An integrated AI and Model-Based approach for railway cyber-physical systems

The design process of models and the integration of documentation, without compromising efficiency, is crucial in Cyber-Physical Systems (CPS), particularly in safety-critical applications. This paper presents a robust methodology for modeling the railway CPS pertaining to the radio system of the European Rail Traffic Management System (ERTMS). The proposed Artificial Intelligence (AI) pipeline systematically identifies the requirements for the Radio Block Center (RBC) to be implemented using a Multi-Agent Systems (MAS). Subsequently, the generated models are parsed to automatically produce comprehensive documentation. An in-depth analysis of the RBC main connection component at the station entrance is also provided. This methodology effectively leverages AI and modeling techniques for complex safety-critical systems while ensuring adherence to essential requirements for explainability and traceability.