Eastern Railroad Discussion > GPS Train Control

The accident in Spain brings to mind the thought that we ought to break away from the concept of ground-based train control systems, i.e., signals as we know them today, and move instead toward implementing a GPS system like we have in the aviation world (and trucks, too). If a properly designed system utilizing computer and GPS technology were in place, we'd know where every train in the system was every second, front end and back end, how fast it was moving, what the distance was to other traffic, what the closing rate was, etc., etc. With mechanical systems in place on locomotives (analogous to the autopilot on a modern commercial airliner, which in practical terms can land a 747 in virtually zero visibility), speed could be controlled, fuel could be conserved, crew fatigue would be reduced, and so forth. I would also imagine such a system to be far. far less expensive than the PTC systems that are envisioned today, and I would also think that we could get far better utilization of existing track infrastructure. Okay, narrow mountain passes, train sheds, long tunnels, etc., might present some technical issues, but not any that could not be overcome.

Many airports today no longer have instrument approaches utilizing ground based equipment. Increasingly, navigation and traffic control are dependent on GPS, computer technology, and on-board systems. What am I missing here?

SpeederDriver Wrote:
-------------------------------------------------------
> The accident in Spain brings to mind the thought
> that we ought to break away from the concept of
> ground-based train control systems, i.e., signals
> as we know them today, and move instead toward
> implementing a GPS system like we have in the
> aviation world (and trucks, too). If a properly
> designed system utilizing computer and GPS
> technology were in place, we'd know where every
> train in the system was every second, front end
> and back end, how fast it was moving, what the
> distance was to other traffic, what the closing
> rate was, etc., etc. With mechanical systems in
> place on locomotives (analogous to the autopilot
> on a modern commercial airliner, which in
> practical terms can land a 747 in virtually zero
> visibility), speed could be controlled, fuel could
> be conserved, crew fatigue would be reduced, and
> so forth. I would also imagine such a system to
> be far. far less expensive than the PTC systems
> that are envisioned today, and I would also think
> that we could get far better utilization of
> existing track infrastructure. Okay, narrow
> mountain passes, train sheds, long tunnels, etc.,
> might present some technical issues, but not any
> that could not be overcome.
>
> Many airports today no longer have instrument
> approaches utilizing ground based equipment.
> Increasingly, navigation and traffic control are
> dependent on GPS, computer technology, and
> on-board systems. What am I missing here?


The day someone trips over the extension cord. Can't dump all your eggs in 1 basket, because some day, the bottom will fall out.

SpeederDriver Wrote:
-------------------------------------------------------
> Many airports today no longer have instrument
> approaches utilizing ground based equipment.
> Increasingly, navigation and traffic control are
> dependent on GPS, computer technology, and
> on-board systems. What am I missing here?

And OZ214 never happened. Right?

Ray

Speeder Driver, you would think that railroads would be trying that.

NYC6001 Wrote:
-------------------------------------------------------
> Speeder Driver, you would think that railroads
> would be trying that.

They have been. The biggest stumbling block in designing a purely GPS-based system, is the fact that trains operate on tracks separated by inches, which is below the minimum "resolution" that a GPS based system can recognize. Airliners, on the other hand, operate at minimum separation distances that are recognizable by GPS. Trying to come up with a reliable ... key word being reliable ... purely GPS-based system that can tell the difference between two trains closing on adjacent tracks, instead of the same track, hasn't been easy. That is why the GPS systems being implemented here, are "layered" into the existing signaling systems.

From a pure speed enforcement standpoint, existing GPS technology could have likely prevented the incident in Spain (based on the assumption it was purely an over speed event). But, once you talk about keeping distances from trains or equipment on other tracks, and maintaining operational ability in a manual mode when the system goes down, it gets far more complicated.



Edited 1 time(s). Last edit at 07/27/13 11:17 by bioyans.

To go into a little more detail, in aviation, the Required Navigation Performance (RNP) on the final (read "most accurate") phase of GPS approaches is generally 0.3 nautical miles, and this requires a considerable investment in redundant systems for the aircraft:

http://en.wikipedia.org/wiki/Required_navigation_performance

Another thing that you are not considering is that there are plenty of circumstances at ground level where satellite reception is blocked, preventing a GPS-based position calculation from being made at all. A PTC system must then use something else (like an inertial navigation system) to calculate position. In fact, this is the basis of GPS-based trackers on the cranes that move containers in major terminals, as the "canyons" prevent GPS reception at all times.

Ray

PTC as being installed by the Class Is and commuter railroads uses a combination of GPS, odometer wheels, accelerometers, and "beacons" (any kind of device along the track that transmits a signal from a known location, and can therefore help locate the train). This is easily enough to resolve distances down to a foot or so. Of couse, switch position is also tracked, so if you know the position of a switch when the train passed over it, you know what track the train is on (if the train is in fact NOT on that track, you have other problems, like a derailment).

BNSF is the leader of the pack on this. They and Metrolink in Los Angeles expect to have working PTC by next summer. The rest of the industry is doggedly dragging its feet. They're now asking for a postponement of the statutory requirement to 2018, with a possible extension to 2020. That would be, at worst, 12 years after Chatsworth.

I saw working PTC demonstrated in the Iron Range of Minnesota 26 years ago, with 8086 computers and 4800 baud digital radio. This isn't rocket science.

"The rest of the industry is doggedly dragging its feet" couldn't be farther from the truth. All the class one's are actively investing 100's of millions of dollars per year to bring PTC on line. They are also working as fast as possible - but as someone directly involved with PTC deployment I can say that the devil is very much in the details. The AAR site has great overview of PTC and the facts at https://www.aar.org/safety/Pages/Positive-Train-Control.aspx#.UfPOlo2sir0

Read FRA's Report to Congress on PTC here http://www.fra.dot.gov/eLib/Details/L03718 for a realistic assessment of PTC implementation.

As for BNSF leading the PTC pack, Matt Rose in the Sept., 2013 issue of "Trains", had this to say:

" ... anybody looking at this, they would have to admit the industry is doing everything physically possible to meet the timeline; it's just not going to be met by 2015. If that's really the case, the safety regulators and Congress and the public policy leaders shouldn't want us to implement it until it is right. BNSF will accept whatever delay is provided and use the time to work out the kinks and challenges that will come in this huge transformation in our industry."

The real question is whether the billions spent are the best investment for safety.

Could the money have better results being used for grade crossing updates etc etc.

I would think that once the system is full implemented the costs of maintaining will
be huge. You put in an overpass and you have reduced your future costs. You put in
a relatively high tech PTC and you have to maintain it plus you are probably looking at
delays all over the place from problems.


Bob

All I am saying is it would seem easy for the train to "know where it is" and stop speeding incidients like the one in Spain. If not from GPS, then from transponders at trackside.