Although it is accepted that system design affects train driver performance, the literature related to this phenomenon – in relation to urban railways and metro systems in particular – is scarce. Metro systems differ significantly from mainline railways, being closed systems, with shorter headways, a greater number of stations and more signals encountered. This paper aims to investigate the effects of design-related performance shaping factors on metro driver performance, by analysing historical incident records for the 2011–2013 period on the Tyne & Wear (T&W) Metro (UK). Bivariate statistical analysis has been used, to assess the potential inter-dependency of the performance shaping factors and other common causal factors, for various driver-related incident types. In addition to category A signals passed at danger incidents, station overruns, and incidents associated with station procedures have also been assessed. The results show the significant importance of the location (design) based performance shaping factors in incident propagation mechanisms in the Metro. The two years under investigation display increased consistency between driver-related incidents and locations, rather than time of day, or season. In addition, the highest correlation between incidents has been found in terms of locations. Deviations from a standardised T&W Metro station design were found to associate with either an increase or decrease in incident rates, depending on whether additional complexity or simplicity was introduced. Although the features of metro systems suggest improved route knowledge and system familiarity among drivers, the results show that this can actually lead to an unsatisfactory safety-related performance during non-routine operations, e.g. engineering works, or driving in sidings.
Analysing and Modelling Train Driver Performance
