The future as at best a haze. We can make or break it. My rants, visions, ideas and dreams hope to make a better future. Lets learn to live a better life. Raise your glasses (no it doesnt have to be alcohol - I am using orange juice!), and toast the future!

Wednesday, February 16, 2005

Trains & Sensors

Trains have had a problem for many years. With the current UK privatised system, different companies have scheduled control over different sections of track for different times. What this means is that a red light may mean that it is unsafe to go somewhere, but it may also mean that it is commercially not allowed. The problem being that drivers are encouraged to save time, and will sometimes skip signals on the assumption that they are commercially driven and none to do with safety. The mistake lies in the fact that they could at any time be wrong. This is a serious issue, which led to tragedies like the paddington disaster a few years ago.
One way of solving these would first be to lift the crappy beaurocracy that led to such a situation in the first place - this is definately the best option.

But even given that, drivers still make mistakes. People run over lights, trains go over points that werent at the position they should have been.

Train companies are currently looking into using expensive to track their rolling stock. But is that really necessary?

Robot People (like ) and safety groups have been arguing for years that trains should have sensors. These were always rubbished as "too expensive" and updating the whole fleet of rolling stock would be a logistics nightmare - surely GPS is even more expensive, and has the same logistics nightmare - it also does not even guarantee the same level of safety. Here are some of the ideas which should probably be considered instead:

Each train has on its front and its rear a long-range sonar proximity scanner, which with a processor, takes into account the doppler shift (that is sounds change frequency when reflected or source from a subject moving towards or away from you) and distance of oncoming objects. It then uses this to govern the maximum allowable speed of the train - or bring it to a stop - this is made failsafe by simply disabling the locomotive if the governing signal is not continuous.

The idea of intelligence in the trains, not the track - only makes sense in the commercial light, not in passenger safety, or maintainability or value for money. It is a great deal more simpler, and easier for a computer to flag if something is wrong, if all points had sensors to indicate their current position, and strategic sections of track had sensors to indicate if rolling stock is passing them. These with some fairly simple electronics and computer systems - could maintain the system, and even in the commercial environment, be used - with an RFID on the rolling stock - to charge companies for there use of rail (meaning they no longer need slots but are actually charged for their usage). Combine with logging speed governers - this would give a much safer, flexible and even commercially viable system than the GPS plan.

Anyway, there are many more improvements to be made on our rail system - but this would be a good start.

Other links:


Post a Comment

Links to this post:

Create a Link

<< Home