Amtrak Crew Practices / Freight Crews

[Tadman]

Allow me to interject with something that I've brought up before.

There is a reason that you see freight operators sliding out of passenger service contracts. I'm not sure if you realize it (even though I've brought it up before)but freight crews can just "operate" passenger trains unless it is an en route emergency. This has been in effect since
CFR Part 242 Conductor Certification was implemented in 2012.

I'll save you the trouble of digesting the entire regulation(even though it is worthy of reading and it may answer why some of your ideas and thoughts aren't allowed) and post the Cliff Notes.


Conductors (and engineers for that matter) must be federally certified and currently, there are two levels of service.

Although I was not aware it was in the CFR, I am pretty familiar with union shops and work rules having been in management and ownership of one. I mentioned something about a few extra dollars per hour in my original post. Perhaps I should have expanded on that more.

I don't, for even a second, think that such an idea would be feasible to just write a contract and start moving guys off coal trains into passenger trains. There is a genuinely different skill set for passenger that is layered on top of the basics of railroading. Dealing with passengers, HEP, train handling, etc... If there ever was genuine claim for extra pay for extra duties or beefing up work rules, this is it. That was why I mentioned extra dollars, and I should have been more clear that if such an idea were to be acted on, it requires a lot more than just writing a contract and changing the employer of record.

I guess I suggested the idea because I've had good luck in the past cross-training personnel and giving folks a bump in their pay. It's kept the company competitive or allowed us to enter new areas of business, while the guys take home more money most of the time. We were lucky to have a really good relationship with our unions but there was a lot of communication and give-take to get there, and it wasn't always that way.

[amtrakhogger]

Why draw a line where none exists? Maybe a $X/hour bonus for completing training to run passenger and meeting certain standards?

I've been digging a little for the start date of the changeover from host railroad crews to Amtrak crews in order to better inform this post. I believe the date was sometime around 1974 for non-NEC routes. The NEC takeover from Penn Central was 1976. Was also trying to check and see if the Amtrak Improvement Act played a part in this but couldn't find the answer right off.

Setting aside the question of whether or not there should or shouldn't be a separate labor pool at the end of the day it is no more economical for Union Pacific to be able to cover Amtrak than it is for Amtrak to be able to cover their operations. I think that given all the improvements over the past 20 years host railroads now run such tight labor pools that they couldn't have extra crews available without experiencing noticeable additional costs.

FWIK, NEC T+E was under Conrail from 1976-83 and then Amtrak 1983-current.

[SouthernRailway]

I don't know enough about freight crews or freight locomotives to have an opinion, but when I looked (years ago) and costs of LD trains, it seemed as though a large portion of their costs (then) were overhead, such as station costs.

No airline would make a cent if it had to operate its own airports for just 1 flight a day. Similarly, Amtrak being responsible for a range of infrastructure and other costs, just to operate 1 train a day, would be ruinous.

Doesn't it make sense for Amtrak to farm out what it can to freight railroads, if Amtrak doesn't have volume but freight railroads do: locomotives, crew, repairs, etc. Even paying a premium for those things could be cheaper than Amtrak handling its own.

[mtuandrew]

Decisions aren't always made on a straight labor cost, folks... Metra pays more to have BNSF and UP operate commuter service than they do their own employees.

BNSF, UP, etc. are far better able to bear the cost of things like healthcare and liability. I don't know if Amtrak reaches the level to be able to self-insure, and I'd have to think liability insurance would be lower by using your own folks at the throttle than whatever Amtrak has to pay today.

Metra is also in a (deadlocked?) contract dispute with UP, so who knows whether Omaha will continue to run trains for Chicago commuters.

And while UP, BNSF, and the like are either self-insured or have huge insurers backing them, Amtrak is backed by the biggest insurer of them all: the full faith and credit of the United States of America.

[east point]

If you go to freight crews I expect crew costs will be much higher. Probably twice as much on routes that are now covered by one crew would be 2 or more freight crews. That especially when you change RRs. Imagine the Michigan trains Freight RR to Detroit, back to NS who would have to hire more crew for their sporadic freight service, then CN for what 5 miles, then NS to CHI. Instead of one Amtrak crew.i

[eolesen]

That assumes NS doesn't have trackage rights (or the ability to get them) on the CN for that 5 mile stretch.

How many changes of host railroad *don't* take place at a crew change point today? And no, I'm not talking about the short stretches like the NS-CN-NS mentioned on the Michigan trains...

[mtuandrew]

Presumably should Amtrak be forced to bid out service (which it shouldn’t), the bid process would protect the right of the bidder to set its own crew change points and for them to qualify crews over the entire route. After all, even if Amtrak itself wasn’t crewing the trains, they’d be working under the auspices of NRPC and therefore have all the same rights and responsibilities of access to the host railroad. Therefore, NS or CN or CSAO or Watco could (but won’t) choose to provide crew change points where convenient (home rails, probably) and qualify them for passenger service over the entire route.

[Tadman]

Trip optimizer is a very interesting issue. If an algorithm can be written to run a freight train, an algorithm can be written to run a passenger train. Automated trains, until recently, have only been on isolated mines and are basically an abstract concept. Not anymore. With the massive R&D effort going into autonomous vehicles and drones, it's a safe bet that engineers will not have a job in 2050. There will probably be 20 humans in a control room in Omaha or Jacksonville that supervise 100 trains each. This is currently how many intermodal and port cranes are run, a similar harsh operating environment to railroading.

If you go to freight crews I expect crew costs will be much higher. Probably twice as much on routes that are now covered by one crew would be 2 or more freight crews. That especially when you change RRs. Imagine the Michigan trains Freight RR to Detroit, back to NS who would have to hire more crew for their sporadic freight service, then CN for what 5 miles, then NS to CHI. Instead of one Amtrak crew.i

It's pretty common nowadays for crews from one road to be qualified over other roads. The NS approach to Chicago for Detroit trains is a good example. NS trains their crews already for both the Detroit line and Chicago line, as they hold the freight contract on the middle and east ends of the Detroit line.

As to other hand-off points like Denver, Portland, MSP, Cleveland, I don't think there should be handoffs there. I'm pretty strong in my views about regionalizing the ends of LD trains. The SEA-PDX stub of the starlight should be another Cascade slot. The benefits of doing this are so great to passengers, carrier, and host.

[STrRedWolf]

Trip optimizer is a very interesting issue. If an algorithm can be written to run a freight train, an algorithm can be written to run a passenger train. Automated trains, until recently, have only been on isolated mines and are basically an abstract concept. Not anymore. With the massive R&D effort going into autonomous vehicles and drones, it's a safe bet that engineers will not have a job in 2050. There will probably be 20 humans in a control room in Omaha or Jacksonville that supervise 100 trains each. This is currently how many intermodal and port cranes are run, a similar harsh operating environment to railroading.

There's one problem with having an algorithm to run a passenger train (assuming commuter rail and Amtrak here). It's called PEOPLE!!!!! (cue Gossimer the Hair Monster screaming off into the distance).

What do you have with freight? Well, you have the engine(s) which have their systems that coordinate with each other and monitor the road conditions, PTC, Automated Track Protection, and know what consist they have so they can plot when and where to speed up, slow down, stop, wait, whatnot. This assumes the trackside detection systems (hotbox, drag, etc) can talk to the trains digitally (not over the radio in English). I can see that now with current technology.

But what do you have with passenger trains? Well... the same as freight, but your haul weight isn't really known even along the trip because:

• You got stops to handle (okay, we can code that in)
• People lie about their weight and the average American is heavier than what the FAA says (grabbing the nearest relevant standard)
• The amount of actual people on the trip is unknown even along the trip because last minute purchases and cancellations
• People's luggage is NOT weighed
• Person may be handicapped and in addition to their luggage, they gotta load their mobility scooter.

Add onto that the actions of those people and how much time they take... and the mistakes they make. What's to stop some passenger from pulling the emergency cord because they overslept their stop five stops ago? Or the train's getting hijacked? Or someone's having a stroke or heart attack?

You can fully automate a freight train RIGHT NOW. You can't automate a passenger train.

BTW, note I restricted it to commuter rail and Amtrak. People movers are a separate story equal to elevators.

[Tadman]

But what do you have with passenger trains? Well... the same as freight, but your haul weight isn't really known even along the trip because:

• You got stops to handle (okay, we can code that in)
• People lie about their weight and the average American is heavier than what the FAA says (grabbing the nearest relevant standard)
• The amount of actual people on the trip is unknown even along the trip because last minute purchases and cancellations
• People's luggage is NOT weighed
• Person may be handicapped and in addition to their luggage, they gotta load their mobility scooter.

Add onto that the actions of those people and how much time they take... and the mistakes they make. What's to stop some passenger from pulling the emergency cord because they overslept their stop five stops ago? Or the train's getting hijacked? Or someone's having a stroke or heart attack?

You can fully automate a freight train RIGHT NOW. You can't automate a passenger train.

BTW, note I restricted it to commuter rail and Amtrak. People movers are a separate story equal to elevators.

The same key that accounts for the variables in elevators and people movers is what you use to account for variables in passenger trains. Using something like actuarial data, it probably wouldn't be hard to find the average weight of male/female passengers and their luggage. That data could then be refined through a ticket scanning system that discretely weighs every passenger as they board. Then machine learning refines the data and incorporates it into the algorithm.

All that could be a moot point. Assuming the following - P42 at 268k lbs, Surfliner at 150k lbs, average passenger 150lbs, six cars, 75 people capacity. The train itself weighs 1.3m lbs dry. A full passenger load is 67,000lbs, half is 33,500lbs. Is that enough to change the train handling characteristics? The weight of the passengers is about the weight of a caboose. I've never experienced that from the controls.

[bdawe]

weight is so important to trip optimizer because you're trying to save some marginal amount of fuel. Which is not the case for passenger trains since you are explicitly trying to go fast

That said, there are tilting trains that keep track of loaded weight in order to disable the tilt system if there is to much weight aboard

[ThirdRail7]

I guess I suggested the idea because I've had good luck in the past cross-training personnel and giving folks a bump in their pay. It's kept the company competitive or allowed us to enter new areas of business, while the guys take home more money most of the time. We were lucky to have a really good relationship with our unions but there was a lot of communication and give-take to get there, and it wasn't always that way.

Again, this has nothing to do with "union" rules or cooperation. This has to do with a federal rule was passed and you have not made a business case for why a host railroad would want to shoulder the additional costs to train and keep their personnel qualified on an operation (passenger) that doesn't pertain to their business model (freight).

Trip optimizer is a very interesting issue. If an algorithm can be written to run a freight train, an algorithm can be written to run a passenger train. Automated trains, until recently, have only been on isolated mines and are basically an abstract concept. Not anymore. With the massive R&D effort going into autonomous vehicles and drones, it's a safe bet that engineers will not have a job in 2050. There will probably be 20 humans in a control room in Omaha or Jacksonville that supervise 100 trains each. This is currently how many intermodal and port cranes are run, a similar harsh operating environment to railroading.

A lot of these areas are confined, have sealed right of ways or as you stated are on isolated routes. As long as you have people in vehicles challenging trains to a joust at grade crossings, wandering down the tracks as it were a sidewalk, playing chicken on the tracks(forgive me but I don't recall people playing "chicken on the port crane") and lawyers continuing to defend these actions, I'm pretty sure you'll have someone present in the locomotive cab.

They may have a different title but I'm pretty sure someone will be on the train if something occurs.

Again, if you go back to the CFR which undermines the premise of your post and thread, it was only 8 years ago that the FRA enacted rules that made the presence of a conductor MANDATORY.

[east point]

I believe that the US government should declare aa national emergency for essential workers. That includes RR engineers. For the engineers that have retired require them to return to work especially Amtrak engineers that have recovered. Let the RR retirement board requirements be dammed!
There is already a shortage for CAL Z and also Florence, SC.

[STrRedWolf]

The same key that accounts for the variables in elevators and people movers is what you use to account for variables in passenger trains. Using something like actuarial data, it probably wouldn't be hard to find the average weight of male/female passengers and their luggage. That data could then be refined through a ticket scanning system that discretely weighs every passenger as they board. Then machine learning refines the data and incorporates it into the algorithm.

Not quite. The elevators can more readily measure weight because it's pushing/pulling it up and down. People movers have a bit more work to do because it has to keep track of how much power it's pushing to get the speed it needs. Trains you'd have to do it in aggregate, but if your load is constant through the trip, it's an easier calculation -- something that doesn't need a PC that makes hard-core gamers drool into lake-sized pools to do.

All that could be a moot point. Assuming the following - P42 at 268k lbs, Surfliner at 150k lbs, average passenger 150lbs, six cars, 75 people capacity. The train itself weighs 1.3m lbs dry. A full passenger load is 67,000lbs, half is 33,500lbs. Is that enough to change the train handling characteristics? The weight of the passengers is about the weight of a caboose. I've never experienced that from the controls.

FAA regs used to say 179 lbs. This was adjusted recently to 200 pounds for men, 179 for women, and 76 for kids under 13. They also say to use 16 lbs for carry-on luggage. This on top of having you weigh your checked luggage.

weight is so important to trip optimizer because you're trying to save some marginal amount of fuel. Which is not the case for passenger trains since you are explicitly trying to go fast

That said, there are tilting trains that keep track of loaded weight in order to disable the tilt system if there is to much weight aboard

True, and the force needed to tilt can be measured, which then translates back to weight... which is still variable over the trip.

A lot of these areas are confined, have sealed right of ways or as you stated are on isolated routes. As long as you have people in vehicles challenging trains to a joust at grade crossings, wandering down the tracks as it were a sidewalk, playing chicken on the tracks(forgive me but I don't recall people playing "chicken on the port crane") and lawyers continuing to defend these actions, I'm pretty sure you'll have someone present in the locomotive cab.

That then gets to what Tesla, Alphabet (Google), Uber, etc are doing with self-driving cars. It's far from perfect because they ARE running over pedestrians. It's slightly easier on the rail, but you still have the local environment, and you'll want something fail-safe. What's better, a train that keeps going when it hits a person, or one that stops and calls the dispatcher for help?

I'm with ThirdRail. We'll still have human(oid) engineers and conductors on board for quite a while yet.

[RRspatch]

I believe that the US government should declare aa national emergency for essential workers. That includes RR engineers. For the engineers that have retired require them to return to work especially Amtrak engineers that have recovered. Let the RR retirement board requirements be dammed!
There is already a shortage for CAL Z and also Florence, SC.

I'm pretty sure COVID-19 is effecting all railroad workers including those working close together in "control rooms". Now lets see, it's been 23 years since CETC so no calls from the 215 area code but maybe the 817 area code. *Looks at phone wondering if I need to activate the Do Not Disturb option* ....

As for RRB I'm sure they would wave the 17K limit on earnings if ordered to do so by the FRA and or congress.

As far as COVID-19 is concerned it's possible I already had the mild version -

https://www.webmd.com/lung/news/2020040 ... ms-study#1

I had two weeks (!) of diarrhea and elevated blood sugar readings. The diarrhea has cleared and the blood sugar numbers are slowly coming down. Because I wasn't short of breath, coughing or running a fever I couldn't get a test. The only way to be sure is an antibody test which is still in the works. Without an antibody test I, and others I assume, would be hesitant to go back to work not knowing if we really had it or not.