after having been in the flying business for 20+ years a few thoughts on PTC are in order. The FAA first came out with the TCAS (traffic collision avoidance system) proposals in about 1985. First implemented about 1990 TCAS I had about 5 versions in two years. Then TCAS II about 9 through 2000. TCAS III is on version 7.2? now and just now has enough bugs worked out to implement on helicopter operations. Lets see— 23 years. and boy do I believe in TCAS now. But even though research is ongoing it will probably take more than 4 years. RWM is so right that it will take a lot of work. Even though some software from TCAS might work in PTC - GPS is not accurate enough for application here. TCAS works on measuring radar transponder time, distance, altitude, and relative location of all other aircraft. On RRs train length needs to be accounted for, Switch positions for every switch on your route, including dark territory as well, a need for open coding so all systems whatever the RR will work together, multiple tracks, Other RRs running next to your track, track geometry, slow orders, work zones, etc. Bet RWM can list two dozen more. And PTC eventually on all RR lines!
The point is------it will take time. During the interim maybe more visible signals, more double track, straighten curves, 2nd man, other items spoken of.
The issue seems to be that PTC is being viewed by many (not just political leaders) as the silver bullet that will eliminate human error. It’s also expected that the railroads, transit agencies and other operators will eat the cost of development and installation without passing it along to their customers. Absolute safety will be impossible to attain as long as the human factor is involved.
Blue Streak – you can go buy this morning all the PTC hardware and software you need, off the shelf, plug-and-play, ready-to-run, works with seven-nines (99.99999%) reliability, for any system from a dark single-track railroad to a multiple-main track CTC high-speed passenger and freight railroad, and get it delivered and installed in a matter of months.
Technical feasibility is not the issue here. Economics isn’t even the issue here. U.S. politics, U.S. law, and the U.S. regulatory system is the issue here. The FRA has been charged by Congress (and Congress charged by the voters) with creating an impossibility – a regulatory framework that is 100% perfect for all people at all times and upsets no one ever.
We have precisely the system the U.S. voters want to have, and when the system doesn’t work they blame everyone other than themselves. We have met the enemy and he is us – the voters. Those guys in Washington are just a bunch of hired hands that we put there; we voted for them. We got what we wanted.
(you could have have PTC, ATS, cab signals etc. and still something like this would happen…and then would come all the complaints about train delay caused by guess what?)
If man builds it or conceives it, he still can most certainly still screw it up. The politicians had best go back to their caves and get their fires started, it’s gonna be a long winter legislating common sense.
Can it put a train in emergency if it runs a red signal?
Can it reduce the train speed or put the train in emergency if it is speeding?
Can it in one way or another monitor if the engineer is actually controlling the train, or if he is fallen asleep, having a heart attack, and in that case have the train in emergency and the dispatcher alerted?
In my opinion, if it could do those three with 99.9999% accuracy, it should be installed on every major rail line ASAP.
As always, consider the source, although this seems to be pretty much on the mark.
And what is it with putting a train into emergency? That’s a last resort and fraught with danger. The action of choice is a full service application, so as to bring the train to an orderly, controlled stop.
Well that’s a relief to know that a federal agency is working on a solution. The census bueau was tasked with doing an electronic census. they have spent over a billion dollars for a handheld computer system that doesn’t work in any way. Yet UPS can track your package anywhere in the world at any time. I can’t wait (to not ride those trains).
The Wiki entry is OK from my point of view, that is, if you know what it’s supposed to do you can read it and understand it. Beats me if it makes any sense from a non-expert point of view.
One major error is that it says that PTC provides reactive braking in case of authority excursion or overspeed. Actually it’s predictive. That’s the key advantage, along with much lower cost and higher reliability than cab signals or ATC/ATS.
I’m curious as to what definition of reliability is used to derive that figure. If it is one false clear in 10E7 opportunities, that figure may be achievable (i.e. heavy use of fail-safe design). A broader definition of reliability would tend to imply an availability of seven-nines (down for 3 seconds per year), which would be difficult to believe - as that would require anticipating and correcting for a lot of very low probability failure modes.
RWM - dont’ tell us, tell Dianne Feinstein. Not that she’d understand OR care. Also, please clue in the MSM. Not that they’d understand OR care either.
The biggest human factor accident cause is careless drivers at grade crossings … PTC does nothing exept to prevent the rarest accidents of all , collisions.
It’s the likelihood that a system error will lead to an unsafe condition instead of a fail-safe condition, or one in 10 million. At a minimum, CFR 49 236 Subpart H requires that a PTC system or any microprocessor based train-control system be proven to have no higher a rate of unsafe conditions as existing train control systems such as CTC.
Seven-nines reliability is not uncommon; for example, it is standard for life-safety critical radio communications such as digital microwave networks supporting 911 numbers. You can get nine-nines reliability too.
I didn’t tell you that. The problem is hardly that recent. Our gobbledygook railroad safety policy dates to the 1800s and has become only more byzantine with time. Us voters have had thousands of opportunities for more than a century to change the way we want this country run and haven’t lifted a finger to do so, which means either we like it just fine or we don’t care. So which are we – apathetic, stupid, or satisfied? I think the latter, because despite all our carping we return 90% or more the people who “screwed it up” right back to the same job, regardless of pa
If a train was to stall on a hill or because of locomotive failure, how would you bring in another locomotive or train to couple on to a stalled train on the mainline in PTC ?
Exactly the same way you would do it with CTC, TWC, or DTC: You issue a joint authority to both trains, which requires both to operate at Restricted Speed (able to stop short of train, engine, stop signal, derail, open switch, men & equipment, keeping a watch out for broken rail, not to exceed 20 mph in any case). The PTC system will enforce the 20 mph speed limit; alternatively, you can set a parameter in the software at any lower speed you wish. If you wish, you can put a parameter in the software that finds the length of the stalled train off either the train sheet or a GPS transmitter in the EOT device, and limits the helper locomotive approaching it from the rear to any lower speed you wish when at any distance from the end of the train you wish.
Existing signal and control systems usually have a permissive signal which allows that option, I’m sure that it would be included in any PTC installation.