SoSEng

The aim of the project is to significantly shorten innovation cycles in manufacturing and logistics, and thereby strengthen the international competitiveness of European industry. To this end, “classical” automation systems will be extended with LLM‑based agents that can autonomously take on sub‑tasks or provide advisory support enabling the management of complex systems (system‑of‑systems). The agents obtain access to production resources via the Model‑Context‑Protocol (MCP), while critical decisions are safeguarded through human authorization and a fine‑grained role‑ and permission‑management system. The agents do not require uniform syntax and semantics standards for protocol elements; they can tolerantly interpret informally described machine and plant interfaces and consequently control tasks, ultimately shortening the innovation cycle considerably and increase operational performance. These capabilities enable the engineering and control of highly complex, flexible, and heterogeneous systems-of-systems as well as their dynamic re-configuration. In several demonstrator set-ups, subsystems of intra‑logistics and material handling, robotized production cells and maritime logistics will be used to validate the system‑of‑systems approach.