Speeding up LD Amtrak Passenger trains cheaply (two questions)

Trains Magazine Passenger
1

Two items I want to discuss here.

First, the use or employment of flag stops on LD Passenger train schedules. A flagstop to me would be defined as a station stop which does not take place unless a passenger is going to board the train (existing reservation) or wishes to depart the train. Computers and technology have come a long way and I think we are in the age of almost real time passenger train manifests since now they scan the Amtrak LD Train Tickets via hand held scanner and my thinking is the result updates almost immediately via cellular signal into the Amtrak HQ Computer system. If this presumption is the case then why can’t Amtrak turn a lot of these small town and small patronage stops into flagstops via their schedule and only halt the train at the station if there is a passenger that wasnt to disembark or a passenger that wants to board. LD trains I thought were reservations only so this should be relatively easy to achieve financially. It would result in slightly faster LD train schedules and save fuel as well. Are there any operational or dispatching issues here as the LD train schedule would be slightly less predictable?

Second, in the former privately run LD passenger train era and long ago in Europe they used to place passengers in specific coach cars based on their destination vs sit anywhere there is a seat. Does Amtrak still do this? In the old days they would only need to open one vestibule at a station stop and only make one stop instead of two at intermediate stations along the route. Seems Amtrak might have abandoned this approach for whatever reason. I really think the two stops one for sleeper and one for coach is rather shortsighted too. Maybe this is done because there isn’t the train staff available to assist the sleeping car passengers with their luggage there once was in the past (not really sure). So OK, keep the two stops one for sleeper and one for coach but does Amtrak batch passengers on the coaches by their destination like the old days or is it general seating still in LD coach?

2

I have ridden coach on the Texas Eagle from Dallas to Temple or Taylor at least six times. On each occasion I was directed to a specific coach. Passengers for Austin, San Marcos and San Antonio were sent to another coach.

Passengers for stations west of San Antonio are assigned to the three times a week through coach that is coupled to the Sunset Limited.

3

The persistent issue with ‘flag stops’ has always been not “departing” passengers (or ‘stops only to receive passengers’ on all-reserved trains). It is that, for the practice to have much value, it involves the train reaching or passing the ‘flag stop’ locations ahead of the scheduled time.

Greyhound has adopted the extremely annoying practice of boarding passengers and luggage as quickly as possible at sequential stops, meaning that they may be departing over half an hour sooner than the time on your reservation for departure. The ‘excuse’ is to tell you to be at the stop at least half an hour before indicated departure time – and if you’re intending to ‘catch the bus’ without a reservation, which seems like a very common reason for having ‘flag stops’ in the first place, you’ll be SOL unless you have the tracker app running and perhaps even if you consult it more regularly than it auto-updates (currently about every 5 minutes, with no allowance for stoppages or other emergent delays).

For some of the longer LD runs, like the Builder, going to extensive ‘flag stops’ for all the tiny towns that essentially depend on the train for “transportation” (which, after all, is the only thing Amtrak thinks they’re responsible to provide) means that God only knows when you need to be ready to board at the station. Perhaps this is a factor in the disappearance of ‘reference’ timetables or even route maps with scheduled arrival times listed.

It was a frequent practice when I rode trains in the Northeast Corridor to have dedicated cars or even dedicated doors for some of the uncommon stops – this might be the case also for ‘dedicated boarding zones’ at smaller stations corresponding to logical sections of the train for those passengers. Two problems immediately come up: first, ‘transportation’ passengers for coach vs. ‘sleeper’ passengers who pay much more for their ‘service’; second, the accommodation (ADA or otherwise) of differently-abled passengers on bilevel equipment without elevators at each car. I would suspect that even one or two stops requiring registering the train with an external bilevel ‘platform elevator’ for wheelchair or mobility-impaired people would ruin the “prospective big savings” in time from adopting the flagstops – this is certainly addressable, and it seems to me without too much objective difficulty, but whether Amdreck SOP would run it correctly, especially in inclement weather, is a different matter.

Certainly if you have a consist that’s going to make multiple stops (for a long consist or whatever), reducing the absolute number of stops may vastly improve both scheduled performance and amount of fuel consumed, even if the train lollygags in typical padded Amtrak style so that it passes each intermediate stop location ‘on or after the advertised’.

4

There is a saying that many in supervision of many aspects of human activity use -

If you aren’t FIVE MINUTES EARLY to a identified time - YOU ARE LATE!

5

This makes no sense. Is this any way to run a bus service? Or a railroad?
I remember reading many years ago that the policy was to depart one minute after the scheduled time and never early. Of course, that was in a book about railroads written over a 120 years ago.

6

All flag stops should be scheduled at the minimum time from the last scheduled stop. Always assume your flag stop is the minimum scheduled from last scheduled stop not any intermediate flag stops.

7

I don’t get why you see this as a problem. When a train is late or off schedule it does not dwell a long time at stations when it is behind schedule this is still true in the Amtrak era. Seems to me with the flagstop implementation at a very min, a late train could make up time a lot faster. Additionally, with computer technology being what it is today on Amtrak LD trains the onboard crew should be able to tell who is scheduled to be picked up at what station stop while the train is traversing down the line in almost realtime. If Amtrak does not have that onboard ability, adding it would not be hugely expensive in most cases. So really they would only need to adhere to schedule for the major cities. In the case they miss a passenger, which would be rare. They could contract with the same train crew taxi service the class one railroads use for frieght crews. How hard would it be to reivise those contracts for Amtrak to use? In fact might not be a bad idea to use those services where cutting a stop off the schedule to make up time for one passenger or just 2-3…train taxi might make more sense service wise there to get the train back on schedule for the majority of passengers.

“The needs of the many, outwiegh the needs of the few” …which I think could be a customer service model Amtrak could adapt from time to time.

8

We are explicitly not talking about ‘late trains’ – by definition a late train, whether flagstopping or not, will be arriving ‘past the scheduled time’ so nobody expecting it will miss it.

The problem for Amtrak is that a train that is more than a few minutes late in the ‘slot’ provided in PSR traffic management is likely going to lose its ability to run 79mph or more, and I suspect we can immediately determine its subsequent ‘precision scheduled’ arrival at all the coming stops just as if it were another PSR fleeting train… again, that’s not the problem being discussed here. What might be considered is whether (as is done by many local bus services) the train makes periodic actual ‘stops’ that are intentionally prolonged until ‘back on schedule’ (as well as dawdling to the padded schedule passing the various stop locations). (Such extended stops might be useful ad hoc opportunities for ‘ghost kitchen’ meal deliveries…)

Part of the reason I point this out is that the original thread title is “Speeding up” LD trains – and the actual ‘speeding up’ cannot and perhaps should not involve any actual speeding up to ‘get ahead of the schedule’ so that any subsequent delays have less impact on passengers’ timely arrival. I am course always willing to have a FYIGM approach on this question, and there are places where it gets applied (the NYC ex-IRT 2 and 3 expresses run that way independent of coordination with the 1 locals, and I believe some of the Caltrain ‘expresses’ will do the same) – but the specific context here is the longest LD operations, and their specific purpose as the ‘only’ available “transportation” for large numbers of stops.

9

I think your being a little too rigid.

Just to be clearer, the intent of the thread was to apply to all overnight or longer duration LD passenger trains and not sub-categorize them. There is usually only one LD train per route each way. It might be just coincidence that I never encounter these over regimented PSR dispatchers you might have encountered on your trips possibly? I don’t think I have been on an Amtrak train yet that was restricted to less than 79 mph because it missed a slot. If it happens it is for an hour or two because perhaps another train was allowed to proceed and then the train recovers or passes the train in front of it …it is not usually for the remainder of the route.

I might not have been clear before in my past responses but on a train traveling from say Dallas to St. Louis, you can arrive on time in Little Rock (I don’t think many care at Little Rock due to the time the train arrives/departs) and you can be on time in St. Louis but you can be late at the intermediate points in between. The flag stop approach would be to recover the lost time without exceeding the speed limit between those two points by skipping intermediate points where there might not be passengers. I also might point out the number of scheduled stops on a LD train route where there is little or no patronage day to day. So for example my favorite example on the Texas Eagle route is Hope, Arkansas (which I would refer to as a political stop more than a practical stop). I think it was 1200 passengers a year or something miniscule but it does not even come out to 1 passenger a train or I seem to hit that station when nobody gets on or off the train.

Really do not see the issue with arriving at a train station early with passengers waiting if you know shortly after they board you have all the passengers on board, what is exactly stopping the train from departing that train station earlier than scheduled? As I stated earlier I’m pretty confident the technology exists for this to happen today without inconviencing anyone. So what exactly is the issue here with departing earlier than scheduled? You cannot presume it will throw off PSR because in most cases the routes traveled by Amtrak are not high density freight lines the whole LD route 24 by 7. In order to throw off PSR the track capacity has to be used up or fairly full of freight trains. I noticed on UP on high density areas they are not so high density during some days of the year like say…holidays or times of economic slow down. So even in highly traffic density areas the flagstop approach would probably not even impact PSR during some days of the calendar.

I see the increased use of flagstops as increasing flexibility of getting the Amtrak train over the railroad route. I am still not seeing where exactly it is decreasing flexibility or causing a huge problem, even with PSR.

10

I believe this was in the era (on the DB at least) of the Kurswagen, coaches added and cut off trains to go to alternate destinations. Most/all trains have had automatically opening or “Press green button to open” doors in all coaches for years.

11

Why not require passengers who wish to board at a flag stop to have a ticketed reservation in place? No-shows would be charged the fare

12

Here in the U.K. passengers for Oban are in the rear carriages, those for Fort William/Mallaig sit in the front portion. Until recently the same applied to passengers going to Thurso or Wick, Scotland.

David

13

One thing that must be remembered - when setting passenger train schedule (at least when I was working) the schedule time between O & D will not be the best actual running time between those points.

On single track lines time will be inserted in the schedule to allow for ‘scheduled’ train meets between trains - both passenger and freight that are anticipated to happen.

On multiple track territory additional time is added to the schedule to allow for trains to crossover between tracks (to my knowledge no US carrier is using crossovers that allow their use at 79 MPH.

All trains will operate at maximum track speed for their class of service unless otherwise restricted by signal indication, slow order or local statutory requirements.

14

I think your basically saying they pad the schedule to account for routing flexibility. I think that is OK and I only have an issue with the massive padding of the schedule Amtrak does with LD trains which I think is way over the line in most cases and reflects a basic laziness by Amtrak management on addressing issues with each LD train vs just sweeping them under the rug via adding more and more padding to the schedule of each train until finally it is almost impossible tor the train to ever be really late. The traveling public ends up holding the bag with the Amtrak approach in place now.

15

Just to be clearer, the intent of the thread was to apply to all overnight or longer duration LD passenger trains and not sub-categorize them. There is usually only one LD train per route each way.

The situation would be common to any LD service with multiple trains a day; many routes with ‘one train’ actually have to account for two if there is bidirectional service on the same segment the same day, and the need to establish a ‘window’ for each of them to run at passenger speed is there regardless of direction.

It might be just coincidence that I never encounter these over regimented PSR dispatchers you might have encountered on your trips possibly? I don’t think I have been on an Amtrak train yet that was restricted to less than 79 mph because it missed a slot.

I’m going specifically by some of the material in the docket for the suits against carrier railroads for ‘delaying’ Amtrak service. Whether an Amtrak train is ‘held’ for opposing slow trains or to open up a ‘higher-speed’ window for it to move is immaterial to overall train delay, and obviously that’s a frequently-expressed concern.

I also go by operations as I saw them through Douglasville, Georgia around the turn of the century: traffic would be held in both directions until Amtrak came through (at track speed of ~79mph) and then trains would be ‘fleeted’ behind it at what would be expedient freight speed; I think the most I saw was five, on no more than 5 to 10 minutes’ headway. This would seem logical for ‘real’ PSR where train movements are conducted sensibly to optimize fluidity on a single-track line (and there is adequate facility capacity to yard the following trains).

It also demonstrated (to me) an understanding that Amtrak trains were ‘mandated’ to run both with no delay due to facing trains, and with no speed restrictions due to following slower traffic. To accomplish that, dispatch would need to keep a slot ‘reserved’ and be cognizant of how train length in both directions might affect ability of freights to take siding. It seems to me that true PSR would be able to ‘know’ the standing and stretched train lengths to facilitate the process, and I think it appears from the Justice Department proceedings that it is by far the easiest response affected carrier railroads might engage in.

If it happens it is for an hour or two because perhaps another train was allowed to proceed and then the train recovers or passes the train in front of it …it is not usually for the remainder of the route.

“Yes, but…” it appears from the record that there may be multiple trains needing to pass, or trains that won’t fit in unimproved sidings (note that this not only implies Amtrak’s “shorter” consist needing to take siding, but also precludes ability to run around slower trains).

I might not have been clear before in my past responses but on a train traveling from say Dallas to St. Louis, you can arrive on time in Little Rock (I don’t think many care at Little Rock due to the time the train arrives/departs) and you can be on time in St. Louis but you can be late at the intermediate points in between.

I think that is more due to including fixed dwell at major stops (this used to be due to baggage and mail/express handling) I am sure Amtrak currently treats much of the dwell as 'padding. The implication that has to be made here is – if the intermediate passengers don’t care about arrival promptness, as you say – the train simply be allowed to proceed ‘late’ so long as it reasonably deprts the major stop On Time.

The argument for flag stops reducing operating cost remains, and I think it may be substantial. Something in another thread indicated the cost of an older MKT train as involving from 75 to 150 gallons of fuel oil, and corresponding water, for each stop and re-acceleration. Obviously this would be less extreme for modern diesel-electric locomotives, but it stands to reason that on particular sections, and in bad weather, the cumulative cost would be meaningful.

The flag stop approach would be to recover the lost time without exceeding the speed limit between those two points by skipping intermediate points where there might not be passengers.

IIt will certainly allow makeup of ‘lost time’, as I said; the issue is and was only whether one or more ‘arrivals’ at a flagstop location would be ‘in advance of the advertised’.

I also might point out the number of scheduled stops on a LD train route where there is little or no patronage day to day. So for example my favorite example on the Texas Eagle route is Hope, Arkansas (which I would refer to as a political stop more than a practical stop). I think it was 1200 passengers a year or something miniscule but it does not even come out to 1 passenger a train or I seem to hit that station when nobody gets on or off the train.

Hope (which I think is starting to lose the political importance it had in the ‘Clinton years’) is a perfectly good example. This is on the high-speed ex-MoPac line, and there’s no point in stopping there if there is no one to detrain or get on. On the other hand, if there IS someone – even the likely sort of activist that will come out of the woodwork to make the train stop more often than objective traffic would warrant – who wants to use the station, you have to allow for a stop ‘and all that that implies’. And you can’t pass Hope early… there might be someone expecting they can just arrive and get on.

R

eally do not see the issue with arriving at a train station early with passengers waiting if you know shortly after they board you have all the passengers on board, what is exactly stopping the train from departing that train station earlier than scheduled?

That’s a fair point, and in a world where you MUST have paid for your ‘transportation’ before boarding (iow no ticket sales on the train, even at additional cost) you could easily determine running schedule from ongoing ticket sales with only a couple of minutes’ latency. On the other hand, you now introduce the question about people who have confirmed reservations, but are not at the station by the time the train actually gets there. They may very well be assuming that because they have their documentation ready, thay can get there only a minute or two before train time and just clamber aboard.

You cannot presume it will throw off PSR because in most cases the routes traveled by Amtrak are not high density freight lines the whole LD route 24 by 7. In order to throw off PSR the track capacity has to be used up or fairly full of freight trains.

Note that with various kinds of fake PSR, everything gets optimized around the cheapest possible operating paradigm. If Amtrak trains could be restricted to notch-5-restricted speed and CBTC were properly implemented (which it surely should have been as part of implementing the PTC mandate) there would be little issue; in fact it would be possible to fleet one or more priority freights close behind a given Amtrak train and have them use the following interval to minimize cost and dwell during required flag stops. The problem with that is the same as on any one-speed railroad… and to get something faster to run around your scheduling will be best achieved if you reserve the appropriate track time, and then stretch that ‘window’ as Amtrak may be late entering it.

I noticed on UP on high density areas they are not so high density during some days of the year like say…holidays or times of economic slow down. So even in highly traffic density areas the flagstop approach would probably not even impact PSR during some days of the calendar.

Where there is excess track capacity, the ability for Amtrak trains to get off a given segment of track ‘more quickly’ is a bit circumstantial. It is nonetheless an important consideration even for fake-PSR-crazed railroads. The argument here is entirely one of perceived customer service, and indeed Amtrak (which has fuggled so much else in customer relations!) need do no more than Greyhound in “solving” it by (1) requiring advance ticketing as you indicated, and (2) requiring riders to be physically at the station and ‘checked in’ by half an hour (or more) before scheduled train time. Were I running Amtrak, I might in fact implement such a policy just for the CYA it affords against people complaining about missed trains (I suspect, as you do, that there are actually not very many).

There is the further implication (in your method) that the passengers detraining at the ‘later’ flagstops would see shorter and shorter travel times, and this implies that even operation well ahead of ‘timetable’ would not inconvenience them. A version of this was in the aforementioned name trains that stopped ‘only to receive’ or ‘discharge’ passengers at secondary or suburban locations like Englewood, IL or Rosemont PA…

I see the increased use of flagstops as increasing flexibility of getting the Amtrak train over the railroad route. I am still not seeing where exactly it is decreasing flexibility or causing a huge problem, even with PSR.

Succinctly: absent a policy like Greyhound’s (which I believe you yourself have criticized) you cannot run a train ‘early’ past any given flagstop location without potentially causing some person to miss the train. (And given that Amtrak considers its sole purpose to provide ‘transportation’ – not on-time or rapid transportation – that possibility alone might be anathema.) That is really the only functional objection to the idea, and in the great majority of LD cases I think it would be benefitial, if not actually advisable, to make the necessary operating changes to permit it.

16

Dispatchers have to deal with two realities - the territory that they dispatch and the trains that are given to them to dispatch.

Many think it is not possible to ‘overload’ a multiple track territory - THEY ARE WRONG.

Dispatchers are trying to move all traffic in the most expeditious manner possible, given the time and power constraints that each train on their territory has - both passenger and freight.

1 Like
17

Thanks for the responses. On the above point. I thought all Amtrak LD trains require a reservation still? Even if they are not all they need to do is ban initial onboard ticket sales (upgrades still OK) here and the above issue is taken care of. Change the reservation system to forbid ticket sales 30 min before scheduled arrival (that might be expensive but might already be in the system and be a minor code change). So if you buy a ticket the trains conductor should be updated as to where to stop along the route. That sounds high tech and expensive but it is not the Feds use cellular monitoring like this in other areas of commerce (which I actually know at least one application technically…so I am not pulling that out of thin air). Not a big expense, in my view to implement if not already implemented via the reserved trains concept.

18

One way to speed up passenger trains (is what happens here in the U.K.) is reduce the number of station stops.
Other trains stop at all stations on a much slower timetable; thereby not arriving late?
(So called progress.)

David

19

Thats a good idea but with Amtrak Long Distance we have sometimes what are known as political stops. Someone wanted the train to stop but the annual patronage really did not support the stop so they got the stop added via political means. Part of the problem as well with Amtrak LD is they seem to have no real lower threshold of annual patronage for a station stop to continue to exist. So getting a stop removed once it is added…almost never happens. What should happen and this was pointed out by MN DOT while the Borealis ridership was being studied. Amtrak should take US Census tract demographics and apply to station location to see if the markets being served have the best station locations. In the case of the Borealis MN DOT pointed out that a suburban stop West of Milwaukee (suggested Pewaukee, WI) could boost ridership as the Milwaukee Airport station has (primarily South Milwaukee riders). Amtrak and WisDOT have so far not acted on that suggestion for Wisconsin. Amtrak really should do that all over their system but as with most items that make business sense. Someone needs to hand hold Amtrak Managerment through suggestion implementation.

BTW, this process is done annually for US Military Recruiting stations to see if they are optimally placed demographically for the 18-27 year old age group. The process of realignment was called draw down once upon a time. They open and close military recruiting stations based on population shifts and preferences. I don’t think Amtrak has ever done it system wide and relies on the ridership or surrounding communities to formally ask Amtrak for a station (ridiculous, but that is Amtrak).

20

True, but I think that is the price to be paid to get congressional support from the rural states, without which there might not be any LD trains at all.

1 Like