Dependability and safety evaluation of railway signalling systems based on field data

Abstract: Railway signalling systems can be considered a group of complex systems that together provide control, supervision and protection of railway operation. When a failure occurs in the railway signalling system, other safety mitigation measures are put into place to continue railway operation, e.g. reduced speed, with the driver responsible for safe operation. Briefly stated, failures of the railway signalling system affect both the capacity and safety of the railway. The purpose of this study is to propose tools to assess the dependability and safety of infrastructure signalling systems in the operation and maintenance phase, to support continuous improvement. The overall research strategy is a case study at Trafikverket (Swedish transport administration). Quantitative data have been collected through the fault reporting system (0Felia) for a specific railway line and through interviews, document studies, and observations. The type of failure is either random or systematic, according to the railway standards. The analysis of random failures is based on a theoretical framework of RAMS (Reliability, Availability, Maintainability, and Safety) for signalling systems. Failure data have been analysed using such methodologies as FMECA (Failure, Mode, Effects, and Criticality Analysis), Markov analysis, and reliability analysis for repairable systems. Systematic failure data analysis is based on theories and expert judgement in the area of configuration management and information logistics. The results of the study can be divided into two parts. The first part comprises three tools that can be used to assess the dependability and safety performance of railway infrastructure signalling systems based on random failure data. The first tool in this section adapts the FMECA logic in an analysis model to assess the operational RAMS performance for different levels of infrastructure signalling systems. The second tool combines statistical methods with expert assessment to perform a reliability assessment of complex long-lived repairable systems, such as railway signalling systems. The third tool evaluates the safety and availability of the railway operation based on the state of signalling systems. The second part of the study is a guide on how to apply configuration management and information logistics to deal with systematic failures related to infrastructure signalling systems. These tools can be used during the operation and maintenance phases to assess the dependability and safety of a signalling system in the railway infrastructure. For example, the FMECA-based tool reveals that a maintenance approach based on the causes of failures instead of particular systems can be more appropriate to support improvement of the maintenance of railway signalling systems. The results obtained from the case study show the importance of external influences on the dependability of signalling systems and the need to improve the quality of collected data. This was manifested by a high variability of the reliability and maintainability of different systems with the same architecture and operational characteristics. This has two implications for the study. First, it shows the focus of the adapted tools to analyse failures must be on a system level, instead of on more detailed indenture levels. Second, it suggests how to improve maintenance support by enhancing the configuration management and information logistics. The results can be used improve the management of all types of failures, thereby reducing risks and supporting the safety integrity of signalling systems.