Neues Forschungsprojekt “LLM-SE: Large Language Model unterstütztes Systems Engineering”

One of the greatest challenges in the development of interdisciplinary systems is the complexity, which increases with the number of subsystems, disciplines, technologies and their links and dependencies. In the case of mechatronic systems, the coordination and synchronization effort between the development work of the various specialist disciplines such as mechanics, electronics and software is particularly pronounced. Model-based systems engineering (MBSE) is a holistic approach to comprehensive modelling, coordination and parallelization that helps to master complexity and thus supports the goal of improved resource utilization in the engineering process. Systems are not designed using traditional documents, but on the basis of machine-readable models, which are typically created in the form of UML (Unified Modeling Language) or SysML (Systems Modeling Language). Despite the numerous advantages, MBSE is not yet sufficiently widespread in practice. This is due in particular to the high barriers to entry in SMEs, which are caused by the large number and complexity of the usability of diagram types, their creation, management and continuous further development, among other things, in order to be able to fully describe a corresponding system.

The LLM-SE research project sees the potential and pursues the goal of reducing entry barriers and at the same time increasing development efficiency and quality in MBSE through the introduction of Large Language Models (LLM), which are currently receiving particular public attention through image and text processing methods. To this end, an assistance system based on the philosophy and structure of MBSE is being developed using large language models to partially automate the engineering process of mechatronic systems from requirements analysis to virtual commissioning. This is intended to interpret user input such as requirements and adaptations, transform them into individual, model-based solutions through the availability of company-specific product catalogs, historical project data and best practices, and secure these through suitable validation and verification mechanisms. This significantly increases the efficiency of planning processes and reduces and avoids errors at an early stage through automated verification and validation. This simplifies coordination and reduces the operational effort in engineering projects and enables Bavarian SMEs in the special machine construction sector to manage the transformation from document-driven engineering to end-to-end digital MBSE.

You will find further information on FAPS’ Web page.