If my $200 Garmin GPS can tell me the current speed limit within 100 ft of the speed limit sign, and have the numbers turn from black to red when I exceed the speed limit in any area of the USA, why can’t they use this technology to over ride a train conductor and limit the trains max speed based on location?
The conductor can go as slow as he wants, and can stop for whatever, whenever. But when he approaches a corner, the train will self govern at 50 mph, for example.
You should build this and get rich from it.
First, how often is inadequate speed control a cause of accidents?
Second, is it worth it for say… BNSF or Union Pacific to retrofit all their engines with that technology?
Third, who’s going to decide what the adequate speeds are, and where those adequate speeds are? Are you going to use existing railroad speed limits?
Fourth, who’s going to keep the engines AND the master database updated? In other words, who’s going to go update every engine, and who’s going to keep the master updated in case east BFE decides they want trains to go slower in their one-horse town?
This sounds to me like a solution in search of a problem.
This has already been proposed. My understanding is that implementation is lagging due to technology and practicality problems.
this
I can assure you, it will be a lot easier to automate than the Dc-10 jumbo jets that they started automating landings on, about 15 yrs ago. Trains have a fixed 2D path, and weather and wind speed direction are moot.
It could also deploy the parking brake when unattended (esp with haz mats).
We have seen how this simple technology could have saved dozens of lives, and an entire city, just this month, in 2 unrelated events.
Speeding through curves is an incredibly rare cause of train wrecks. Much more common is derailment/splitting a switch and collisions. But GPS isn’t precise enough to differentiate parallel tracks, and isn’t reliable in all weather and all locations.
My idea was for each segment of track to have a simple RFID tag with a unique number. Each locomotive would sense what segment it was on, and call (well, text) in to ask the central computer for a track warrant to occupy that segment. Of course, if one couldn’t be issued, the train would come to a stop.
Presumably similar ideas were proposed before the very complex PTC system was chosen as the one to implement nationwide.
Closer to 50 years ago actually. Autoland is a mature technology.
Uh… The technology already exists (ERTMS, European Rail Traffic Management System) and it is in use in certain stretches of track.
I guess that this thread was prompted by the accident in Spain a couple of weeks ago. The biggest problem there was that the accident took place where a new high-speed line links with an already existing standard-speed line.
The High-speed line was equipped with the ERTMS system, but the existing line was equipped with the older ASFA system. Copy-paste from the relevant article (emphasis mine):
“The high-speed track has ERTMS-compliant signalling, which is designed to slow or stop a train whose driver is ignoring signals or the speed limits. However, the new high-speed line joins a conventional track shared with low-speed trains, at the curve where the accident happened. The conventional track only has the older ASFA signalling system, which will warn drivers if they are exceeding speed limits, but will not automatically slow or stop a speeding train. There is a different system capable of stopping a train if it passes a red signal but that was irrelevant in this case.”
The train had left the ERTMS system and was only under control of the ASFA system when the accident took place. Apparently the driver ignored the speed warnings from the computer.
But yeah… The system, as described by the OP, already exists and is operative. It has not been implemented throughout the whole rail network because there is a LOT of rail to cover. The first lines to be fitted with it have been the high-speed lines, and slowly other stretches of track are being covered.
Thanks JoseB. I am not saying that it happened in this case, but I could see how a train driver would get a little complacent and just say “I’ll leave the speed control on maximum and just let the computer slow me down when necessary.” Then he leaves the ERMTS zone and enters the ASFA zone and forgets the difference.
Again, I’m not saying that had anything to do with this accident, but I would think that might be a risk of using mixed technologies. Maybe the system should require the train to come to a complete halt before switching from one technology to the other, just like when you are leaving a freeway to go down a residential street.
Well, at least for the mid-term, the two technologies will coexist. ERTMS coverage is being extended, but as I said, there is a lot of rail to take care of, and switching from one system to another, although not particularly difficult, is not straightforward. There will be many places where both will coexist, and I don’t think it would be practical to require the train to completely stop in the middle of the tracks before switching from one control system to another.
It comes down to making sure that the drivers are attentive, and that the ERTMS gets implemented more and more. In the end possibly the whole network will be under ERTMS control, but it will take time for that to happen.
You’re still using a solution to chase after problems. If the engineer deliberately ignored warnings then there are an infinite number of ways that could happen.
And if we start relying on such technology to do the obvious then there is the risk of dependency failure. The chances of mechanical malfunction is probably higher then that of a person making the same mistake when all of the systems are taken into account. I just flew over Lake Michigan for the first time 2 days ago and damn if GPS didn’t lose the signal of every satellite. I didn’t pick them up until I crossed the shore line. The one place I couldn’t look down and see my location on a map. And of course it’s a cloudy hazy twilight and of course it’s my first flight of the year not counting a check ride. But it’s basic flying skills that were needed. I was already on a heading and I stayed on it.
Dependency failure leads to stupid accidents like the Asiana flight into SFO. 4 pilots including a check-ride pilot manage to fubar a standard approach that a student pilot would recognize as severely compromised all because the ILS was out and they didn’t flip one switch. If my plane was dependent on the GPS I’d be dead. But that was the easy part of the flight. My destination included rain and a thunderstorm cell. I flew over the intended airport at night at 1000 feet AGL and still couldn’t see the runway lights. All I could see were the approach lights and the storm cell prevented me from making a normal approach. I ended up diving for the runway with the airspeed in the yellow and I crossed the numbers at 50 ft going sideways in a slip which I rode down the centerline for a couple of hundred feet to bleed off airspeed. A computer would not have been able to group together basic observations and make a decision outside of what is programmed into it.