing integration risks and fostering holistic development from concept to component. As a work-in-progress framework accepted for presentation at the 2026 IISE Annual Conference, its anticipated impact lies in automating the tedious translation between system-level intent and detailed design, promising a more informed and efficient engineering methodology.
- Proposes an AI-assisted MBSE framework that integrates systems engineering directly into designer workflows using an open-source PLM platform.
- Uses vision-language models to automatically extract requirements and generate system architectures (like DSMs) from ingested source documents.
- Creates a unified digital thread by semantically linking documents, 3D B-rep geometry, and relational system graphs to bridge design and systems engineering.
Why It Matters
Automates the gap between high-level system requirements and detailed CAD design, reducing integration errors and accelerating complex product development.