Using the Decision Ladder to Understand Road User Decision Making at Actively Controlled Rail Level Crossings

Rail level crossings (RLXs) represent a key strategic risk for railways worldwide. Despite enforcement and engineering countermeasures, user behaviour at RLXs can often confound expectations and erode safety. Research in this area is limited by a relative absence of insights into actual decision making processes and a focus on only a subset of road user types. One-hundred and sixty-six road users (drivers, motorcyclists, cyclists and pedestrians) completed a diary entry for each of 457 naturalistic encounters with RLXs when a train was approaching. The final eligible sample comprised 94 participants and 248 encounters at actively controlled crossings where a violation of the active warnings was possible. The diary incorporated Critical Decision Method probe questions, which enabled user responses to be mapped onto Rasmussen's decision ladder. Twelve percent of crossing events were non-compliant. The underlying decision making was compared to compliant events and a reference decision model to reveal important differences in the structure and type of decision making within and between road user groups. The findings show that engineering countermeasures intended to improve decision making (e.g. flashing lights), may have the opposite effect for some users because the system permits a high level of flexibility for circumvention. Non-motorised users were more likely to access information outside of the warning signals because of their ability to achieve greater proximity to the train tracks and the train itself. The major conundrum in resolving these issues is whether to restrict the amount of time and information available to users so that it cannot be used for circumventing the system or provide more information to help users make safe decisions.

Language

  • English

Media Info

Subject/Index Terms

Filing Info

  • Accession Number: 01600295
  • Record Type: Publication
  • Files: TRIS
  • Created Date: May 16 2016 2:36PM