Modelling of Railway Signalling System Requirements by Controlled Natural Languages: A Case Study

The railway sector has been a source of inspiration for generations of researchers challenged to develop models and tools to analyze safety and reliability. Threats were coming mainly from within, due to occasionally faults in hardware components. With the advent of smart trains, the railway industry is venturing into cybersecurity and the railway sector will become more and more compelled to protect assets from threats against information & communication technology. We discuss this revolution at large, while speculating that instruments developed for security requirements engineering can then come in support of in the railway sector. And we explore the use of one of them: the Controlled Natural Language for Data Sharing Agreement (CNL4DSA). We use it to formalize a few exemplifying signal management system requirements. Since CNL4DSA enables the automatic generation of enforceable access control policies, our exercise is preparatory to implementing the security-by design principle in railway signalling management engineering.