THE first thing the off-duty policeman saw was a car stopped on the train tracks, where the road to Ufton Nervet crosses the main line to the west of England, 70 kilometres from London. Almost at once, the barriers of the level crossing came down to stop further traffic, which meant a train was little more than 30 seconds away. The policeman rushed to the telephone beside the crossing to warn signalling staff. But it was too late. A crowded express travelling at 160 kilometres an hour hit the car and came off the rails. The accident, last Saturday, killed seven people.
The crash highlights a worldwide problem. In the US, a train hits a vehicle on a level crossing every 150 minutes on average. In Europe, accidents on level crossings kill 300 people a year, though the UK has one of the better safety records: more people die on level crossings in Germany, France, the Netherlands, Portugal and Belgium.
Millions of cars and trains use level crossings every day without problem. But on the rare occasions when things go wrong, we need a better back-up than a passer-by, a telephone and signalling staff trying to radio a warning to the train driver. The inadequacies of this cumbersome procedure were all too obvious last Saturday.
Advertisement
Level crossings could be eliminated by closing roads or building bridges, but such measures are unpopular or prohibitively expensive. There is a better way. Radar is already used to detect flocks of birds near airports, to control aircraft movements in crowded airspace and in collision-avoidance systems for passenger aircraft. Some cars have simple radar systems to warn drivers when they are reversing into an obstacle.
Level crossings are small, well-defined spaces, so monitoring them for obstacles with radar and linking this up to the signalling system should be straightforward. The hardware already exists. If it detects a car on the tracks, it could alert the signalling staff and train driver, and even automatically apply the train鈥檚 brakes.
The system would have its downsides. A stalled car that quickly restarts might cause a slight delay, but this is a small price to pay, given the catastrophic consequences if it fails to move. There is also the question of who should pay: should it be the railways or the highway authorities?
None of this is impossibly futuristic. One British technology company was already in talks to install a pilot system on UK railways (see 鈥淩adar could prevent disaster鈥�) even before Saturday鈥檚 accident. The system will doubtless need some development. But it promises to be a simple and relatively cheap way to cut the global death toll on level crossings.