Satellite train trial gives the all-clear

[zardoz]

From issue 2487 of New Scientist magazine, 19 February 2005, page 25
By Mick Hamer


REPLACING railway signalling equipment with satellite navigation systems could sla***he cost of maintaining the network and provide a financial lifeline for lightly used rural railways by cutting their costs.

A nine-month trial using a train on the UK rail network, reported at a conference in London last week, showed that the satnav technology could successfully replace trackside signals. It reliably displayed signalling information inside the driver's cab, while equipment in the signal control room showed the position of the train.

Maintaining conventional signalling systems can eat up as about 10 per cent of the cost of running a railway. The new system has been geared to help operators of lightly used railways, says Gareth Close of satellite consultancy SciSys, based in Chippenham, Wiltshire, in the UK.

Railway signalling systems divide the track up into short sections, each guarded by a signal that remains at red while the section ahead is occupied. These trackside signals are expensive to maintain and vulnerable to vandalism. In some countries it is not uncommon for people to steal lengths of cable and sell them for the copper they contain.

The trial, which was backed by the British National Space Centre, was the first on a publicly used railway. It followed trials of a similar system on a test track in Belgium last year. The test train carried a GPS receiver capable of determining its position to within 4 metres. Using a link to the Iridium satellite communications network, or a mobile phone network if the satellite link failed, the system then broadcast the train's position to signal boxes.

Software feeds this position data into the railway's existing signalling system, which keeps track of all the trains over a broad area of the network. This works out whether to display a danger signal or the all-clear on the panel in the train cab that takes the place of the trackside signals.

While GPS navigation is used extensively in aircraft and ships, railways were expected to pose more problems. Deep cuttings could block the train's view of the GPS satellite's signal, and electrified sections could interfere with the signal itself or with the cellphone link. The trial proved these fears groundless, and the only blind spots were tunnels. "We knew where the train was to within 4 metres of accuracy," Close says. "We were able to follow it right up to the buffers of [London's] Paddington station, even under the glass canopy on the train shed."

But Peter Ladkin, a safety analyst at the University of Bielefeld in Germany, warns that the system may be unreliable. "It is easy to jam GPS, inadvertently or deliberately."

[BaltACD]

CBTM is not all that new. CSX has been testing a system for several years and I believe the other Class I carriers also have their own test beds.

The biggest problem, for US roads will be inter-operability. The ability of the 'Final System' to operated correctly on all the railroads and under all the various Rule Books that those railroads have.

The old saying 'The Devil is in the details', is what makes and will make CBTM (Communications Based Train Management) a difficult nut to crack for US railroads.

[Anonymous]

Amtrak used a GPS system as part of its high speed experiment, ITCS (90mph), on its track from New Buffalo,MI to Kalmazoo,MI. Don't know what they are doing now. Lawrence.

[Anonymous]

Amtrak used a GPS system as part of its high speed experiment, ITCS (90mph), on its track from New Buffalo,MI to Kalmazoo,MI. Don't know what they are doing now. Lawrence.

[kenneo]

We had quite an extensive thread on this last year. Does anyone have an update on the US tests sponsored by USDOT?

[daveklepper]

Of course this could be tied in to grade crossing protection with TV cameras and pattern recognition equipment and solve (or mitigate greatly) both problems at once!

[tree68]

One of the problems pointed out in earlier threads had to do with that 4 meter resolution. That's ~13 feet. Not sure what the current standard spacing of double tracks is, but if it's less than or equal to 4 meters, the system has no way of knowing which track a train is actually on. Now you're back to track sensors.

I'm all for technology. If something like this can help safety and productivity on lines where it's appropriate (I'm thinking mostly single track, combined with voice confirmation of things like taking sidings), then let'er rip. Reporting past waypoints would theoretically be a thing of the past since the DS could see it on his screen.

On the other hand, on double tracks, it sounds like they still have some bugs to work out. And there is that interoperability thing.

[Anonymous]

QUOTE: Originally posted by zardoz
But Peter Ladkin, a safety analyst at the University of Bielefeld in Germany, warns that the system may be unreliable. "It is easy to jam GPS, inadvertently or deliberately."

Considering this, it might be prudent to wait until all the bugs are worked out. Ladkin's statement does not fill me with confidence in the system.

[FThunder11]

I like the satillite idae, but wont the railroads have a lot of layoffs of the signal maintaners if this happens?

[tree68]

QUOTE: Originally posted by FThunder11

I like the satillite idae, but wont the railroads have a lot of layoffs of the signal maintaners if this happens?

As long as there are crossings there will be something for the maintainers to do. Besides, this represents a tech shift. Somebody still has to maintain all that equipment...

[Hugh Jampton]

Odd that I both live in the UK and work in the rail industry,, yet I've never heard of this...
4 metres is not accurate enough for rrail ,, how do you know which track the train is on given that the distance between tracks is half that??

[dehusman]

I don't understand how a TV camera at a grade crossing will prevent crossing s. In a significant portion of the s the car runs into the side of the train and the vast majority of the rest of the crossing s the car is moving. All a TV camera will do is record the .

Dave H.

[dldance]

QUOTE: Originally posted by equinox

QUOTE: Originally posted by zardoz
But Peter Ladkin, a safety analyst at the University of Bielefeld in Germany, warns that the system may be unreliable. "It is easy to jam GPS, inadvertently or deliberately."

Considering this, it might be prudent to wait until all the bugs are worked out. Ladkin's statement does not fill me with confidence in the system.

Since the airline industry is also strongly considering a GPS based navigation system, I would have to question the comment about the ease of jamming GPS. From a statistical basis, it does not appear to be that easy to jam. There are about 28 satelites in orbit (several are spares). The average location on earth is in direct line of sight to about 1/3 of them. The GPS system only needs to see 3 to calculate a location. Further, the GPS system knows where the satelites are supposed to be because the orbits are highly predictable. Thus a reliable system can detect and disregard a jammed signal.

Further a transportation GPS system will be augmented with ground based reference signals to further improve accuracy. The commonly quoted 4-meter accuracy is based on the commercial signal that formerly had a dithering error added. This error generator was turned off about 6 years ago. With the use of ground-based reference signals and with the error generator off, GPS can be accurate to about 1/2 meter or better. This is sufficiently accurate that GPS systems are frequently used in land survey and measurement today. Farmers are even using the system to keep the rows in their fields straight.

dd

[jeffhergert]

Do they contemplate getting rid of track circuits altogether or just the wayside signals?
If it's just the wayside signals, you could go to cabsignals without the satellite. If it's the entire track circuit, how do you detect broken rails?
Jeff

[Anonymous]

I personaly don't like the sound of it anyway,I mean this will put the Public at risk!
Some how.

[Anonymous]

And by the way......
This will make Railfanning kind of BORING!