Model-Based Evaluation of the Availability of a CBTC System

A metro control system is a software/hardware platform that provides automated mechanisms to enforce the safety of a metropolitan transportation system. In this field, the current technical trend is the Communications-based Train Control (CBTC) solution. CBTC platforms are characterized by a continuous wireless interaction between trains and ground controls. Several degrees of automation are provided, from basic traffic monitoring to unattended train operation. Besides safety issues, a CBTC system is also required to guarantee a high level of availability. These platforms are normally composed of several subsystems and devices, and estimating the overall availability of the system is not a trivial task. Stochastic Activity Networks (SAN) are a powerful formalism that allows modelling and evaluating complex distributed systems. In this paper, a study is presented that shows how SAN models can be employed to evaluate the availability attributes of a CBTC system. The current results show that the SAN technology and the analysis tool adopted, named Möbius, are mature for a profitable employment in industrial practice.