PTC

[Bob Bufkin]

http://thehill.com/policy/transportation/257566-agreement-reached-on-automated-train-extension

[davefoxx]

That might freak some people out, but check out the statistics from the U.S. Department of Transportation on deaths by transportation mode;

http://www.rita.dot.gov/bts/sites/rita.dot.gov.bts/files/publications/national_transportation_statistics/html/table_02_01.html_mfd

For example, in 2012, there were 35,699 total deaths by some form of transportation.  Of this figure, 33,782 are by highway, 449 are by air, and 677 are by rail.  That is, deaths by railway accidents in 2012 was only 1.89% of the total of transportation fatalities.  Even fatalities by air was small (1.25%), but deaths on the highway accounted for 94.6% of the deaths.

It would seem that a lot of money being invested for PTC will only potentially prevent a very small percentage of fatalities by transportation modes.  However, consider also that approximately one-third of the railroad deaths in 2012 was highway-rail grade crossing deaths (230 fatalities).  I'm not sure that PTC is even going to make much of a difference there.

I'm not saying that PTC isn't worth it, because how do you place a dollar value on a human life?  But, it would appear that PTC is not exactly going to save thousands of lives per year, so a three-year minimum extension hopefully won't cause a major uproar in the general public.

DFF

[C855B]

Ah... but... that "677 by rail" is not train accidents. That big number includes grade crossing collisions and "other" that are not train vs. train collisions or derailments resulting in a recorded fatality. Stuff that PTC can't fix.

The actual number of fatalities due to train accidents in 2012 is 9. That's right, nine.

So... exactly what is it we're getting with PTC?

[Scottl]

Uh, isn't this about freight trains?  As you point out, passenger rail and related fatalities are very small. 

I personally suspect PTC is just the first step towards train automation.

[C855B]

I "drove" a UP cab simulator in a PTC demo, where the supervisor switched the PTC on and off to show what it would do "for" me if I didn't respond in time. Heck, it was blowing the horn at grade crossings if I didn't. PTC definitely highly automates the routine chores of running a train, enough so that a train can become a form of "drone".

I think we're on the verge of remote operators monitoring cab video feeds only for exception conditions, while PTC 2.0 (or whatever) handles everything else. A single engineer could probably handle four trains simultaneously.

[Rich_S]

I "drove" a UP cab simulator in a PTC demo, where the supervisor switched the PTC on and off to show what it would do "for" me if I didn't respond in time. Heck, it was blowing the horn at grade crossings if I didn't. PTC definitely highly automates the routine chores of running a train, enough so that a train can become a form of "drone".

I think we're on the verge of remote operators monitoring cab video feeds only for exception conditions, while PTC 2.0 (or whatever) handles everything else. A single engineer could probably handle four trains simultaneously.

Mike, with PTC and LEADER the railroad is getting very close to being able to operate trains over the road without any engineers.  When the PTC system is fully implemented, it will be aware of every mile of main line track, every turnout, every signal, every curve, grade, speed restriction, including temporary speed restrictions and grade crossing. It will also be aware of the location of every other train on the main line. The next piece of the puzzle is LEADER, which determines the best speed and throttle position to operate the train over a specific section of main line. When everything is said and done, the only thing that will be missing is the ability to tie locomotives onto a train, perform an air test then tell them when it's OK to go. Would PTC have prevented the Amtrak accident new Philly? yes. Would PTC have prevented the accident in California where a freight train ran into the rear of a standing passenger train? yes. Will PTC prevent highway / grade crossing accidents or someone walking out in front of a train? no.  Will PTC prevent derailments when a rail turns or breaks? only if it can be detected by the signal system.

My biggest issue with PTC, because I do have skin in the game so to speak is, back in the late 60's the government gave railroads 10 years to remove roof walks from freight cars and they even received a 4 year extension on that dead line, but the government wants to play hard ball with PTC. The railroads were given 7 years to install a system that did not exist in 2008 when the law was passed. In other words, the government gave the railroads 7 years to design, manufacture, test and install one of the most complex computer systems for railroads to date.  Wonder what would have happened if the government told Bill Gates he had to create a perfect operating system in just 7 years? The most frustrating part, some members of government have stated the railroads are putting profits before safety, but the truth is the railroads have already spent billions of profits on installing PTC equipment.         

[davefoxx]

Assuming that PTC and LEADER were to result someday in trains operating without a crew, what would happen in the event of a broken knuckle or pulled drawbar?  I just can't see it being practical to eliminate the crew.  So, if you have to pay a crew, why not keep the engineer?  I'm not sure I would ever trust automatic systems anyway.  I know I would never trust one of those automatically-piloted highway vehicles, e.g., Tesla.

DFF

[Doug G.]

I completely agree, Dave. The trust many place in totally automated systems is much misplaced.

Doug

[Scottl]

I suspect that advanced sensing systems could deal with some of the routine emergency systems.  A dynamic load sensing system could likely identify many maladies. Time will tell,

[C855B]

... broken knuckle or pulled drawbar... if you have to pay a crew, why not keep the engineer? ...

Train separations are already "remotely" detected by the FRED. Head-end may feel the abnormal slack action, but they won't know what's going on until the brake hose separates and puts everything into emergency stop.

The engineer, as I see it, becomes 1/4 of an engineer. And I don't count remote/automated control of trains in the same category with self-driving Google cars. With trains, you're only concerned with 1 axis, either forward or backwards. With road vehicles, two dimensions and ad-hoc interaction with other vehicles in the 2D plane completely change the ballgame.

Personally, I think the automated systems suck a big one. It will put thousands out of work, for one thing. But as a technologist, I see the writing on the wall. What is happening with PTC, etc., is simply the logical progression of the technology.

[davefoxx]

Train separations are already "remotely" detected by the FRED. Head-end may feel the abnormal slack action, but they won't know what's going on until the brake hose separates and puts everything into emergency stop.

That's my point.  Once the train goes into emergency with no crew on board, the train sits (possibly and dangerously blocking highway access) for who-knows-how-long waiting for someone to get to the site, determine the problem, and make the necessary repairs.  As if it doesn't already take a lot of time as it exists to walk the train and fix the problem, wait and see how long it will be when there's no railroad personnel on the train to deal with it.

On that note, what would happen if the rail conditions deteriorate (e.g., wet leaves) or a locomotive in a consist goes down, suddenly requiring a doubling or tripling of a hill?  Can PTC do that? 

DFF

[John]

I think we're on the verge of remote operators monitoring cab video feeds only for exception conditions, while PTC 2.0 (or whatever) handles everything else. A single engineer could probably handle four trains simultaneously.

wanted --- Locomotive engineer -- must have trainsim 2.0 experience on NEC

[John]

you will end up with a 1 man crew .. and video feed from the cab to central operator .. we do it with drones now ..

[Rich_S]

Train separations are already "remotely" detected by the FRED. Head-end may feel the abnormal slack action, but they won't know what's going on until the brake hose separates and puts everything into emergency stop.

That is not 100% true Mike, the EOT can sense movement but it's biggest job is to be able to dump the air remotely from the HOTD (Head of Train Device) which links via radio signals to the End of Train device. Train separations are remotely controlled by George Westinghouse's invention, the Automatic Air brake. Let the air escape too rapidly from the brake pipe and the train goes into emergency.

[Rich_S]

That's my point.  Once the train goes into emergency with no crew on board, the train sits (possibly and dangerously blocking highway access) for who-knows-how-long waiting for someone to get to the site, determine the problem, and make the necessary repairs.  As if it doesn't already take a lot of time as it exists to walk the train and fix the problem, wait and see how long it will be when there's no railroad personnel on the train to deal with it.

On that note, what would happen if the rail conditions deteriorate (e.g., wet leaves) or a locomotive in a consist goes down, suddenly requiring a doubling or tripling of a hill?  Can PTC do that? 

DFF

Dave,
   Even though the technology is getting closer, I do not think for the foreseeable future the railroads will go with no crews or one man crews. Even if they went with a one man crew, you'd still have the same issues as a  train without a crew. The engineer cannot operator the locomotive and re-couple the train in the case of a pull apart or broken knuckle.  SOP now with a broken knuckle is to have the conductor walk the train to find why they went into emergency. Once the conductor finds the location, the conductor has to secure the rear half of the train. Then the conductor would walk to the head end, throw off a spare knuckle on the ground and get permission to pull the train forward. Then the conductor would remove the broken knuckle and renew it with the new one off the locomotive. The conductor would then ride the train back to make the coupling, walk to the rear of the train to perform an air test, then walk back to the head end to continue their journey.