Escolar Documentos
Profissional Documentos
Cultura Documentos
CHAPTER 85
ABSTRACT
1 INTRODUCTION
despite the total number of deaths per day being much less. It is only the worst bus
crashes in which large numbers of passengers are killed that seem to capture public
and political concern and lead to the introduction of measures that would not have
otherwise satisfied economic criteria.
The criteria for determining whether a given risk control is economically viable
typically involves a comparison of costs and benefits. The cost of a treatment is
compared to benefits such as the monetary value of mitigated safety loss, which
includes human costs, avoided property damage, avoided delays and other general
Downloaded by [Queensland University of Technology] at 20:17 12 October 2017
costs that are avoided. Human costs are often estimated as a value for preventing a
fatality (VPF) and are commonly calculated using one of two approaches: human
capital or willingness to pay (WTP). Refer to (Tooth, 2010) and (Tooth and
Balmford, 2010) for a discussion of issues in determining the socio-economic costs
of road crashes and accidents at railway level crossings. The WTP approach is
currently considered good practice for determining a VPF.
A key difference between road and rail is who makes the decision on the
implementation of risk controls. In the road environment, the authorities (state
departments of transportation) are responsible for these decisions, whereas in rail,
accredited rail operators (AROs) are responsible. Consistent with a safety
management systems approach to regulation, it is up to the AROs to demonstrate
that they have sufficiently controlled risks to safety in their operations and that they
meet their obligations under the Rail Safety Act.
In determining applicability of a risk control, road authorities would typically
seek a benefit to cost ratio (BCR) of greater than or equal to 1. While a similar
approach seems reasonable for AROs, legislative obligations under the Rail Safety
Act require that risks are eliminated, and where this is not reasonably practicable,
that those risks are reduced so far as is reasonably practicable (SFAIRP) (National
Transport Commision, 2006). The Australian National Transport Council provides
guidance for interpretation of this requirement (Salter, 2008). The guideline states
that in considering the cost of eliminating or reducing the risk, practitioners must
demonstrate that the likelihood of the risk eventuating is remote or that the cost is
grossly disproportionate to the safety benefit (Salter, 2008). While the likelihood of
the risk of a collision at a level crossing eventuating is relatively low, a higher
disproportion factor is likely to be required in order to account for societal concerns.
Legislative requirements for safety in the road environment differ significantly
from the rail environment. In contrast, road safety legislation typically relates to
individuals (i.e. prosecution of individuals for dangerous driving behavior, lack of
roadworthiness of road vehicle, etc.) (Tingvall and Haworth, 1999).
Spatial A. Grade
Separation separation
C1. Signed
Trafc
Control C. Non-
signalized
Time C2. Non-
Separation signed
B.
Signalized
Figure 1. Traffic control hierarchy at an intersection (note that C1. Signed includes road
markings)
A series of traffic control classes have been defined for the two methods of
traffic conflict management. Table 1 provides a comparison between types of traffic
control used at road intersections and railway level crossings, and a brief description
of the decision making process required of the user is provided.
We expect the paradigms for stop and give-way signs to be similar for road
intersections and railway level crossings. Stop signs appear to be more effective at
level crossings than Give-way signs, potentially due to road rules that require road
users to come to a complete stop. Failure to do so is a punishable violation.
Stopping, or at least reducing speed, in theory affords the road user more time to
look for trains and determine whether the level crossing is safe to traverse.
An analysis of the performance of level crossing controls in Australia and the U.S.
was performed by (Independent Transport Safety Regulator, 2011), comparing
flashing lights with half boom barriers, flashing lights only, stop signs and give-way
signs. They observed a significantly lower number of collisions at level crossings with
stop signs compared to give-way signs. The figures indicate approximately 37% less
crashes per million trains and 100 million road vehicles than give-way signs. Of
particular interest to us is the comparison of traffic lights and active warnings at
railway level crossings in terms of differences in risk perception and human factors
issues. The following sections discuss traffic control signals in more depth.
One of the key differences between traffic control signals at road intersections
and at railway level crossings is that communication of system state varies
significantly between the two. While nominal (correct operation) states and failure
states of traffic lights at road intersections are well known, the states of level
crossing warnings are not easily differentiated by the road user. For most road users,
the train approach warning appears to be identical to the indication of failure.
Confusion of what the level crossing warning device is attempting to communicate
can potentially encourage the road user to engage in risky behavior, and this
behavior can influence road user performance at other railway level crossings,
especially if the failure condition occurs frequently or for prolonged periods of time.
This state of confusion is often termed mode error, where an individual applies
the operation appropriate for one mode when in fact they are in another (Norman,
1993). This leads to the execution of an inappropriate action. Mode errors occur
frequently in systems that do not provide clear feedback of their current state.
Table 2 compares the signal states of traffic lights at road intersections and level
crossing warning systems.
In road signal control, vehicle detection is typically used either to change signal
stage (i.e. multiple non-conflicting phases) or to extend the period of the current
stage, depending on where vehicles are detected. The road signals are configured
with a cycle consisting of a sequence of stages. Conflict detection functionality
insures that the signal reverts to the failure mode in the case conflicting phases are
detected (i.e. green aspect for conflicting flows of traffic). Failure of vehicle
detection can result in a delay in passing to the next stage of the cycle. While the
condition is not technically unsafe, sufficiently long delays are likely to encourage
Downloaded by [Queensland University of Technology] at 20:17 12 October 2017
Another key difference between traffic signals at road intersections and railway
level crossings relates to traffic operating rules. At a road intersection, all vehicles
are subject to the same set of rules, whereas at railway level crossings, the rules for
trains vary depending on the type of crossing and jurisdiction. For the most part,
trains in Australia have right of way, regardless of the level crossing signal state.
The level crossing warning signal serves to indicate to the road user when the
crossing is clear, rather than timesharing the intersection. Table 3 compares the
warning and decision making process at different types of level crossings from the
train drivers perspective.
The white or red side lights commonly found on level crossings in Australia are
provided to the train driver to indicate that the level crossing warning device is
working. Some level crossings provide a healthy state indication. According to the
ARTC code of practice (Australian Rail Track Corportation, 2009), the driver must
inform the train controller of a failure, who will subsequently advice the signal
maintenance technician for the area. The train controller will advise any trains or
other rail vehicles in the affected section to approach with caution, and arrange for
any further rail traffic from entering the section. (At the time this paper was written,
the Australian Railway Industry Safety and Standards board were in the process of
developing a national code of practice for Australian Network Rules and Procedures
(ANRP). In the absence of national rules, the code of practice from the railway that
spans the largest number of jurisdictions has been cited).
To the authors knowledge, similar operating rules exist in the U.S. Some
countries in Europe have installed level crossings that provide an obstacle detection
feature, indicating to the train driver the protection state of the level crossing and
whether it is clear. The indication must be installed at a distance from the crossing
sufficient for a train at line speed to come to a complete stop before the level
crossing. In this paradigm, while trains have right of way, they can also be stopped
if the crossing is not clear. Stopping trains, however, may have significant economic
consequences, especially if the obstacle detection system regularly exhibits false
positives.
2013 Taylor & Francis Group, LLC
Several human factors issues have been discussed in relation to traffic control
signals and their meaning to the road user. While the authors acknowledge that there
are other factors at play, this discussion has focused on road user interpretation of
how signals communicate state and signal control issues. Table 4 summarizes
safety issues identified in this discussion.
In developing new interventions for level crossings, a number of issues
Downloaded by [Queensland University of Technology] at 20:17 12 October 2017
5 CONCLUSIONS
The approaches taken to managing safety in road and rail have been shown to be
very different. Some of these differences stem from the divergent performance
characteristics of road and rail vehicles, while others reflect historical differences in
safety cultures. Recent changes in approaches to road safety appear to be creating
more similarities, however. The traffic control hierarchy at intersections (road-road
or road-rail) developed and discussed in this paper provides a novel framework for
identifying potential human factors issues arising from road user confusion.
REFERENCES
Australian Rail Track Corportation 2009. Working of Level Crossings: Rules 1 to 7. TA20 -
ARTC Code of Practice for the Victorian Main Line Network.
Bureau of Infrastructure, Transport and Regional Economics 2011. Australian Infrastructure
Statistics Yearbook. Canberra, Australia
Farmer, C. M. & Williams, A. F. 2005. Temporal factors in motor vehicle crash deaths.
Downloaded by [Queensland University of Technology] at 20:17 12 October 2017