FURTHER RESEARCH OPPORTUNITIES TO IMPROVE RAIL LEVEL CROSSING SAFETY
The research programme led to the identification of several future research areas that could build on the work undertaken. These include field trials of proposed design changes to rail level crossing environments, the application of computer simulation and modelling techniques to explore the impact of design changes, and research efforts to better integrate the findings of human factors research into economic modelling for cost-benefit analysis.
Firstly, there are opportunities for additional evaluation processes to support the potential uptake of recommended design changes. One potential means could be the testing of prototype designs in controlled field trials. For example, a prototype of the mirrors from the Ecological Interface Design concept could be built and tested via controlled field trials. Such trials could initially involve the use of test tracks, with participant observers in stationary vehicles as trains traverse (to mitigate the risk of participants being involved in collisions if there is a poor response to the mirrors), under various environmental conditions. This could be done with a range of road user types, including car drivers, heavy vehicle drivers, motorcyclists and cyclists, and could enable collection of data around reaction times, situation awareness, decision-making and the potential for emergent behaviours introduced by the new design.
Other interventions could potentially be implemented and tested in the real world according to the notion of safe-to-fail experiments proposed by Snowden and Boone (2007). This approach acknowledges that, when implementing changes in complex systems, we are unable to predict the effects until they are implemented in practice. Following the safe-to-fail approach, design concepts could be introduced in stages, beginning with the least novel components (i.e. for the Ecological Interface Design concept - adding guide posts to create an illusion of travelling at higher speed) and ending with the most radical (i.e. the replacement of traditional signage with mirrors). With careful monitoring of behaviour occurring during the transition, it would be possible to incrementally measure the impacts on behaviour, including the identification of emergent behaviours not initially predicted. This enables a change in strategy to occur quickly if monitoring reveals potential risks. Such an approach to evaluation requires a flexible and ongoing commitment to evaluation and thoughtful planning to ensure its appropriateness when conducted in high-hazard domains. Further, it requires a culture that accepts failure and easily adapts to try new innovations, which is unusual in safety- critical industries due to strict requirements for demonstrating the effectiveness of safety interventions prior to implementation. This again highlights where the current protective structure may impede the application of systems thinking in transportation design.