The goal of the project HyMAS is the development of a hybrid multi-agent system for self-regulating steel production. A core aspect of the envisioned solution are autonomous software agents that autonomously, collaboratively, and automatically create production plans, monitor their realization, and dynamically adapt them if needed. Simulation techniques and cognitive algorithms allow for anticipatory and learning behavior. Besides integrated optimizing planning techniques, decentralized negotiation-based approaches shall be employed, too, enabling the system to identify the configuration best fitting the current demand. Further, methods of data analytics are used for enable human experts to analyze the control decisions of the production management system (PMS) and to influence them if needed.

The DAI-Labor of TU Berlin is responsible for major parts of the distributed optimization and the underlying agent architecture. In this context, the DAI-Labor will contribute its own software framework called JIAC, which will support the development of such systems. In the course of the project, the general framework shall get extended with novel algorithms for planning, optimization and collaboration in order to fulfill all the requirements posed by the domain of steel production.

This project is co-funded by the European Regional Development Fund (ERDF).