Building resilience into urban rail transport systems

Author: Pierre Dersin
Day: Aspect Day One
Session: CBTC and Metro

Today's urban rail transport networks are an essential instrument for large metropolitan areas in coping with the growing demand for punctual, reliable and environment-friendly transport services. Alstom's solution to address that need is its URBALIS® communication-based train control system.Resilience is the capacity to recover quickly from difficulties; in this context, it is the ability for the system to continue to perform its transport function, or to return quickly to nominal operation, after disturbances caused by unforeseen situations. The situations URBALIS is designed to deal with are of three types: 1) hardware and software failures or degradations; 2) disruptions resulting from various causes, including passenger usage; 3) malevolent attacks.To deal with hardware and software failures, URBALIS relies on highly redundant architectures, in particular that of the data communication system which is its backbone, but also on the use of innovative maintenance and asset management strategies; in particular, predictive maintenance, supported by the Health Hub™platform, aims at detecting degradations before they result in service-affecting failures. Communication system redundancy involves duplicated and fully separated wired and radio networks for end-toend communication between trackside and trainborne equipment. Vital messages, essential for signaling, are sent through two different channels, and non-vital messages are carried by a third network. Wi-Fi communications between trackside and trainborne equipment are managing frequency reconfiguration in case of radio perturbation or cyber-attack. On-line built-in test equipment with high detection rates enables the detection of partial failures before a function loss happens. Redundancy management mechanisms are made as simple as possible and attention has been paid to critical interfaces such as with power supply sources, to avoid single points of failure. Redundancy integrity and data flow are continuously monitored by the centralized maintenance system. Design protection mechanisms against common cause failures have been applied. The centralized equipment that supports key functions such as automatic train control, interlocking and train supervision can be duplicated to provide a standby redundancy backup that guarantees a short recovery time in case of catastrophic events such as flood or fire, or malicious attack.In addition, predictive maintenance is applied to wayside assets such as point machines, whose failures significantly impact service;it consists of performing maintenance operations based on the condition of the assets rather than scheduled maintenance(time-based or distancebased). Just as for the communication network,the goal is to avoid service-affecting failures as much as possible, this time by detecting evolving degradations and intervening before they result in a failure. Machine learning and domain knowledge are combined to construct health indicators which measure the distance to 'perfect health' and are used for degradation detection, diagnostics and prognostics. Beyond that, Health Hub™paves the way for dynamic maintenance management based on evolving asset conditions.Last but not least, adaptive traffic management algorithms ensure quick restoration of full nominal operation after a disruption (resulting from passenger flow fluctuations, passenger behavior such as door obstruction, or external delays).