Incident Factor Classification System and Signals Passed at Danger

The Incident Factor Classification System (IFCS) is a cross-industry database of incident underlying causes which is managed and populated by RSSB for the GB rail industry. It forms part of the broader industry incident... [ view full abstract ]

The Incident Factor Classification System (IFCS) is a cross-industry database of incident underlying causes which is managed and populated by RSSB for the GB rail industry. It forms part of the broader industry incident database called the Safety Management Information System. This paper will provide background on the current status of the Incident Factor Classification System and describe how the database has been used to support industry in considering human factors issues.

The paper will summarise the range of uses which the database is being put to, particularly focusing on the analysis of Signals Passed at Danger (SPAD) incidents. SPADs continue to be a contributor to railway risk and continue to be closely monitored and managed by both the GB rail industry and more widely for international railway systems.

The train driver task of observing railway signals plays a central role in the occurrence of SPADs. Task performance reliability for observing and acting on railway signals is an outcome of the broad system design which involves, in the GB rail context, different companies (one who owns the signals and another who is responsible for the driver and train). Each has safety management systems and different design processes, procedures, competence management systems and safety cultures. Task performance reliability is crucially dependant both on how the organisations define and manage the task and the staff who undertake the task.

This paper will look at the insights gained in to the human performance and the underlying causes of SPAD incidents using the Incident Factor Classification System. The paper will also outline the qualitative and quantitative human factors approaches (such as workshops with front line staff and managers) which have been used to better understand the incident data. In addition, normalisation data have been collected by the industry. When these data are combined they allow us to develop an understanding of both success and failure in this task, where we can link task performance reliability back to deeper organisational issues. The paper will describe the method and some key results from the analyses.

My submission will have industrial application. I would like the paper to be considered for inclusion in a special edition of the Journal of Rapid Rail and Transit.