THERE are a number of situations where active devices - in particular barriers - provide the most effective form of warning at level crossings. Most importantly, active devices can warn road users of an approaching train even when it is not yet visible. For example if the railway is obscured from the line of sight of a road user stopped at the level crossing, then some form of active warning is clearly needed to ensure the road user can cross the level crossing safely.

This is a particular concern in Australia where road trains can be more than 50m long and weigh over 130 tonnes. In such situations very long sight distances are necessary to allow a road train to safely pull away from a complete stop, traverse and clear a level crossing before a train arrives. Often such long sight distances are not available at the road traffic holding point and therefore some form of active warning is essential.

Where a level crossing has more than one track, a second train approaching the level crossing can be obscured from view by the first train. This can lead to a very dangerous situation where road users often believe that the rear of the first train indicates that the level crossing is clear, and proceed into the path of the second train. Once again, active warning devices improve safety in these situations by warning road users of the second train. Furthermore barriers clearly show road users that the first train departing does not mean that the level crossing is clear.

In Australia, however, many of the level crossings without active warning devices have only one track and are located in open country where there are no obstructions to visibility. Therefore there are a very large number of level crossings where the benefit gained from active warning devices is limited. A well-managed programme of level crossing safety should consider the specific risks at each level crossing before determining the most effective, and cost-effective, means of warning.

A concern for many railway infrastructure managers is that their legal obligations are unclear. It is often argued that since a collision at a level crossing is likely to be more serious than a collision between road vehicles at a road junction, infrastructure managers have a duty of care to provide the best warning devices that are available, and therefore barriers should be provided at all level crossings. However, it is unclear whether railways necessarily have the burden of responsibility to protect road users from every folly of their own actions. If a road user has sufficient sight distance to make a safe decision whether or not to cross a level crossing, it is unclear whether the railway should necessarily be going to the expense of providing additional information - after all there is no need to provide traffic lights at every road intersection. Secondly, there is no clear information available to show that active warning devices in Australia necessarily provide better protection than passive devices.

Aus-crossingSince the law is unclear regarding rail infrastructure managers' responsibilities, and since the available evidence in Australia is mute on whether active warning devices provide any better protection, some infrastructure managers install barriers "just to be safe," out of fear of potential litigation if they don't.

The approach of providing barriers as a precaution may be lauded were it not that in some situations barriers can increase the hazards, for example where queuing over the crossing is highly likely. A descended barrier provides a sense of entrapment to road users who are queuing on a level crossing when a train approaches. In most cases there is no real entrapment, since two-quadrant barriers are overwhelmingly more common than four-quadrant barriers in Australia. Furthermore, most road vehicles can readily break a barrier by simply driving at it. Lastly all Australian barriers can easily be crawled under by a person with any degree of mobility. Nevertheless the sense of entrapment can be very real in the minds of people on a level crossing in the path of an approaching train, and has been a direct cause of a number of fatalities in the past.

Barriers can also create hazards where it is highly likely road users want to race in front of a train. On Australia's expanding network of coal railways, trains can be more than 2km long and travel at 60km/h, which means a coal train can take a minimum of 2 minutes to traverse a level crossing. Adding the 30-second activation time for active warning devices can make road users believe they are being intolerably delayed, tempting some to race in front of the train. The presence of barriers in such situations means that road users are spending more time in the path of the train than they otherwise would.

It can be argued that the folly of road users who drive around barriers and thereby disobey the law should not be a consideration for railway infrastructure managers. But, freeing managers from the burden of protecting road users from their own errors could be a reason for never installing barriers in the first place.

The major objection to the widespread use of barriers in Australia is economic. The best available data suggests that the number of fatalities at level crossings in Australia with passive warning devices is between three and seven per year. This is an approximate number since the data are not of high quality.

Similarly the cost of installing active warning devices is not easy to determine. Many level crossings in Australia are in very remote areas where there is no active signalling equipment; to install track circuits or axle counters, and provide the very long cable runs necessary to provide power is very expensive. A safety manager of one of Australia's largest railways estimates that to replace all passive warnings with active warnings would cost about $A 11bn ($US 11.4bn).

Whilst the exact number of fatalities at level crossings with passive warnings in Australia is unclear, I will assume that seven people per year are killed using them. I will also make the unrealistically-optimistic assumption that installation of barriers will completely eliminate all fatalities at level crossings with passive warnings.

Hypothetically, if barriers were installed at every passive level crossing at a total cost $A 11bn, each barrier installation had a useful life of 30 years, and the programme resulted in seven fewer fatalities every year, this would be a reduction of 210 fatalities at a cost of more than $A 50m per fatality prevented. But would this be a valuable use of public funds?

To provide a comparison of the safety risk that passive level crossings pose in Australia, everyday activities such as getting out of bed or eating dinner kill four and eight times respectively more Australians than are killed at passive level crossings. When considering the danger from more common, and preventable, causes of death such as breast cancer or thyroid disorder it is far from clear whether the very high cost of installing barriers at all level crossings is a justified use of public funds.

Clearly barriers would be more cost-effective if they cost less to install and operate. Modern technology is enabling barriers to be installed for a fraction of the cost of current systems. Using innovative train detection technologies such as solar power supplies, wireless communications, and off-rail train detection, a new generation of low-cost level crossing warning devices (LCLCWDs) may provide a practical way to install barriers, particularly in remote areas, where it is currently not feasible. Whilst LCLCWDs may cost around only 10% of existing barrier installations, many of the systems available don't provide the high (SIL 4) degree of protection against failure that exists in current systems.

Work is underway to determine the complex legal and social issues that may arise from providing lower cost barriers if it also involves a possible increase in wrong-side equipment failure rates. Nevertheless, the new technologies look promising for rail safety and should continue to be thoroughly investigated.

Finally, it may not be wise for Australia to establish a precedent whereby barriers are installed as a matter of course at all level crossings. By comparison, while traffic lights are very common at road junctions, there is no road intersection in Australia where barriers are used to provide a warning that road traffic is approaching on the other road. Arguably there are many more situations on Australia's road network where barriers could provide a useful safety control than there are on the railway network.

If barriers were installed unthinkingly at every level crossing, then logically barriers should also be used at road intersections, especially where traffic lights are currently installed. Following this line of reasoning, barriers should be installed at any intersection between road traffic and any other vehicle: such as pedestrian crossings, or even private drives. Such an absurd approach highlights the fact that whilst barriers can be an effective control in some situations, it is clear that they should not be used in every situation where they could be used, indeed they should not even be used at every level crossing possible, but only where they can provide a meaningful improvement in safety.