Freight Schedules

Noel Weaver: is the going to schedules connected to crew fatigue problems and morale problems caused by calling crews at all hours with little notice? Is this as much about standardizing individual employee call times as it is about operations efficiency?

While I'm aware that FEC does run, more-or-less, a scheduled railroad, my observations of a local regional railroad that makes a huge effort to schedule its operations (and doesn't have Amtrak to worry about, save for joint operations on the UP and BNSF) shows that it's more difficult than you think.

The FEC has a lot of captive hauls; it controls the movement from point 'A' to point 'B' and everything inbetween. It's essentially a dead-end railroad so there's only significant interchange at one end of the system, and it's pretty easy for them to say "well if CSX doesn't pull their train into our yard on time, we're going to leave, and those cars will just wait until later."

On my regional railroad, if UP doesn't deliver the cars to Eugene Yard, there is no train. So maybe the road will have the crew do "clean-up" work to keep busy. The problem is that those cars are out there and will be handled, just not on schedule.

The next day, the cars show up, PLUS the cars that are supposed to be on the train. You now have a train that is heavier and longer, takes longer to get to Albany, takes longer to sort in Albany, and won't fit in sidings. Congratulations, the crew is dead on the law. You now have to pull a crew off another train, taxi them to the dead train and have them bring the train in. But that other train is tied down on a main track somewhere not going anywhere. Meanwhile, the yard is clogged so at least two or three other trains are now delayed - all because ONE train didn't interchange properly.

It's also very hard to tell a customer "I'm sorry, but your cars didn't make the train on time, so they're going to be delayed by a day."

Back on the Class One, what do you tell a unit grain or coal customer, or a solid TOFC/COFC or auto-rack train when it's time to go and only 3/4ths of the train is loaded? Do you tell the loading crews to stop work, the train has to go?

I'm not saying that all of these matters can't be arranged or done, if handled right. Obiviously FEC and CN is doing something right. But keep in mind that FEC doesn't host Amtrak, and CN hosts, what, one Amtrak train (City of New Orleans)? Maybe those two examples aren't the best to use in this context...

When I sit down at my desk, I have 3 1/2 hours before Amtrak hits-I may have 5 on 8 with two hot shots in the middle and amtrak bringing up the rear. Three who will die on the law if they don't keep moving. And on a single track rr when you tie down it is like a stroke waiting to happen.

By one hour in I have to have a basic plan put together. Too bad I blew it; getting mw out of the way, funding out a crew did a last minute mark off and a train stalled. Manifests end up sitting for an hour or more waiting on the passenger because I can't afford to bring them to because eventually they'll end up blocking crossings.

You guys make it sound easy!

trainwreck wrote:You guys make it sound easy!

It is not easy, and it gets much more difficult when the railroad is congested. Perhaps it would go better though if the railroads highly valued good dispatchers, and if those dispatchers knew every inch of the line they were operating so that they could better anticipate problems, and if they had had a few beers with the guys and gals out on the road so that information came in before it was too late to avoid problems, and if the guys and gals in the dispatcher chair had twenty or thirty years in on the segments they run so that they very seldom came across a problem they had not seen many many times before (IE over the years figured out good work-arounds).

Mr. Trainwreck, I've never been a Train Dispatcher - I've only participated in administering your Agreement on my property. During my three year hitch in LR, I only handled one discipline case involving your craft - that was a 'lay off' on rather short notice and for a RATHER dubious 'reason". Had the Chief not been called for a third trick (he was regularly assigned the "tag end'), the matter may never have seen the light of day.

But your posting, including your slang that possibly "sailed" past some of our younger hobbyists, really made me feel I was sitting in the chair.

Thank you for your submission.

I believe one of the major class 1's offered fixed scheduling for train crews ( i.e set start / finish times , but they voted against it , the majority preferred the overtime payments .
To my mind , if youve got a conjested railroad , then scheduling is one way to alleviate it . One way would be to break a long journey up into sections , relating to the crew districts . I quess this would require extra sidings at these points though .

Speaking for the BNSF where I have worked they turn Amtrak loose with a Z train or two behind it. Oposing traffic goes in the hole. This was when I was working on the Fallbridge sub. Amtrak runs on time pretty much every day on that sub.

Info on computer dispatching systems.

http://findarticles.com/p/articles/mi_m ... i_n8565531
http://query.nytimes.com/gst/fullpage.h ... A96F948260

Another focus has been on development of systems, based on algorithms
optimising the meet/pass plan of a railway line. The objective of the algorithms
are normally to minimise total weighted train delay. A typical optimal CATDsystem
has the following features:
· It serves a line between two major stations or terminals. There are normally
several meetpoints on that line.
· Every train has a schedule for that line. That schedule represents the initial
optimum.
· There is a cost function of tardiness associated to every train.
· It is continuously served with new train positions from the signal systems.
· If there has been a disturbance, it calculates a new optimal meet/pass plan
(schedule) that is presented to the dispatcher, who make the final decision.
A CATD-system that in real-time calculates a new optimal schedule has to be
based on relatively advanced optimisation algorithms

http://www.it.uu.se/research/hci/public ... r-algo.pdf

It would appear that during the course of development the computer dispatching systems have move off of what was anticipated to be a pure flow system into one where schedules are made. My original point (when this thread was part of the Amtrak forum) is that the railroads predetermine the importance of keeping Amtrak on time by how they assign the weight of the cost of late Amtrak trains. If late Amtrak is low cost the computer will hold Amtrak for freight traffic more often.

DutchRailnut wrote:Just remember stopping a 10 000 ton freight train and getting it going again takes way longer and cost way more fuel than putting a passenger train in siding , it would only take that passenger train a mile or two to get back at speed, and all in under a few minutes.

Very true. I would much rather be put in the hole rather than wait on the high iron for the freights to clear up in the siding, especially non-bonded sidings. I can clear up into the siding with a 9 car CZ train within seconds and keep that drag coming at me without losing too much time.

slchub wrote: I would much rather be put in the hole rather than wait on the high iron for the freights to clear up in the siding, especially non-bonded sidings. I can clear up into the siding with a 9 car CZ train within seconds and keep that drag coming at me without losing too much time.

Presumably having the freight put in the hole pre-meet is a better option still. Yes the railroad will eat fuel and labor costs, but such is (or was) the cost of agreeing to become stake holders in Amtrak, of operating Amtrak per contract.

icgsteve wrote:Noel Weaver: is the going to schedules connected to crew fatigue problems and morale problems caused by calling crews at all hours with little notice? Is this as much about standardizing individual employee call times as it is about operations efficiency?

It happens all of the time. I don't think the focus with the Mgt. is crew fatigue and avoiding it. For instance, this morning I was called at midnight for a 0400 show. This interrupted my sleep and I was unable to fall back to sleep. When I asked the crew caller about her giving me a 4 hour call, she said she was very busy calling other crews. To me that is absurd and creates a hazard when I have crew and passengers relying upon me to be alert and operate in a safe manner from point A to point B.

One nice thing about Amtrak and having a "schedule" is the train crews assigned to a job can assume that they will show up at a certain time to start their shift. If the train has fallen behind schedule then Crew Mgt. will call the crew to set them back. The only time this does not work out is for those crew members working the Extra Board.

I really enjoyed the UP implementing "Robo Caller" before I left. Basically Crew Mgt. could set the computer up to call the crew at a certain time. You did not have to deal with an upset and irate crew caller at 0300.

The way I understand it computerized operations (dispatching tied into crew calling) was supposed to allow the computer to prioritize crew calling in part by when they last worked in order to maximize crew rest with-in the overall cost parameters of operating the railroad. Crews would never know what train they would be working because trains were not scheduled, but their call times would conform with the needs of humans. The system would allow for setting the parameters so that to some extent train times could be set upon the needs of the crews for rest ( for instance train XYZ can afford to set two hours so that crew 654 can get two more hours of rest). The IT system has never gotten that far, and two-three years ago there was a huge complaint on a lot of roads by crews about screwy call times, so I figured maybe scheduled train times was intended to be part of the fix. That is the question I have.

FYI: I might have my terms wrong..I take "call time" to be the time the crew needs to report. The actual call I figured is a notification of a call time.

been put in the hole is not so much the problem , its been left in the hole for hours .

For the most part, freight railroads such as CSX (at least the former
Conrail portion) even regular jobs did not have a set reporting time at
either end of the railroad, the call time would depend on when the train
was expected to show up or in case of an originating train when the cars
would be made up and ready and if power was available.
On the Florida East Coast the trains will leave Bowden Yard (Jacksonville)
at their scheduled time and if the cars from either NS or CSX are not in
Bowden in time for a particular train, they would run them on the next
scheduled train. Example is trains leave Bowden for Hialeah at 12:00
Noon, 1:30 PM, 5:00 PM and 9:00 PM and for Fort Lauderdale at 5:30 PM and 10:45 PM. These trains operate at their scheduled time, connections
or not. Out of Hialeah and Fort Lauderdale the northbounds do not wait
for late loads either, the cars have to be loaded on time and the trains
will leave on schedule, if the containers do not make one train, they will
go on the next one. Their customers are aware of this and so are their
connections, this is policy.
Some of the trains including the two that operate out of Fort Lauderdale
meet and the crews swap trains and return to their originating terminals.
The crews that operate from end to end and do not swap get a scheduled
train for a return the next day. I don't think fatigue is a big problem here
when you know just when you will be going to work at all times.
During my last couple of months with Conrail out of Selkirk the railroad
tried to work out scheduled operations out of Selkirk and Buffalo with
scheduled trains and returns. The trains did not get called when they
were supposed to be and crews missed their scheduled returns, outlawed
on the road and had many other problems. The set up lasted about two
weeks and the railroad elected to return to the previous set up before
they went broke trying to do this and completely ran out of men due to
excessive deadheading. For Conrail at least in the summer of 1997, it
was a terrible ideal. I was affected by this scheme and I was tired every
trip especially on the eastbound train back to Selkirk.
Noel Weaver

Personal experience tells me that Amtraks get most of the breaks. Freight trains more likely than not will sit in a siding for an hour or more to avoid delaying a passenger train for a couple of minutes. But, sometimes they do get delayed, and then we get to listen to them whine about getting an approach or having to touch the brake.
Does anyone actually know what the exact incentives are that Amtrak pays for on time arrivals?