This paper discusses human factors issues of low cost railway level crossings in Australia. Several issues are discussed in this paper including safety at level railway crossings, human factors considerations associated with the unavailability of a warning device, and a conceptual model for how safety could be compromised at railway level crossings following prolonged or frequent unavailability.
The current paper summarises and extends pertinent literature that must be considered for effective interventions to improve safety and to advance our theoretical understanding of human behaviour at level crossings. Although the results of our research are not presented, we describe our experimental approach to progress the current lack of knowledge in this area. In particular we highlight where we can improve previous research methodology (independent & dependent variables) when investigating right-side failure at level crossings, which can produce results with greater validity and meaning. Our research aims to quantify risk to motorists at level crossings following right-side failure using a Human Reliability Assessment (HRA) method, supported by data collected using an advanced driving simulator. This method aims to identify human error within tasks and task units identified as part of the task analysis process. It is anticipated that by modelling driver behaviour the current study will be able to quantify human reliability. Such a risk assessment for the impact of right-side failure at level crossings is currently absent in the literature. Therefore it is crucial to offer quantification of success and failure of this intricate system. The task analysis allows human error identification for the precursors to risky driving to be achieved. If task analysis is not employed the error reduction method may be unsuitable and eventually unsuccessful.
Our aim is also to determine those contexts that allow the system to operate successfully with the smallest probability of human error. Human behaviour during complex tasks such as driving through a level crossing is fundamentally context bound. Therefore this study also aims to quantify those performance-shaping factors that may contribute to vehicle train collisions by highlighting changes in the task units and driver physiology. Finally we consider a number of variables germane to ensuring external validity of our results. Without this inclusion, such an analysis could seriously underestimate risk.
