Elsevier

Applied Ergonomics

Volume 42, Issue 4, May 2011, Pages 548-554
Applied Ergonomics

Driver behaviour at rail level crossings: Responses to flashing lights, traffic signals and stop signs in simulated rural driving

https://doi.org/10.1016/j.apergo.2010.08.011Get rights and content

Abstract

Australian road and railway authorities have made a concerted effort to reduce the number of rail level crossings, particularly the higher risk passive crossings that are protected by devices such as ‘give way’ or ‘stop’ signs. To improve this situation, passive level crossings are often upgraded with active controls such as flashing red lights. Traffic signals may provide good safety outcomes at level crossings but remain untested. The primary purpose of this research was to compare driver behaviour at two railway level crossings with active controls, flashing red lights and traffic signals, to behaviour at the current standard passive level crossing control, a stop sign. Participants drove the MUARC advanced driving simulator for 30 min. During the simulated drive, participants were exposed to three level crossing scenarios. Each scenario consisted of one of three level crossing control types, and was associated with an oncoming train. Mean vehicle speed on approach to the level crossings decreased more rapidly in response to flashing lights than to traffic signals. While speed on approach was lowest for the stop-sign condition, the number of non-compliant drivers (i.e., those who did not stop) at the crossing was highest for this condition. While results indicate that traffic signals at rail level crossings do not appear to offer any safety benefits over and above flashing red lights, further avenues of research are proposed to reach more definitive conclusions. Compliance was lowest for the passive crossing control which provides further support for the ongoing passive crossing upgrades in Australia.

Introduction

While there has been a recent increase in rail safety research (Wilson and Norris, 2005), the intersection of the railway system with the road transport system, through rail level crossings (RLCs), represents one key area in which optimum efficient performance has not yet been achieved. Adverse outcomes at RLCs, such as collisions between trains and vehicles, impact both the efficiency and safety of the rail and road systems. As an example of the size of the problem, there were 2592 accidents and 892 fatalities at RLCs in the European Union across 2006–2007 (European Railway Agency, 2009). In the US in 2007 there were 2206 crashes and 229 deaths (U.S. Department of Transportation, 2009). Within Australia during 2009 there were 58 collisions at RLCs (Australian Transport Safety Bureau, 2009). These crashes typically involve traumatic injury, and the economic impacts of disruption on the rail and road networks are significant. This is particularly so for heavy vehicle collisions as they have a much greater potential to derail the train.

Across Australia there are approximately 9400 rail level crossings, with 6060 passive (60%), 2650 (30%) active, and 690 (10%) having other forms of control (Australian Transport Council, 2003). Level crossings are not homogeneous; some provide active warnings (e.g., flashing red lights), while others provide only stop signs (referred to hereafter as passive crossings). In addition, there are differences in the volume of rail and road traffic, the type and speed of traffic, overall RLC geometry, and so on. All of these factors shape road user behaviour at RLCs and thus the appropriate solution.

Compliance with level crossing controls has been found to be variable and ranging from 10% to 67% with flashing lights (Meeker and Barr, 1989, Pickett and Grayson, 1996; Tenkink & Van der Horst, 1990) and from 14% to 38% with barrier gates and lights (Meeker et al., 1997, Witte and Donohue, 2000). Yet as noted previously (e.g., Tenkink & Van der Horst, 1990), the various causes of level crossing crashes remain poorly understood, but fall into two broad classes. The first is unintentional error, encompassing situations where the drivers may fail to detect the warnings or to apprehend their meaning, even if the site is known and the warnings are clearly visible. This type of error accounts for nearly half of the fatal level crossing crashes in Australia (ATSB, 2002) compared to 22% for all other fatal crashes. These drivers do not detect potential threats and therefore are at great risk. The second explanation may be that drivers see the lights and are fully aware of their meaning, but intentionally cross on their own judgment (e.g., Witte and Donohue, 2000). Male drivers with strong sensation seeking tendencies and prior frustration with level crossings are more likely to not comply with crossing signals.

A wide range of RLC countermeasure have been trialled in the field, however the evaluation of many of these is not robust. These include road user education, enforcement, speed restrictions, improvements to sight distance, rumble strips, improved train conspicuity, along with a range of active treatments (Edquist et al., 2009). Active treatments are accepted as having superior safety performance than passive crossings. For example, flashing lights at RLCs have been shown to increase driver crossing compliance (Tenkink & Van der Horst, 1990), while the benefits to behaviour and crash performance provided by flashing lights and boom barriers are clearly demonstrated to be superior than for flashing lights alone (Meeker et al., 1997, Wigglesworth and Uber, 1991).

Increasing warning salience and familiarity has been identified as an important measure to improve safety. Traffic lights at RLCs may represent another active countermeasure to improve crossing salience, although the safety benefits in this context remain unclear (Edquist et al., 2009). The use of conventional traffic signals may increase compliance with RLC controls through increased warning salience and by taking advantage of the familiarity that road users already have with this type of signal. In a US study, two months of data were collected and compared before and after the installation of traffic signals at an RLC previously equipped with flashing lights. While there was no significant difference in perception–brake response time (measured as the time from the lights being activated/turning red until the vehicle’s brake lights came on), the number of vehicles crossing during activated/red lights was reduced by 80% (Fambro et al., 1989).

Two more recent studies also demonstrate the safety potential of traffic signals at level crossings. RLC with traffic signals and half barriers (i.e., those that cover a single lane) were found to be slightly more effective in terms of reducing crash rates than those with flashing red lights and half barriers (Schwarz, 2006). In fact, the use of traffic signals with half barriers is now the preferred treatment in Germany, and flashing red lights with barriers are no longer manufactured. The rationale for this change was to capitalise on-road users’ familiarity with traffic signals. A second study in Israel examined driver behaviour at flashing light-controlled RLCs before and after the installation of additional traffic signals positioned 24 m in advance of the crossing (Becker et al., 2008). Following the installation of the traffic signals there was a relative drop of 57% in the number of vehicles crossing while the flashing red lights were activated.

There are challenges to conducting research into driver behaviour at RLC, both in the field and using simulation, that no doubt contribute to the fact that relatively little research is conducted in this area. A primary obstacle is the low rate of exposure to oncoming trains. This presents practical challenges when trying to assess behaviour in the field as a very large number of testing hours are required to collect very modest numbers of observations, which does not make such studies cost-effective. Driving simulation can be used to provide the controlled and increased exposure to oncoming trains that is missing in field studies (Triggs, 2002). However, this must be balanced against the risk of measuring non-representative behaviour through exposure to a high number of low-frequency events within a short time period.

The present research capitalised on the control provided by the use of simulation to study behaviour at RLC. Specifically the project aimed to compare driver behaviour at RLCs with two active treatments: traffic signals and flashing red lights. A further aim was to compare driver behaviour at the two active RLCs to behaviour at a passive RLC with stop signs only. It was hypothesised that RLCs protected by flashing red lights or traffic signals would provide greater safety benefits than would passive RLCs equipped with stop signs, and that the strongest benefits would be associated with the use of traffic signals. The primary anticipated safety benefit at active RLCs was increased crossing compliance.

Section snippets

Participants

Nineteen males and six females aged between 25 and 50 yrs (M = 34.4 ± 9.5 yrs) took part in the study. Participants were recruited by means of notices posted within the local community and advertisements in a local community newspaper, and were compensated $20 for their participation. Ethics approval was obtained from the Monash University Standing Committee on Ethics in Research involving Humans.

The driving simulator

The mid-range simulator is located at the Monash University Accident Research Centre. It consists

Results

The proportion of time looking at the central versus the peripheral regions of the driving scene on approach to RLCs did not differ among crossing types and is not presented here.

Discussion

Flashing red lights, alone and with boom barriers, are currently the primary countermeasures used in Victoria to upgrade passive level crossings to active crossings. Traffic signals represent another potential yet untested possibility in this regard. This exploratory research compared driver behaviour at RLCs with two active treatments, traffic signals and flashing red lights. A further aim was to compare driver behaviour at the two active RLCs with a passive RLC with stop signs only.

There is

Conclusion

It is clear that RLCs with active protection have lower crash risks than those with passive protection. However the costs of installation and maintenance of many active treatments, such as grade separation and flashing lights with boom barriers, make widespread implementation near impossible. For this reason it is important that the likely safety benefits of more cost-effective options be assessed.

Flashing red lights represent one of the primary countermeasures used in Australia to upgrade

Acknowledgements

This project was performed under contract for VicRoads and was funded by the Victorian Railway Level Crossing Safety Steering Committee. We would like to thank the members of the stakeholder advisory group for their inputs in project discussion. In particular we note the following members of the stakeholder group for their input to the development of the simulated level crossing scenarios: Mr. Colin Kosky; Mr. Ken Hall, VicRoads; Mr. Peter Upton, V/Line; and Mr. Jim Warwick, Vic Track.

We also

References (37)

Cited by (100)

  • Safety and associated parameters influencing performance of rail road grade crossings: A critical review of state of the art

    2021, Journal of Safety Research
    Citation Excerpt :

    Simulation results of dual gated RRGCs by Coleman and Moon (1997) showed that aggressive drivers traversing the crossing against the approaching train generally cross the safe stopping distance and penetrate beyond the railway track, which can lead to a serious collision. The driving simulation tool is effective in evaluating behavioral parameters, including stress level, secondary activity while driving, and familiarity with RRGCs (Lenne et al., 2011; Tey et al., 2011;Tung & Khattak, 2015; Young et al., 2018) and the use of driving simulator has been validated for RRGC (Larue et al., 2018). Young et al. (2018) revealed that drivers spend more than 50% of their time texting without looking ahead on driving through RRGCs, which could make them unaware of the instant situation.

View all citing articles on Scopus
View full text