Model-based system development

Methods and principles of systems engineering help to manage the increasing complexity of products and interdisciplinary development processes. Model-based systems engineering is based on formal, consistently linked models for describing the product, taking different perspectives into account. The IMW is researching methods for the development of consistent model chains based on the adapted RFLP approach. The clear distinction between models for mapping requirements (R), functions (F), operating principles and logical systems (L) and the physical realization (P) improves the traceability and transparency of interdependencies, decisions and changes in the interdisciplinary development process. Research at the IMW focuses on

• hybrid modelling and visualization methods for the consistent linking of formal (e.g. SysML) and non-formal, domain-specific (e.g. sketches) models in the development process
• (Automated) identification, creation and reuse of models and model fragments for knowledge transfer between development projects
• Introduction of systems engineering and product-specific modeling methods in companies, including training and support for pilot projects

In most development projects, existing systems or subsystems are reused and adapted. The targeted evolution and variation of system architectures is a prerequisite for the efficient and low-risk development of new systems. To this end, we are researching methods for the model-based mapping of reference architectures and the identification of architecture patterns based on recurring structural features. We develop methods for the integration of context information (e.g. specific requirements, organizational or technological constraints) for the targeted support of the reuse, evolution and variation of architecture models. Topics of our research are

• Methods for model-based architecture description and automated derivation of reference architectures based on recurring structural features
• Methods for the evolution and variation of architecture models through targeted principle and design changes
• Strategies for the modularization of system architectures, taking into account cross-domain function realization
• Integration of context information in architecture models to support knowledge management in the development process

The integration of new procedures, development methods and systems engineering tools often fails due to "grown" processes that specify information flows as well as development tools and methods. Model-based process engineering takes up the central importance of process models in product development and transfers approaches from model-based system development to the engineering and further development of processes. The aim is to research and transparently map the complex interactions between development activities, development results (models), organizational boundary conditions and development tools and methods. The focus is on the following research topics:

• Analysis of development processes and activities with regard to knowledge intensity, degree of cooperation, robustness as well as derivation and formal description of process patterns for recurring development activities in systems engineering
• Integrated modeling of process and product information based on formal languages (Business Process Model and Notation, Systems Modelling Language)
• Graph-based approaches for the visualization and analysis of interactions between development activities, product models, (software) tools and organizational constraints