jeffhergert In one of the front columns of Trains (Technology column?) a couple months back there was an incident in territory where the PTC (or near PTC) system Amtrak uses overlayed on an existing signal system failed to detect an open hand throw switch. (It sounds like the signal system failed to detect the switch.) I was surprised that I didn't see (unless I missed them) any comments on here about that. Since the signal system didn't "see" the open switch, the PTC didn't "see" it either, allowing a train to go into a side track at an unsafe speed. So maybe PTC would've prevented Graniteville, and maybe it wouldn't have.
In one of the front columns of Trains (Technology column?) a couple months back there was an incident in territory where the PTC (or near PTC) system Amtrak uses overlayed on an existing signal system failed to detect an open hand throw switch. (It sounds like the signal system failed to detect the switch.) I was surprised that I didn't see (unless I missed them) any comments on here about that. Since the signal system didn't "see" the open switch, the PTC didn't "see" it either, allowing a train to go into a side track at an unsafe speed. So maybe PTC would've prevented Graniteville, and maybe it wouldn't have.
Methinks that such switches will become a thing of the past in PTC territory. I'm seeing shelters going up along a line near here in the vicinity of switches. This is dark territory.
I have come to the conclusion, though, that while PTC will likely prevent another Graniteville, it will do nothing to prevent another Oneonta, Kingman, or Rockford.
None of the four railroad incidents in my immediate area that I'm aware of would have been prevented by PTC. No lives were lost, but they did tie the railroad up for a while and definitely caused damage to track and rolling stock.
Larry Resident Microferroequinologist (at least at my house) Everyone goes home; Safety begins with you My Opinion. Standard Disclaimers Apply. No Expiration Date Come ride the rails with me! There's one thing about humility - the moment you think you've got it, you've lost it...
While I agree there is no 100% solution, I still think that there has to be something better than a piece of paper for running on a track that allows speed greater than 10mph.
Just my opinion.
It's been fun. But it isn't much fun anymore. Signing off for now.
The opinions expressed here represent my own and not those of my employer, any other railroad, company, or person.t fun any
tree68Methinks that such switches will become a thing of the past in PTC territory.
Probably not. What WILL happen is more thoroughgoing installation of positive position indication on these switches, and some form of robust communication on switch position AND lock status to the "PTC system".
There are far too many places on current Class 1s where power-operated crossover switches aren't cost-effective. This is far more than just a 'siding' situation. But it's comparatively cheap and easy to instrument the switches (and most PTC systems I've studied have just that functionality implemented).
I feel that we need a less 'brittle' control modality if PTC is to work, and part of making the system better is to have positive indication when switches stop 'responding' or start providing improper data for any reason. Most of the 'rest' of the issues with these switches is already well addressed in operating and safety rules -- the PTC would just be doing supervisory checking on that activity...
Overmod tree68Methinks that such switches will become a thing of the past in PTC territory. Probably not. What WILL happen is more thoroughgoing installation of positive position indication on these switches, and some form of robust communication on switch position AND lock status to the "PTC system".
That's pretty much what I meant. I don't see manual switches going away, for cost reasons, as you point out - but totally unmonitored switches will be come exceedingly rare, at least on main track.
Hand throw switches in signalled territory aren't unmonitored, they are interlocked (for lack of a better word) into the signal system. So are some derails. When the main track switch or derail is opened, it is supposed to shunt the signal system. It didn't happen in the instance that appeared in Trains. It wasn't known at that time if a component had failed, or had been intentionally tampered with so as not to shunt the signal system.
The component, which I have seen called a "circuit controller," I don't know if that's the correct name though, is probably what will be added to switches in dark territory to show the position of the points for PTC.
Jeff
zugmann While I agree there is no 100% solution, I still think that there has to be something better than a piece of paper for running on a track that allows speed greater than 10mph. Just my opinion.
Like the old saying, "Speed costs money. How fast do you want to go?"
Going have to spend something as "doing nothing" is no longer going to cut it.
jeffhergert Hand throw switches in signalled territory aren't unmonitored, they are interlocked (for lack of a better word) into the signal system.
Hand throw switches in signalled territory aren't unmonitored, they are interlocked (for lack of a better word) into the signal system.
Hand throw switches in dark territory have no signal system with which to interface. If the railroad sees no reason to install signals, yet has to include some manner of PTC, then some interface has to be added.
The line of which I speak is single track - a GPS based system won't have to discern which track the trains are on. All it has to know is where the front and rear of the trains are.
I do not have any idea as to how many of us non-railroad employees have had an opportunity to really look at hand throw switches used in signaled territory, but I did have such an opportunity when I was living in Wesson, Mississippi. There was a crossover (trailing points) between the two main tracks (double track) of the IC, and I saw that the switches (just as all others for the main tracks) were connected to the block signal system--they HAD to be, for safety's sake.
Johnny
The issue is not so much that they are 'connected to the signal system' as that the method by which they are is redundant and tamperproof. I have said elsewhere that there needs to be positive indication that the throw has been made completely (and perhaps relocked) and separate positive indication that the points are actually hard over and the rod not bent, etc. The usual method of activating the signal system won't assure either of those things.
Meanwhile, all that discussion 'back in the day' of ATC testing about what happens when you have relay-based ABS and back into the wrong block is still relevant in many places. If the ABS indication is the only input to PTC, the system is no more 'positive' than ABS+ATC is. And many sources have indicated that in this day and age, that's insufficient.
Yes, it is essential that the switch points be in the proper position and be locked in the proper position. However, as I recall, if the points are not in the proper position even if the hand throw lever is in the proper postion for normal traffic the block signals will show red.
I may not remember correctly, but, as I recall, if only one crossover switch is lined for a cross over movement the signals for both directions go to stop.
As to being locked, the night that I lined both switches back after a freight that had had to back over onto the wrong main to let the City of New Orleans was back on the right main, the conductor shone his light on what I was doing until I had locked both switches. The engineer did not have to stop for the rear end men to line the switches back; the conductor had told the fireman that I was going to line the switches and thus expedite the train on its way. Note: this took place forty-eight years ago, at a time when non-railroad employees who were known to employees could do things that are absolutely unheard of now.
Bucyrus I could not see the point of driverless cars, but driverless commercial trucks makes lots of sense, especially when we all know that there will NEVER be driverless trains (unless they can develop a robot that can replace knuckles).
I could not see the point of driverless cars, but driverless commercial trucks makes lots of sense, especially when we all know that there will NEVER be driverless trains (unless they can develop a robot that can replace knuckles).
10000 feet and no dynamics? Today is going to be a good day ...
traisessive1 Bucyrus I could not see the point of driverless cars, but driverless commercial trucks makes lots of sense, especially when we all know that there will NEVER be driverless trains (unless they can develop a robot that can replace knuckles). There already are driverless trains. Australia started testing in 2008 and are still using the technology today. The technology is there and if Australia can send a train across the desert with a computer controlling it then any railroad here can too. The public access is the issue.
Public access as in grade crossings. The "driverless" trains in Australia are operated on the Pilbara ore lines on the far side of nowhere, where grade crossings don't exist because nobody lives there.
But why does a grade crossing matter to driverless trains? They can't stop in time to yield to road traffic anyway.
Bucyrus But why does a grade crossing matter to driverless trains? They can't stop in time to yield to road traffic anyway.
But they can stop to take names and notify authorities and render human assistance.
Never too old to have a happy childhood!
BaltACD Bucyrus But why does a grade crossing matter to driverless trains? They can't stop in time to yield to road traffic anyway.But they can stop to take names and notify authorities and render human assistance.
I see what you mean, but couldn't that human assistance be called out as an automatic response to a locomotive colliding with something significant? Generally, it would seem that taking the crew off of a train would require added auxiliary support. ie: Lots of guys in trucks spread out along the way to take names, render assistance, replace knuckles, etc.
I see what you are doing here.
We were discussing the eventual full automation of the railroad the other day in the crew room. It was agreed that eventually, when the already exsisting cameras focused out the front of the locomotive indicate that a crossing accident has occured, a computer program will access the damage and an ATM on the side of the locomotive will dispense the proper compensation. Perhaps a nice sympathy card in the case of fatalities.
zugmann Bucyrus But why does a grade crossing matter to driverless trains? They can't stop in time to yield to road traffic anyway. I see what you are doing here.
Doing? What do you mean?
Bucyrus Doing? What do you mean?
No, no, no... you go ahead. I'll sit back and watch.
Buttered popcorn or plain?
Beer or soda?
23 17 46 11
How many grade crossing crashes have been prevented by the preemptive action of the engineer? I know some are prevented if the train can be stopped short of a stalled vehicle on the track, but PTC will prevent that. If I am not mistaken, that is a planned feature of PTC. So, I do not see how the existence of grade crossing precludes driverless running.
BucyrusI know some are prevented if the train can be stopped short of a stalled vehicle on the track, but PTC will prevent that. If I am not mistaken, that is a planned feature of PTC.
You are correct.
There are in fact a couple of systems independent of "PTC" that serve to alert engine crews if crossing warnings have failed, or if there is a vehicle resident on one of the loops or visible to radar or camera coverage. I don't have firsthand experience with them, or know how well they work.
So, I do not see how the existence of grade crossing precludes driverless running.
I could start semi-snarkily and say 'well, does the presence of grade crossing safety devices preclude collisions?' Systems fail or degrade, and in those cases, and many others, the judgment of an alert human engineer is the best 'protection'.
The trick is to design this so the engineer, and not the safety system or PTC, is actually driving the train. This rules out the usual sorts of 'cruise control' instantiations, or the ridiculous 'vigilance' alerter system. (Note that the FRA/DOT just remain puzzled about why this sort of system doesn't produce safety or satisfaction... pauvre petits)
The real issue with driverless running, whether of cars or consists, is that it's an open field day for plaintiff's bar. ANY problem caused by all those hellish foot-pounds of momentum will be deemed ever more likely to be -- at least in some part -- the railroad's fault or liability. And remember that under joint & several, even a fraction of a percent leaves the railroad liable for 'whatever the other parties did not pay'. Wanna bet the same problem comes up here as in medical-malpractice case law?
I am not holding my breath waiting for tort reform, or expecting some sort of Price-Anderson cap on railroad liability. On the other hand, some clear understanding of this must be developed, and all problems remedied, before autonomous train control becomes do-able (as opposed to technically practicable).
RME
edblysard Buttered popcorn or plain? Beer or soda?
Plain & Hard Lemonade
BaltACD edblysard Buttered popcorn or plain? Beer or soda? Plain & Hard Lemonade
Let me have some of the latter. A lot of the latter. Perhaps as a chaser.
Bucyrus How many grade crossing crashes have been prevented by the preemptive action of the engineer? I know some are prevented if the train can be stopped short of a stalled vehicle on the track, but PTC will prevent that. If I am not mistaken, that is a planned feature of PTC. So, I do not see how the existence of grade crossing precludes driverless running.
While signalled crossings are eventually planned to be integrated into PTC systems, I don't think it's to detect vehicles stopped on them. I've understood it to be that the crossing signal system will communicate it's "health," that is that the crossing systems have no malfunctions that would keep them from operating normally.
To be able to detect a stopped vehicle would mean installing some form of detectors or sensors that could sense metals, weight and/or motion. (Just think of all the rabbits setting off motion detectors and bringing everything to a stop.) I haven't heard of any plans to start installing such items at crossings.
In one of the Professional Icononclast columns many years ago, JGK mentioned using such a setup at crossings to appease the public when using fully automated (no onboard crew member) trains. He said a computer could react faster than a human and that some crashes would happen anyway because they would be too late for man or machine to react to. He also said that the detector systems could be programmed to take into account the size of a object, and he thought if the object was small enough that it wouldn't harm the train, the train shouldn't be stopped. Just let it barrel right on through.
Jeff,
Yes perhaps you are right that this entire crossing vision will not be part of the first rollout. I can see the "smart" crossing (diagnostic) concepts being added at the get-go, but not the far ahead sensing of obstruction. Besides the cost of reliable sensors, there is the problem of separating objects ahead that will move in time from those that won’t. You can only secure the road so far in advance without unduly disrupting commerce ahead.
I think this first application of PTC will begin an evolution of train operation that basically never ends. There is a lot of energy expended in developing advanced products to sell to railroads. The vision is practically unlimited at this point. You pour diesel fuel in one end and transportation rolls out the other. Everything in between is up for grabs.
jeffhergertTo be able to detect a stopped vehicle would mean installing some form of detectors or sensors that could sense metals, weight and/or motion. (Just think of all the rabbits setting off motion detectors and bringing everything to a stop.) I haven't heard of any plans to start installing such items at crossings.
Jeff, look at the papers in the FRA digital library. There is a pretty good discussion of one approach for 'instrumenting' a crossing that covers many of the issues. (I can provide you a copy via PM if you can't locate it on line. Use the 'elib' link if you Google for it -- FRA has changed their document server tech.)
If technology is implemented that stops trains when there is a vehicle 'blocking' a crossing I can see a new game - Stop the Train. And laugh all the way home.
With mile and more stopping distances, the train will, in many cases, not even be in sight of the crossing when the technology commands the train to stop because of the 'blocked' crossing.
I cannot see PTC or any other technology being implemented that will 'prevent' trains striking vehicles at road crossings, without that technology seriously disrupting rail operations, as in the 2+ minutes that would be requred to bring the train to a stop short of the crossing, the vehicle causing the initiation of the braking event gets it's act together and departs the crossing (never having realized what actions were initiated on its behalf).
Our community is FREE to join. To participate you must either login or register for an account.