Predicting the future is fraught with difficulties, but Colin Hakeman has provided a convincing narrative. This seems a good article to discuss on the forum. Please note that Colin does post on these forums, so we should probably be polite where we disagree.
I am posting from Australia, and my friends have generally indicated that we wouldn't expect to have train crew on board for a line haul in twenty years' time. This is based on Rio Tinto already running trains without crew on line haul, and to a lesser extent, the Sydney Metro system which operates automatically. there are personnel on board, but the controls are locked up and only used in case of a major failure.
I think that voltages well above 600 volts could be used to directly drive coventional AC traction locomotives. I am pretty sure that 1500 volts DC can be fed directly to the DC bus of an inverter powering AC traction motors. The New South Wales Railways have ordered dual mode diesel and 1500 volts DC vehicles for regional and outer suburban commuter service. These are not currently intended to use batteries, but on a number of significant grades the track is already electrified for shorter distance commuter trains, and a significant fuel saving is expected. It also will effectively eliminate exhaust fumes in the major terminal station in Sydney. Interestingly, the trains were initially ordered as straight diesel-electric with AC traction, and it was realised that the electric mode could be added for not much more than the cost of a pantograph and a high voltage circuit breaker.
The other thing I might expect is that visible signals will be eliminated. Rio Tinto eliminated visible signals before 2004, possibly twenty years ago, and all data was displayed on the screens of the locomotives, including a recommended speed (it showed zero for Stop). That same system is supplying data for the "Autohaul" trains now. However, BHP, who still have manned locomotives, have also eliminated visible signals on the main line, at least five years ago, relying on data displayed on screens in the cab. So the introductory photo in the article might not be appropriate (although I doubt Colin had any say in that).
Colin has assumed ECP braking would be adopted, at least for some traffic, by 2040. In Australia, all four private iron ore roads have adopted ECP braking for all traffic. This would imply that more than two thirds of iron ore traffic in Australia uses ECP braking. About half of the coal traffic in Australia uses ECP braking, and none of this is on isolated private systems. The coal trains are intermixed with intermodal and commuter trains in the Hunter Valley and with passenger and intermodal trains in Queensland, and in both locations, there are older coal trains with conventional Westinghouse brakes. So I find it hard to understand why ECP has not been adopted in the USA, particularly since the equipment used in Australia all meets AAR recomended practices.
And I would hope we don't have to wait twenty years for double track to be extended all the way around the Tehachapi loop, as suggested by the article. Surely UP could double track the "tunnel" under the loop at Waylong in a single weekend...
Peter
Is there any information on the failure rate for Rio Tinto trains? I would guess it's within an "acceptable" limit.
While we don't have the full blown automatic capability, we do have some of the 'auto throttle' feature made by some of the same suppliers to Rio Tinto. I will say it works pretty good, but it does fail on occasion. We also have break in two zones where we have to shut off the automatic control on certain trains because it tears too many apart.
You won't see ECP on trains anytime soon here. Even on the trains that might benefit from it. As long as PSR/share holder value uber alles is the name of the day they won't spend money if they don't have too.
Jeff
I'm disappointed that a second thread on this subject was started by another poster, even if he believed that his ideas were copied. I understood that forum protocol was to use an established thread, even if you are disputing something...
Sadly any discussion of the content appears to have disappeared.
In my opinion, it would be highly logical to discuss the technical specifics in the article in this present thread, and leave the other one to its discussion of ethics and perceived remediation or author credit. As I have not read the Trains article, I can't comment on any specifics other than those already mentioned as present. I say this with no disrespect to Don Oltmann, whose blog entry on this topic remains just as thought-provoking and valid as it was when written.
I also think we can at some point 'compare and contrast' this with other 'visions' including that which Don Oltmann so splendidly wrote up. As some of the differences in technology involve things dear to me, such as incremental implementation of islanded catenary with either full-hybrid or 'dual-mode-lite' power in lieu of full buildout for electric operation, it might be useful to concentrate on discussing the differences rather than the similarities ... here.
I said in the other thread that I thought it would have been better for Trains to publish an article on "Visions of railroading in 2040", not just the one adapted article that is turning out to cause so much fuss. One very obvious alternative would be that put forward by ttrraaffiicc and Bruce Gillings, in which autonomous road chassis increasingly, and then in their view completely, supplant rail freight transportation. That may not be palatable to many reading a rail-enthusiast magazine (just as it was when presented to many posters here) but it is at least thought-provoking, and some of the details will certainly contribute to shaping railroading a couple of decades from now.
There is little question that 'standardizing' catenary transversion (or any other method of energy provision or storage) should be made around properly filtered and spike-protected DC-link voltage for AC synthesis drive and control. To the extent that may differ by locomotive type (or conversion kit) it should be relatively simple to set this, perhaps as a version of multipower (say, 11/12.5/25kV vs. 1500 to 3000VDC) where American practice might call for it.
I also think there are numerous reasons, and a credible technical upgrade path, to adopting ECP braking for large numbers of freights, and ultimately to full interchange compatibility. An intermediate step might be implementation of computer controls for brake systems that make some of the effects of the changeover less dangerous to crews who have become familiar with one or the other system and have to operate using the other 'conventions'. Both major suppliers of ECP equipment have provided versions that are easily convertible from one to the other system in the field; this makes adoption or 'cutover' a relatively simple, and prospectively massively parallel, process; I can't imagine a country-wide 'cutover' to substantial ECP operation involving more time than gauge conversion in the 1880s -- probably even with penny-pinching modern (mis) management
Multiple-tracking many 'bottlenecks', not just Tehachapi, are or ought to be priority projects -- especially for properly-administered government 'improvement' initiatives that get the best 'bang for the buck' even with continued private ROW ownership.
In my opinion it was ridiculous to continue any kind of wayside legacy block signaling in an age of PTC, CBTC, and intelligent (and robust, and redundant) cab-signal display. We have been paying, sometimes in blood, for retaining defective ideas of both block and route wayside indications, while not bothering with the one real place fixed waysides are essential: home and distant indications.
Now we've spent heaven knows how much on replacing signal plant that could have been utilized for, say, ECP transition or locomotive valve replacement to fleet compatibility, and still have accidents related to signal confusion.
Perhaps the most interesting 'timeless topic' is that of autonomous operation. For a wide variety of reasons reached over an extended period of time and thought, I cannot imagine any train not having a manned 'presence' -- but not the engineer. It is perfectly possible to implement effective 'telepresence' even using current PTC bandwidth to allow trains to be operated semi-autonomously, backed up by a suite of AI/ES, under the supervision and where appropriate active continuous control of engineers working, like current dispatchers, out of centralized facilities convenient to where they live. When they're called they easily drive to a known location, and when they mark off they go out and go home -- no more silly vans to the middle of nowhere, no wasted hours to take a train, easy reassignment to other jobs if a given train isn't "ready" -- etc. etc. etc.
The person on board the train for emergent issues, fixing potential mechanical issues, doing manual switch changes, etc. is an appropriately abled "conductor" -- remember 'superconductors'? Those now become flying maintenance squads, but not substitutes for a physical presence in the engine cab at all times.
Certainly until full proportional ECP is reliably present, with some of the worst programming errors taken out, it would be impossible to handle the large mixed or blocked freight consists in modern "PSR" operation fully autonomously, no matter how effective the technology can be made to simulate full proportional operation. There is simply too much momentum, and too many environmental uncertainties, to make that more than an armchair engineer's pipe dream. What could be built instead, though, even by the early 2040s, is highly interesting.
Since the other thread is more about plagiarism, I thought I'd try to revive this thread about what things might be like in 20 years.
I have an issue with about 12 or 13 things in the article. Some you might partially see, others I don't think you'll see at that time.
First there is also a terminology mistake. When you want to control remote DP consists independently (manually) from the lead locomotive, you don't "drop the fence," you put up the fence. Currently you couldn't control all 4 consists independently. You have some choice on what combination you want controlled in sync with the lead and controlled independently. Linking and unlinking DP consists is a little bit more involved, requiring someone to do set up on both consists. I think the almost automatic connecting/disconnecting in the article won't happen. I do think there will be utility employees at yards to help out during these processes.
Currently, your cell phone has to be OFF and stowed in your luggage. Maybe they will relax the regulation, but I doubt it. I'm not sure about the finger print thing. To log into and initialize PTC we have to enter our employee ID number and AVR (automated voice response) password. And I don't expect they'll allow a lone employee to listen to music while toolin' down the rails.
Once with PTC working, you don't see other trains on the operating map display. I doubt in the future this will change. There's no reason for it. While it's always nice to know why you're getting hosed at a control point, it's not critical information the train crew necessarily needs to know. Plus, if you have the PTC screen why do you still have a cab signal? If you have "rolling blocks" (no permanent track circuit block boundries) you don't need waysides or cab signals. (Personally, I don't agree with the need to get rid of fixed blocks and the signalling that goes with them. If PTC fails, and it does and it will at times, you have something to fall back on. Verbal Absolute Blocking is more unsafe than leaving legacy signalling in place. Unless when PTC and your other alphabet soup of control systems fail you plan on letting a train sit until rescue by another working engine comes along.) Also, I think even with a communications based, "rolling block" system there will still be some kind of track circuit blocks. Not for signalling but for broken rail detection.
Speaking of crews. I do think most trains will be single person. They may not be called locomotive engineers anymore. They will normally, outside of terminal areas and when switching, mostly monitor the computer who will be operating the train. It'll be what some call, attended automation. The employee will only operate over the road when the computer, for whatever reason, cuts out. And it at times will. The part where errors, and to me implied failures, will be rare is BS. They may be minor, that may be temporary-only for a portion of a trip, and a single employee may go for weeks or months without personally having a glitch. But it will happen. (That's why earlier I asked about Rio Tinto's failure rate.)
I don't think you'll see all those sensors, and especially, the back up cameras on cars. They are working on some of these things, but I think the vast majority of the car fleet won't have those items yet. I'm not sure about ECP. I think you'll see some of the fleet equipped for dual ECP/Conventional operation. I think 20 years is to soon for the entire fleet to be converted.
I don't think you'll see electrification outside of some of the major metro areas on the (mostly) east and west coasts. I could see a move away from diesel fuel elsewhere, but not widespread electrification. Unless they can get the government to pay for it all. Highly doubtful.
I can see the railroads, the major ones, leasing or selling much of their secondary and branch line networks and becoming mostly line haul companies. With or without forced reciprocal switching. I think yards that remain with class ones may be operated by contract switching companies. This is already being done in a few places. Maintenance of Way and equipment and communications may also see more work being contracted out. There is already a trend to use contract for some MOW and communications project. The major carriers are pushing for more.
Which brings me to something not mentioned in this article. I haven't read Don's original story in quite awhile and don't remember if he touched on it either. The major railroads are heavily unionized. They operate under the Railway Labor Act. One provision, that gives labor what power it still has, is it allows Union Shop agreements. That is, to work for a company that is party to such an agreement, you have to join one of the recognized railroad unions. (You can, at least in Right to Work states, partially opt out of full membership. Only paying dues that go to the maintenance of your working contract and conditions. Such as if you go to a disciplinary investigation, you still are represented by the union and you are paying for that.) There have been attempts to repeal or revise parts of the RLA as part of a national Right to Work law. If this happens I look for the unions to eventually fade away. They will have little negotiating power, if they aren't outright decertified. Many of this future vision will either have to be negotiated if there are unions, or unilaterally imposed if there aren't. I would think for some of these future predictions to happen in the time frame given, it will have to be imposed rather than negotiated.
The biggest hurdle to the future isn't technological or employee related. It's management and the vocal activist short term investors. As long as there are enough well heeled stockholders who are happy making as much money as possible in the short term on stock price and sucking out as much value as possible on the same or declining volumes, and eventually revenues, the future as presented doesn't have a chance.
Sorry for rambling and/or jumping around a bit on items.
jeffhergert (Personally, I don't agree with the need to get rid of fixed blocks and the signalling that goes with them. If PTC fails, and it does and it will at times, you have something to fall back on.
(Personally, I don't agree with the need to get rid of fixed blocks and the signalling that goes with them. If PTC fails, and it does and it will at times, you have something to fall back on.
I'm not convinced that there is an effective replacement for the track circuit.
I didn't think the music-listening was realistic either. Whether through speakers loud enough to drown out the loco noise, or via headphones, it seems like engineers will always need to be able to hear what's going on. Like alarms, things that just don't sound right and need investigation, radio calls, or even yelled emergency warnings from trackside individuals. "Stop! There's a semi stuck on the crossing ahead."
The article is an intesting read, though, even though a lot of the tech stuff goes over my head.
I'll add my two cents as well.. The engineer in Colins story won't be picking up any blocks at Commerce Ramp in 2040, because by then it will no longer exist. California's HSR proposed route runs right through the current facility. BNSF and CHSRA have already came to an agreement that the facility will be eliminated once construction beigns in the LOSSAN corridor. BNSF will be getting a new facility in Colton. Eventually..
This of course assumes that common carrier railroads still exist in 2040 and haven't been completely undercut by autonomous trucks.
ttrraaffiiccThis of course assumes that common carrier railroads still exist in 2040 and haven't been completely undercut by autonomous trucks.
And that the roads the autonomous trucks run on continue to be maintained...
And that said roads can handle the additional traffic.
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...
ttrraaffiicc This of course assumes that common carrier railroads still exist in 2040 and haven't been completely undercut by autonomous trucks.
Maybe there won't be long distance trucking by 2040. It could happen that the fuel/power situation (price/availablity,etc) is such that long distance moves will be by rail. Short distance delivery/pick up by truck.
Even with trucks I think fully autonomous operation won't be realized. I think it will also be "attended automation" for a few reasons.
jeffhergert ttrraaffiicc This of course assumes that common carrier railroads still exist in 2040 and haven't been completely undercut by autonomous trucks. Maybe there won't be long distance trucking by 2040. It could happen that the fuel/power situation (price/availablity,etc) is such that long distance moves will be by rail. Short distance delivery/pick up by truck. Even with trucks I think fully autonomous operation won't be realized. I think it will also be "attended automation" for a few reasons. Jeff
And 2040 is only 20 years away.
Never too old to have a happy childhood!
People usually look at 20 years down the road as a long time away. Yet when you look back 20 years, it seems like it was only yesterday. Perception is a funny thing.
In reality, 20 years from now things may not be too different from the way they are now. Some changes yes. Wide sweeping changes, probably not so much.
jeffhergertAnd I don't expect they'll allow a lone employee to listen to music while toolin' down the rails.
Agree. Not happenin'
jeffhergertOnce with PTC working, you don't see other trains on the operating map display. I doubt in the future this will change. There's no reason for it. While it's always nice to know why you're getting hosed at a control point, it's not critical information the train crew necessarily needs to know
Only if "why" becomes "why not" from a managment perspective. Getting the dispatcher's track line view into the cab display is really easy to do. Can't hurt. Might help. Improves employee moral. You want employees to act like owners, treat them like owners.
jeffhergertPlus, if you have the PTC screen why do you still have a cab signal?
Yep. No cab.
-Don (Random stuff, mostly about trains - what else? http://blerfblog.blogspot.com/)
oltmanndOnly if "why" becomes "why not" from a managment perspective. Getting the dispatcher's track line view into the cab display is really easy to do. Can't hurt. Might help. Improves employee moral. You want employees to act like owners, treat them like owners.
I'd rather have music.
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
jeffhergert (Personally, I don't agree with the need to get rid of fixed blocks and the signalling that goes with them. If PTC fails, and it does and it will at times, you have something to fall back on. Verbal Absolute Blocking is more unsafe than leaving legacy signalling in place. Unless when PTC and your other alphabet soup of control systems fail you plan on letting a train sit until rescue by another working engine comes along.) Also, I think even with a communications based, "rolling block" system there will still be some kind of track circuit blocks. Not for signalling but for broken rail detection.
Broken rail protection for sure. RRs are so wedded to fixed block with PTC built on the back of that - going away by 2040 is unlikely. If huge, one man crew trains are the rule by 2040, then rolling blocks don't buy you much. RRs aren't capacity contrained by signal blocks currently and won't be then.
jeffhergert The employee will only operate over the road when the computer, for whatever reason, cuts out. And it at times will. The part where errors, and to me implied failures, will be rare is BS. They may be minor, that may be temporary-only for a portion of a trip, and a single employee may go for weeks or months without personally having a glitch. But it will happen. (That's why earlier I asked about Rio Tinto's failure rate.)
Agree completely. In fact, I would require the operator to "run manual" for some portion of most trips, to keep sharp.
jeffhergertI don't think you'll see all those sensors, and especially, the back up cameras on cars. They are working on some of these things, but I think the vast majority of the car fleet won't have those items yet. I'm not sure about ECP. I think you'll see some of the fleet equipped for dual ECP/Conventional operation. I think 20 years is to soon for the entire fleet to be converted.
If you do some sort of advanced ECP, then you do the DPU via trainline rather than radio (should make linking, etc, simpler). Having a "brain" and comm on each car makes adding sensors pretty cheap and easy. Hot box, stuck truck, "anti-lock" braking and more. It the one thing the industry really should be getting going on....but isn't.
jeffhergertWhich brings me to something not mentioned in this article. I haven't read Don's original story in quite awhile and don't remember if he touched on it either. The major railroads are heavily unionized. They operate under the Railway Labor Act.
I sorta did. But, rather than being RR related, I had the service and retail industry unionizing and pushing AI and automation ahead faster.
tree68 ttrraaffiicc This of course assumes that common carrier railroads still exist in 2040 and haven't been completely undercut by autonomous trucks. And that the roads the autonomous trucks run on continue to be maintained... And that said roads can handle the additional traffic.
That was completely off my radar in 2014 when I wrote my blog. I'm sort of conflicted about that now. I can't get Alexa to reliably do what I say without phrasing things just right, or sometimes, at all. So, she's not driving me anywhere, ever! I'm thinking automated cars and trucks are a ways off beyond 2040. Maybe auto convoying trucks, though...
If self driving trucks come, self driving trains will be here, too. And, that might just be another sea-change. Instead of few, long trains, maybe short, frequent, origin to destination will be the rule.
zugmann oltmannd Only if "why" becomes "why not" from a managment perspective. Getting the dispatcher's track line view into the cab display is really easy to do. Can't hurt. Might help. Improves employee moral. You want employees to act like owners, treat them like owners. I'd rather have music.
oltmannd Only if "why" becomes "why not" from a managment perspective. Getting the dispatcher's track line view into the cab display is really easy to do. Can't hurt. Might help. Improves employee moral. You want employees to act like owners, treat them like owners.
If you listen to the Engineers, they already know how 'their' train needs to be Dispatched without the need to see the model board with all the other trains on the territory. Facts would only cofuse things.
jeffhergert People usually look at 20 years down the road as a long time away. Yet when you look back 20 years, it seems like it was only yesterday. Perception is a funny thing. In reality, 20 years from now things may not be too different from the way they are now. Some changes yes. Wide sweeping changes, probably not so much. Jeff
Currently, nobody in the industry is picking their head up to look much more than 5 years down the road. I wish someone would just take a swag at 20 years and make sure the current investments (or lack of) don't turf out a viable future.
Sorry, no music for you...
BaltACD zugmann oltmannd Only if "why" becomes "why not" from a managment perspective. Getting the dispatcher's track line view into the cab display is really easy to do. Can't hurt. Might help. Improves employee moral. You want employees to act like owners, treat them like owners. I'd rather have music. If you listen to the Engineers, they already know how 'their' train needs to be Dispatched without the need to see the model board with all the other trains on the territory. Facts would only cofuse things.
I've ridden enough trains and heard enough dispatchers and train crews to know that reality covers a whole range of good, bad and ugly. From "let me help you out" to "shut up and leave me alone". YMMV.
One thing Colin put in was automatic couplers. I thought about it but decided it was a "bridge too far". Much more expensive to implement that ECP since the thing would have to be adapted to the coupler - yoke - draftgear arrangement along with the plumbing for the air and you'd have to have a better alignment/gathering mechanism than currently exists.
You CAN do ECP and have a car be backward compatible, if you cared to. Not so with couplers, I don't think.
Another thing Colin put in is "relability centered maintenance" where you replace parts before they fail. I believe you can replace mechanical parts before they wear out and fail, but you can't do much with electronics. A transistor is as likely to fail on the first time it "switches" as the next. They don't "wear out" like bearings and piston rings.
Generally, you are stuck with replacing most parts when they break and you build in reliablity with redundency.
oltmannd One thing Colin put in was automatic couplers. I thought about it but decided it was a "bridge too far". Much more expensive to implement that ECP since the thing would have to be adapted to the coupler - yoke - draftgear arrangement along with the plumbing for the air and you'd have to have a better alignment/gathering mechanism than currently exists. You CAN do ECP and have a car be backward compatible, if you cared to. Not so with couplers, I don't think.
The (completely) automatic couplers really aren't going to save much. What car load traffic remains will probably be in small blocks. That and intermodal block swapping will only be done at points where you are going to maintain utility employees who will handle the ground work. Assuming that branches and secondary lines are spun off and class one local/yard work is contracted out, the owners/lease operators will probably use two person crews.
The only time a class one train will work between yards will be to set out a bad-order car and there will be a road utility person to help with that. They did some experimenting with automatic air connections (with a knuckle coupler) back in the 1970s. A person still had to manually operate the uncoupling lever to uncouple, but the air connection was automatic. It never really went anywhere and I would guess because they figured it didn't really save them much.
Regarding track circuits for broken rail protection.
I was reading an article in Railway Age about a new CBTC (communications Based Train Control) system. The author, a "contributing editor", was saying how this system was completely independent of track circuits. That not only could it be adopted on railroads without any signalling, exsisting signalling could be removed. He said broken rail protection wasn't really needed in many places.
He was involved in the developement of this system. He gave me the impression that anyone who didn't see that their system was the greatest ever developed was a moron. He didn't care about broken rail protection because their system didn't have it. Reading many articles like this I come to realize that the "contributing editors" actually are trying to sell their product or service.
jeffhergert Regarding track circuits for broken rail protection. I was reading an article in Railway Age about a new CBTC (communications Based Train Control) system. The author, a "contributing editor", was saying how this system was completely independent of track circuits. That not only could it be adopted on railroads without any signalling, exsisting signalling could be removed. He said broken rail protection wasn't really needed in many places. He was involved in the developement of this system. He gave me the impression that anyone who didn't see that their system was the greatest ever developed was a moron. He didn't care about broken rail protection because their system didn't have it. Reading many articles like this I come to realize that the "contributing editors" actually are trying to sell their product or service. Jeff
There lots of stuff in the transit world that doesn't "scale up" very well. Automatic couplers with electrical and air built in are just one thing. I can't see the FRA letting RRs get by without broken rail protection until the number of broken rails drops to very near zero. Some smart C&S guys did tell me that you can to really long "broken rail" blocks using higher frequency AC track circuits or some such.
rdamon Electronics have MTBF (Mean Time Before Failure) values calculated allowing for preventative replacements. These values are determined by the reliability of the individual components of the assembly. The software on these devices also have reliability values assigned. Obviously, there are out of box and other types of failures that require replacement before the time, but this is also true for mechanical parts as well.
We once calculated the relibility curves for locomotive components - they were all mostly flat. Scheduled replacement of critical parts was done by monitoring condition, generally, so we never saw the inflection in the curve from when they quit from wear.
Some limits were set by experimenting. Roller bearings on freight cars don't ever need additional lubrication. The wheel wears out before the grease quits. That wasn't always the case. Seals and grease were improved and wheelsets tested and bearning torn down until it was known that the design was solid.
UP 8517, an SD70ACe. I was the third engineer to have it on it's maiden run out of Chicago. The odometer on the computer screen was about 335 when I got on it. The new, fresh paint smell was almost intoxicating.
When the odometer said about 405, we struck the iceberg. The traction motor on the lead axle froze up. A brand new locomotive and it failed. Who woulda thunk it.
Our community is FREE to join. To participate you must either login or register for an account.