Platform for digital engineering with semantically linked data
In the development of complex, safety-critical systems, a wide range of heterogeneous models and code artifacts are created — for example, requirements, designs, and implementations of hardware and software components — which often overlap in content (such as geometric, electrical, and software interfaces of sensors, actuators, or control units). Due to the large number of differently standardized formats for control software, CAD data, and electrical engineering design, inconsistencies frequently arise, increasing coordination effort as well as development time and cost. At the same time, finding information across different domains and data formats is becoming increasingly difficult and time-consuming.
The DigiSemNet project therefore investigates how the use of heterogeneous engineering models across multiple disciplines can be improved. The DigiSemNet platform developed for this purpose builds on semantic networking and the use of generative AI, in particular large language models (LLMs). To achieve this, heterogeneous engineering data are linked within the platform using a data graph. This graph captures precise semantic relationships between elements of models from different engineering domains, enabling efficient traceability and comprehensive access to all aspects of mechatronic system components.
In the DigiSemNet project, key contributions to integrating Model-Based Systems Engineering (MBSE) with other engineering disciplines are being explored. A major focus is on the semantic networking of heterogeneous data across domains, disciplines, and life cycle phases to break down information silos and enable a holistic understanding of systems. To achieve this, the project enables the implementation of a digital thread that ensures traceability and consistency of system information across diverse data sources and tools. Another key contribution is the automated consistency checking of linked data to detect contradictions and inconsistencies at an early stage.
In addition, the project investigates approaches for leveraging heterogeneous data with generative AI to make documented knowledge more accessible — for example, through semantic search or adaptive assistance systems. For all these solutions within the emerging DigiSemNet platform, mechanisms are also being developed to protect confidential information, ensuring the security of sensitive data even in the context of data networking, AI-based analysis, and consistency verification.
Central Innovation Programme for SMEs (ZIM) of the Federal Ministry for Economic Affairs and Energy (BMWE)
01.11.2025 - 31.10.2027
