Paul_D_North_Jr Railway Man In theory, is is possible to automate operation of trains. In practice, it is not even remotely feasible at this time. Nor any time in the near future. Nor in my lifetime. This is much harder than many people think. The operation of one train is easy. Now do it with thousands of trains, in a dynamic system that evolves from one day to the next. The systems management requirements to take in all that data, process it, categorize it, and publish it is quite difficult and costly. Even if it was, I cannot imagine any regulatory environment in this country that would permit such a thing. RWM Agreed, including your subsequent post excluding 'closed' systems such as Pilbara. But what about remote control as discussed above, and with the 'drones' ? Do you care to offer a prediction - which you are typically reluctant to do, I know - or any insight into that ? Thanks. - Paul North.
Railway Man In theory, is is possible to automate operation of trains. In practice, it is not even remotely feasible at this time. Nor any time in the near future. Nor in my lifetime. This is much harder than many people think. The operation of one train is easy. Now do it with thousands of trains, in a dynamic system that evolves from one day to the next. The systems management requirements to take in all that data, process it, categorize it, and publish it is quite difficult and costly. Even if it was, I cannot imagine any regulatory environment in this country that would permit such a thing. RWM
In practice, it is not even remotely feasible at this time. Nor any time in the near future. Nor in my lifetime. This is much harder than many people think. The operation of one train is easy. Now do it with thousands of trains, in a dynamic system that evolves from one day to the next. The systems management requirements to take in all that data, process it, categorize it, and publish it is quite difficult and costly.
Even if it was, I cannot imagine any regulatory environment in this country that would permit such a thing.
RWM
Agreed, including your subsequent post excluding 'closed' systems such as Pilbara.
But what about remote control as discussed above, and with the 'drones' ? Do you care to offer a prediction - which you are typically reluctant to do, I know - or any insight into that ? Thanks.
- Paul North.
I can't see "remote control" aka drones having a signficant cost or efficiency advantage, and maybe a disadvantage. Nor can I imagine it being any more acceptable from a regulatory standpoint than automated control.
I am not an airline pilot but I follow that business for a hobby (since basically I live on an airplane), and I fail to see from all my reading in the business and technology that it is one millimeter closer to automated operation than railroads. The pilots I know sneer at the media claim that a plane can fly and land itself, and invite anyone to try -- see "Ask the Pilot" in Salon.
Sure, like a train, it is hypothetically possible but one would have to create a many-billions-of-dollar infrastructure to enable it, just like railroads.
When people hear the railroads talking about using remote control outside of yards, I think they get the wrong idea. I could be wrong, and frequently am, but I don't think they are talking about operating trains from a central office (Omaha, Ft Worth, Bombay, etc) by remote control. I think the RRs are talking about using the belt pack when doing pickups/setouts at intermediate points. If they could do that, it would be one more argument for a one person crew.
Switching with RCs in yard areas with better controlled access is one thing. Even then, you still see people trespassing at times. Switching cars in or out of your train in a small town where the main track bisects the town is another. In my town for example, we have a few streets that dead end at the tracks. At one time some had crossings that have since been closed off. Still people walk across to get from one side to the other. Even at open crossings if you have 4 quadrant gates you can't stop people on foot. Sure, an engineer may not see someone in time to be able to avoid them. Still, I think the general public would feel "safer" with the thought of someone on the point being able to, if not avoid an accident, respond to an accident quicker.
Someone posted about help being minutes away when a train has problems on the road. If you're in the right place, yes help may be minutes away. That's assuming the help isn't off helping someone else. Most of the time you are on your own. For example, the conductor who in one of those blizzards we had around Christmas who had to walk his (key) train 4 times.
I think you will always see at least one person on board. The time may come when most of the operation will be automated, but there will still be someone on board with enough training to limp the thing into the next terminal when the automation fails. Also, when the cost of tying up a main line gets into five and six figures per hour, you'll want someone who can change out an air hose, replace a knuckle etc ASAP. You won't want to wait for the repair truck who's 50 miles away working on another train.
Jeff
Railway Man I can't see "remote control" aka drones having a signficant cost or efficiency advantage, and maybe a disadvantage. Nor can I imagine it being any more acceptable from a regulatory standpoint than automated control. [snip] RWM
[snip]
OK, RWM, thanks for your response. To me, it seems like it would be similar to RCO - just from much farther away ! As mentioned above, for over-the-road trains in remote areas with no station work, I could see advantages and savings - mainly from eliminating actual crew changes and Hours of Service issues, and lost/ unproductive time when delayed or waiting in sidings, etc. The completely automated operation - no, that's just 'computerized dispatching' writ much larger, and the impracticality of that has been thoroughly discussed here before, mainly by you, too.
But Jeff Hergert's last 3 paragraphs above describe the practical challenges quite well. And as long as a human is on the train, then crew changes will still be needed, so that advantage from the remote control would be largely negated anyway.
Thanks again for your thoughts.
From the social aspect, I think you'll find that automated trains without an engineer will become commonplace, as soon as the public accepts pilotless airplanes and driverless semi trucks on the Interstate. Given that most folks have serious reservations about using the self checkout lanes at Walmart, I don't see this happening in my lifetime.
Thanks to Chris / CopCarSS for my avatar.
The main problem is that mainline freight RRing is not an exact science. Not like a closed subway system (and we all remember how great that automated Washington DC Metro system performed about a year ago).
There hardly is a trip that goes by where something stupid doesn't happen. Power fails, hotboxes, dragging equipment, kickers, signal issues, the list is endless. Let's take an example: a drag freight stalls on wet rail on a single-tacked portion of a busy mainline. Now we have 3 intermodal trains with guaranteed times that can't get through. With a real human-run operation, we may tell another nearby freight to tie down its train and run their power up there to assist. now, can a robo-train do that? Nothing like paying UPS for the privledge to deliver their trains.
It'snot like we're running trains in a giant loop around a Christmas train. There's almost always work or a problem to do enroute. So while it may be able to work a certain percentage of the time, that percentage that it won't work is what will bankrupt you. And it won't work until they can develop a program that is as sensitive to a train's operations as an engineer's buttcheeks.
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
zugmann [snip] There hardly is a trip that goes by where something stupid doesn't happen. Power fails, hotboxes, dragging equipment, kickers, signal issues, the list is endless. [snip] And it won't work until they can develop a program that is as sensitive to a train's operations as an engineer's buttcheeks.
And it won't work until they can develop a program that is as sensitive to a train's operations as an engineer's buttcheeks.
Zug, we even share the same terminology for such occurences - although my list usually starts with grade crossing collisions, weather damage, trains ahead delayed and held up or out . . .
And as I said before - ''Feel that slack runi in and out . . . ''
Murphy SidingGiven that most folks have serious reservations about using the self checkout lanes at Walmart,
spokyone Murphy SidingGiven that most folks have serious reservations about using the self checkout lanes at Walmart, Good one!
Three times in the last seven years I used the self-checkout. (Two of those times the staff basicaly shanghied me into doing it.) At Target, the card didn't swipe properly. At Walmart super center, several bar codes didn't deploy (and I had worked as a checked at the competition, Target, for a while prior and know how to wave the wand over all six sides of a rectangular box). More recently, at Menard's, I had the gall to pay cash for an item under $5.00. The register came up shy by a penny and neither it not the staff would let me go OR keep the durn one cent, which is what I wanted to do. (I imagine the staff were under some manager's mandate to get close to a hundred percent success rate.)
Like so many things digital (the new TV signals come to mind), if you don't get it perfectly you get it not at all.
And I guess it's human to fudge the stats. Anyone rememer the late Robt. McNamara's "body counts"?? - al
Think beyond the train, Paul, and you'll begin to discern the size of the problem. How do you manage the system configuration to deal with:
And it goes on, and on, and on. All of these things are managed on the real railroad by compartmentalizing the operating functions down to the level of the yard, the train, the train crew, the dispatcher, the roadmaster, the track inspector, the signal maintainer, the clerk. Now you are asking to compile all of these into a single massive piece of software that can look eagle-eye at the whole railroad (and maybe the interchanging railroads too) 24-7-365 and know of all changes and all conditions, all in real time, all at the moment it occurs, not a moment too soon or a moment too late, update itself, and publish the information to all of the right trains at the right time. Very hard.
In terms of mathematical modelling, a huge number of random variates, including the problem of people with nefarious intentions 'hijacking' the control frequencies, or hacking coded programs as it were. Maybe that's easily dealt with...?
-Crandell
The people with nefarious intentions will be background noise among the effects of the incompetent, careless, and uninformed.
Railway Man The people with nefarious intentions will be background noise among the effects of the incompetent, careless, and uninformed. RWM
I think that would be more likely than the nefarious.
Remember the one about our becoming a paperless society because of the computer?
IIRC there was a series of blogs a while back from someone who actually collected all kinds of computer/technology related goof ups, glitches and fatal errors that occured. He even started a series of columns in some computer geek journals regarding EM effects of solar storms/flares---a lot of them too
I'm just wishing I could remember his name----
Any argument carried far enough will end up in Semantics--Hartz's law of rhetoric Emerald. Leemer and Southern The route of the Sceptre Express Barry
I just started my blog site...more stuff to come...
http://modeltrainswithmusic.blogspot.ca/
I just think that there is something ethically wrong about trying to completely eliminate human beings from all work.
Why are we reaching this point?
Erie Lackawanna I just think that there is something ethically wrong about trying to completely eliminate human beings from all work. Why are we reaching this point?
It seems the mentality is that humans cost money---and we want to save the dollar bill
The Beaverton, Fanno Creek & Bull Mountain Railroad
"Ruby Line Service"
KBCpresidentI know this might be a little too philosophical for this hour, but if we computerize too many jobs, (which I doubt we will, ) we will effectively put everyone out of jobs, and launch ourselves into the worst recession in recorded history This would lead to the de-computerizing of jobs. Anyone care to comment on that one?
There are an awful lot of jobs that simply can't be done by a computer - there are just too many variables involved, as the preceding discussion has so amply pointed out. Computers may help us do things better (digging a ditch, f'rinstance), but there is still human input needed to accomplish the task.
IIRC, it's been found that instead of reducing the number of jobs, automation has often simply re-distributed them.
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...
Given the social aspectds of operatorless most anything, I think you've nailed it. Checkerless check outs are an oxymoron, they usually have an attendant because the public actions from a computer standpoint are too undisciplined.
As for drone vehicles on streets and highways, one only has to wastch the TV program "Wrecked" to get an idea of the daily mayhem in one city; one that has human piloted vehicles on the streets and highways.
A previous boss of mine who was a Traffic Engineer used to remark from time to time(paraphrased),"... His plans would work perfectly, if you could remove the human element from them..."
In the aircraft business there are jokes about the "glass cockpits" and the need for a three individual crew...Pilot,Copilot, and a big mean dog...
Until we run our trains on guarded corridors, there will be problems. Even with 'closed ROW's' there will be problems with trespassers.
IIRC wasnt there a song in the late 1960's from Zager and Evans called "(In The) Year 2525"?
There have always been dreamers along the line of technocrati who thought one could replace all human labour with machines. But I suspect there will only be redistribution even then---
Railway Man Think beyond the train, Paul, and you'll begin to discern the size of the problem. How do you manage the system configuration to deal with: changed superelevation in a curve a new turnout that's been cut in a signal that's been respaced a new private grade-crossing that's been opened a shipper that likes to sometimes park his trucks too close to the main track a parade in a small town that has the crowd streaming back and forth across the main track a new freight car type open-top loads And it goes on, and on, and on. All of these things are managed on the real railroad by compartmentalizing the operating functions down to the level of the yard, the train, the train crew, the dispatcher, the roadmaster, the track inspector, the signal maintainer, the clerk. Now you are asking to compile all of these into a single massive piece of software that can look eagle-eye at the whole railroad (and maybe the interchanging railroads too) 24-7-365 and know of all changes and all conditions, all in real time, all at the moment it occurs, not a moment too soon or a moment too late, update itself, and publish the information to all of the right trains at the right time. Very hard. RWM
A. Well, it seems I've been less than clear that I'm not advocating total automation = computer control of trains, with no human intervention. All of the above are the kinds of things that would preclude a high degree of confidence in it, or make the management of the database cumbersome and expensive. The first 4 are 'structural' = more or less permanent, and will also cause problems for a Positive Train Control database. I know you're involved with implementing that, but the points are valid anyway. The last 4 examples are more transient phenomena, which 'fixed' systems like computers and databases don't respond to quickly real well.
B. Nevertheless, I'll accept the misunderstanding for the moment because it led to the 'compartmentalization' analysis/ remark as the evolved organization's systematic way of dealing with / defending against these challenges. That's an interesting and useful insight - I'd mainly thought of specialization of labor as a means of obtaining best efficiency in a technical skill. But here, it is being used as a sensory organ instead - just like the ears will hear something the eyes can't see, the track guys will see something that the operating people won't notice or care about, etc. By bringing to bear each department's core competencies and following advantages and concerns, such changes are likely to be better scutinized than if they were just entered into the system by rote.
C. Adopting and following Larry's /tree68 3 -prong test of social, economic, and technical acceptability of these kinds of changes, and applying it to the proposal for remote control 'drone' operation, which I beleive may have some merit instead:
1. Social - I'm not going to address that here and now;
2. Technical - with next to no technical background to support this opinion: I'm inclined to believe that this is 'do-able' today - though at a price of probably several million dollars per unit for the cameras, ultra-reliable broadband Internet connections to carry the video feeds and commands, etc.
3. Economic - I'm wondering how many 'train-starts' or 'crew-starts' occur daily on any large railroad system or across the US, where it can be confidently predicted or known that almost all of the time there will be no need for a crewman on the train to do anything other than perform the engineer's function = ''just drive the train'', from the departure tracks of one yard to the receiving tracks of the next. In other words, no switching at all, no lining of turnouts into and out of sidings, high-quality equipment such that the probability of a brake or train line problem is nil, etc. I suspect that such circumstances are less than 10 or 20 per cent of the crew-starts - so there's not that much of a magnitude of possible significant monetary savings through crew size reduction to a 1-man remote control engineer, without all the crew transportation and lodging expenses, etc.
In conclusion, I'd say that while remote control may be technically feasible and economically worthwhile on the 'closed' operations as RWM summarized further above - I don't think it will be able to pay for itself on most Class I operations. It's cheaper for the railroad to buy the services of the more-adaptable human computer. And the social acceptability issues are probably enough to preclude it from happening anyhow.
An high-class version of 'Never attribute to malice that which can be explained adequately by incompetence.''
Reminds me of the Dilbert 'decisions made by drunken lemurs' cartoon that was referenced here a few months ago. The upshot was that it's the people who don't know what they're doing that create the overwork for those who do.
Someday I'm going to start a thread of the ''Best / Funniest/ Most Trenchant Quotations'' here on the Forum. It will be hard to keep RWM from dominating it - but this one is a definite 'keeper' (along with a certain book ''might be worth more if someone crayoned on the pages'').
EDIT: For the Dilbert cartoon itself, see -
http://faierie.blogspot.com/2007/12/drunken-lemurs.html
"Just in case the image doesn't work... Dilbert is standing by a garbage man and asks him "Why does it seem as if most of the decisions in my workplace are made by a bunch of drunken lemurs?", to which the man replies "Decisions are made by the people who have the time, not people who have the talent". The final scene has Dilbert asking "Why are talented people so busy?" and gets the reply "They're fixing the problems made by people who have the time." "
Paul_D_North_JrAn high-class version of 'Never attribute to malice that which can be explained adequately by incompetence
----along the lines of that comes one from a book H.P.Lovecraft wrote--"The Case of Charles Dexter Ward":
---Do not call up that which you cannot put down---
Railway ManThink beyond the train, Paul, and you'll begin to discern the size of the problem. How do you manage the system configuration to deal with: changed superelevation in a curve a new turnout that's been cut in a signal that's been respaced a new private grade-crossing that's been opened a shipper that likes to sometimes park his trucks too close to the main track a parade in a small town that has the crowd streaming back and forth across the main track a new freight car type open-top loads And it goes on, and on, and on. All of these things are managed on the real railroad by compartmentalizing the operating functions down to the level of the yard, the train, the train crew, the dispatcher, the roadmaster, the track inspector, the signal maintainer, the clerk. Now you are asking to compile all of these into a single massive piece of software that can look eagle-eye at the whole railroad (and maybe the interchanging railroads too) 24-7-365 and know of all changes and all conditions, all in real time, all at the moment it occurs, not a moment too soon or a moment too late, update itself, and publish the information to all of the right trains at the right time. Very hard. RWM
-Don (Random stuff, mostly about trains - what else? http://blerfblog.blogspot.com/)
Erie LackawannaI just think that there is something ethically wrong about trying to completely eliminate human beings from all work.Why are we reaching this point?
Our community is FREE to join. To participate you must either login or register for an account.