to high tech?

[Anonymous]

Are locomotived becoming to high tech and to big? The vintage locomotives, like, the SD-7s where east to work with and repair. No need to have have a huge text book to find the problem, no need for a computter to find the problem. If one chip gose out, the whole locomotive is usless. Once, there was an Amtrack train the had a brand new GE unit, and a chip burned ouot. It took them 4 hours while tying up the main line to finly find the problem only to have to set the unti out wich remaind ideling for 3 days untill finly a tech guy told the roadswitch train to move it to a repair shop. Now had that been, say, a F-7, the trian might have been able to keep goind, despite being late. Its no wounder why people are getting less smart, the need computters to think for them. I have heard reports of some of the newest locomotives, burnning there tubro chargers because of a faultu comoutter, or contorl stands that have to be ducted taped to the floor. What even happened to a reliable locomotive? Locomotive builders are ignoring the idea that simple is better.
And along those lines, we need more new 4 (four) axle locomotives. The locomotives are getting to big for branch lines, and road switching. We are losoing focus of on line business by building locomotives that are to big.

[Randy Stahl]

I AGREE!!!!!!!!!!!!! A COMPUTER TO RING THE DAMN BELL???? OR TO BLOW THE WHISTLE>>>>> CMON
\Randy

[Anonymous]

U go randy! [:D]

[csxengineer98]

locomotives are becomeing just like the car parked in your driveway.... a computer system for eveything...and i think alot of it is unnessisary... to fix your car anymore you need to be a computer programer as well as a gear head.... its not much differnt with the new technotoaster locomotives... the biggets problems with new power are computer issues that make the locomotive not work right... you get a computer fualt of some kind and it makes your locomotive a severl ton paper weight.... the upside is...just like your home computer..alot of problems can be solved by just shutting it off and let it reboot... but that is easier then it sounds when your motorting along and your unit takes a dive...
very manufactor of just about eveything wants to put as much computer controlled electronics in eveything.... computers fail..and they fail at the worst posable times....
get ride of some of the unnessasary computer systems..and you will have a better locomotive with less down time due to software crashes...
csx engineer

[Paul Milenkovic]

Don't know if computers are the problem.

Last time I had a trouble code from my Ford Taurus I used this gizmo I bought off the Internet to hook the car up to a Windows laptop for $150, and it told me that I had "insufficient EGR." Like the guy looking for his contact lens under the streetlight, I cleaned the EGR valve but that did not fix the problem. I took it to a Ford dealer and they replaced the EGR sensor used to control the EGR valve. Why didn't I try that? The Ford dealer may have had a better gizmo with more detailed codes as to the trouble, but had I known it was the sensor, getting at it was beyond my ability because everything on that engine is so packed in.

I don't think computers are the problem -- if you have a way to read the trouble codes AND if you can get at things you are golden.

How are the current generation of locomotives from the standpoint of access to the parts? I heard that one reason the Fairbanks Morse was drummed out of the locomotive business is that their Diesel is two-stroke opposed-piston, and you can't just slip out a "power assembly" for repair without removing the top crankshaft.

[Train Guy 3]

What happened to the days when you just pulled a handle..... didn't that work fine?

[Anonymous]

It is true though, it seems that nowadays when you step into a locomotive it resembles the starship enterprise more than it resembles a locomotive.

Like stated above though, everything is like that nowadays from boats to planes to cars and trucks.

[Anonymous]

It's getting worse. the class 1 railroads are going to be fighting to do away with 2 person train crews in this next contract. safety and efficiency be dammed...they want to cut labor costs..and let a computer do the driving. then one wonders..are trains going to be more secure from terrorist threats? what happens when chips fry? this will be an interesting contract

[Anonymous]

I can't speak from experience with locomotives, but I can as a diesel mechanic. The electronic controls on today's engines allow them to run more efficiently and make more power because they are monitored so closely and manged so well. The electronics on commercial equipement is not the same as a consumer level device, diagnostic parameters are far more descriptive. Options like "snapshot" are available, where a list of parameters can be selected, and when a certain value is reached, a snapshot is taken. Individual parameters can be selected, and an alarm selected for a given value. Preset diagnostic routines, automatic cycling of engine rpm, cylinder cutout can be selected. All of these things help in diagnosing a problem.

Systems are designed for failure. Parameters can be set to limit rpm or power in the event of some failure in a "ramp down" or a "shut down" mode. If there is some failure of electronics the unit can operate in a "limp mode" for a preset period of time, and if a unit is in an application where it is absolutly necessary to avoid any shutdown, this can be overridden temporarily or programmed.

All this said, there are still failures that cannot be shown by any diagnostic system, and for that, there is no substitute for a working knowledge of the equipement.(you still have to be a gearhead)

[csxengineer98]

QUOTE: Originally posted by jruppert

I can't speak from experience with locomotives, but I can as a diesel mechanic. The electronic controls on today's engines allow them to run more efficiently and make more power because they are monitored so closely and manged so well. The electronics on commercial equipement is not the same as a consumer level device, diagnostic parameters are far more descriptive. Options like "snapshot" are available, where a list of parameters can be selected, and when a certain value is reached, a snapshot is taken. Individual parameters can be selected, and an alarm selected for a given value. Preset diagnostic routines, automatic cycling of engine rpm, cylinder cutout can be selected. All of these things help in diagnosing a problem.

Systems are designed for failure. Parameters can be set to limit rpm or power in the event of some failure in a "ramp down" or a "shut down" mode. If there is some failure of electronics the unit can operate in a "limp mode" for a preset period of time, and if a unit is in an application where it is absolutly necessary to avoid any shutdown, this can be overridden temporarily or programmed.

All this said, there are still failures that cannot be shown by any diagnostic system, and for that, there is no substitute for a working knowledge of the equipement.(you still have to be a gearhead)

thats fine in a perfict world..but computers fail......but the problem is not so much the maintance aspect as it is an opporational aspect that is the issue here.... the computer systems might make your job easier at finding and fixing issues...but when the software desides to take a **** on you while your underway..it makes my job that much harder...besides..almost half of the problems i have had with onboard computers have been with airbrake systems that dont want to come online... or traction motors that want to cut out on thier own..just becouse of some wheel slip from wet rail...or the computer not allowing you to do something.....like cutting in or out the airbrakes..which falls under the air brake computer issues...why do you need a computer to cut that air brake system in and out..when a manual valve has been doing it for years without any problems... i think thier are something that the computer shouldnt have any parts in..and one of them is the brake system...
just my 2 cents for what is worht
csx engineer

[Junctionfan]

There is nothing wrong with the technology; the problem is it lacks the basic practicality to make it more favourable than switches and pulleys. In a car there are buttons that say check oil or low fuel. There should be a system that says change harddrive or change chip B section 8 or something like that. There should be spare chips and back up hard drives in a cabinet of the locomotive that can be simple enough to pop in and out where needed like a lego block with a cable. Technology doesn't have to be complicated; right now the only way to fix things is take it to a repair shop. You can't fix computer things yourself. It is like that for cars too. It shouldn't be like that.

[jchnhtfd]

Like all changes in technology, there's an upside and a downside...jruppert pointed out one major advantage of computer conrol, at least as it applies to prime movers (the diesel): the computer can control the thing a lot more accurately than all but the very best maintained mechanical systems, and under a much wider range of conditions, resulting in lower fuel costs and emissions. On the other hand, when the magic box goes phoot (and they do!) there should be -- and almost always is -- a fail operational/limp home option, which is rarely much worse than the mechanical controls.

With regard to junctionfan's, comment, the newer magic boxes (dash 9 and -2) are all set up as card racks (as I understand it), and 'fixing' something involves swapping cards -- which is a lot easier than actually repairing a big piece of electromechanical gadgetry (Randy Stahl: I'll bet you have a comment on this, and can correct me!).

Another point is that the newer gadgetry, if properly set up, can radio ahead to tell the mainentance folks that it's coming in with a problem -- and a pretty good idea as to what the problem is.

Downsides? Earlier computer driven hardware just wasn't built heavy enough to take railroad service. Period. And a lot of failures and bad raps came from that. Also, some failures really do result in a dead engine, where in the F-7 days you might have been able to jury rig something.

[Anonymous]

"DAVID"
"yes HAL"
"The AE-35 unit is going to mafuncion"
"How soon HAL"
"My probability calculations say within twelve hours"
"so what do you propose HAL"
"we need to go EVA and retrieve the unit"[alien][alien][alien][alien][alien][alien][alien]

[Anonymous]

All this tech stuff and you STILL can't play pacman going down the rails. Who said computers will rule, while engineers play gameboys to pass the time,waiting for the computer to say the track is clear.

[Anonymous]

........."HAL, open the *&^$%#@% POD BAY DOOR!!!!!"
"David, I'M sorry, you didn't say please"[:D][:D][:D][:D][:D]

[Randy Stahl]

The biggest problems with these high tech pieces of crap isn't the hard ware itself , it's usually some stupid wire with a poor connection someware, The computer crap works at a Low voltage, like 5 -15 volts, so it don't take much to cause an intermittent problem that is next to impossible to find. Usually an electrician will change out EVERY electronic part on the engine. I think this is going to get worse as the locomotives get older..... I can't wait!
Randy

[Overmod]

There is almost certainly a way to build computerized automatic test equipment for locomotive systems. Note that most military systems... and microcomputers as early as the first IBM PCs... have 'built-in test equipment' (some of this manifests as the POST that runs at cold boot) with very informative test results and recovery options. Locomotive geeks figured that wasn't important on purpose-built systems, and saved a bunch of bucks by leaving it out. Instead, they used the IBM FRU (field-replaceable unit) concept: If it doesn't work, walk some sort of diagnostic MAP (maintenance analysis procedure, or similar acronym), and start pulling boards and replacing with known-good spares until Mr. Problem goes away to stay. Remember old Engine Charlie Wilson in paraphrase: what's good for EMD is good for America. Just have at least one of everything in your grip... or suitcase... or trunk... or truck... oh, dear.

It would be laughably simple to put an artificially-intelligent expert system into the locomotive software system, running on a separate processor, that would continually monitor the overall state of the system, detect many problems early, and provide common-sense (and polite) answers that crews could understand, respect, and respond to. Hell, pay me enough to get SWMBO off my tail and *I'll* build it.

Part of the key is to avoid 'deterministic' systems, where everything has to have its own complicated little answer. Anybody who has played with the IBM S/34 and S/36 minicomputers knows that the fattest manual... printed in small type... is the 'error codes' manual; there's a little four- or five-digit number that shows up in a corner of the screen when something breaks, and you look up what it means and what to do about it. Rapid, mind-numbing complexity... even assuming you have folks in the trenches who are willing to read the damn thing.

NAJPTC is a glorious, shining example of why you don't let geeks program real-world railroad safety systems. No number of clever and expensive 'scenarios' will let you recover from one idiot's implicit assumption that trains occupy 'zero length' when going between blocks...

[Anonymous]

When steam engines took 3 minutes to find the problem, and 3 days to fix...

Or the Steam turbines which took 3 days to find and 3 minutes to fix...

[kenneo]

QUOTE: Originally posted by Train Guy 3

What happened to the days when you just pulled a handle..... didn't that work fine?

Well, when someone has to pull a handle, you need a human being, with needs, buildings to house them, food to feed them, potty rooms, wages, benefits, and so on.

Now, with a computer, you don't need any of that. "Where have all the people gone, long time passing?" (Peter, Paul and Marry)

[Anonymous]

I agree with Jay1. We have 2 unrebuilt GP7's (we chopped the noses thats all the prime movers are still original) that run fine for the most part. Most of the their problems are minor and easily fixed. Our geeps are over 50 years old and original and they still run every day. In 50 years i want to see an SD70,80,or90 still running


Theres a reason why ALCO went out of business

[jchnhtfd]

Randy's absolutely right about the connection problem -- not only is the voltage low, but the currents involved are very small. It doesn't take much corrosion to screw up these low power connections, with all kinds of weird difficult to trace problems resulting (they're almost always intermittent, at least at first, which really has the mechanic talking to himself!). As usual, there's a cost/benefit tradeoff -- it is possible to use connectors which have much better properties but... they cost a lot more.

And he's also right in that it's only going to get worse as the locos get older -- particularly if the bean counters insist on replacing stuff with the cheapest will fit rather than the best...

An all mechanical -- or almost all, anyway -- machine like a GP7 (or my 1970 Chevy truck) will run just about forever with very simple routine maintenance and care. But... with the proper maintenance and care, so will an SD70. As witness there are a lot of Boeing 747's out there, still flying reliably, 40 years out -- and they are definetly not in the 'simple' category!

[Hugh Jampton]

Any infant technology always proves to be unreliable at first. Early steam and diesels were unreliable. As the technology matured though, reliability improved. The SD 70 80 or 90 may be unreliable, but the SD 100 110 or 120 will probably be OK. It has to be for the company to stay in business really.

[oltmannd]

There are two issues that have driven locomotives to computerization. Both are economic.

One is performance. A non-electronic wheel slip system on a DC locomotive gets you ~18% adhesion. A solid state electronic one, 21%, a TTL logic system 24%, a microprocessor based system, 26+%. There is a huge difference in the tonnage rating of an SD9 and an SD70 and without increasingly sophisiticated wheel creep systems, it would not be possible. This is just one example - there are many others. Another would be more precise auxilliary HP control to reduce fuel consumption.

The other is simplicity of design. Look in the electrical compartment of a GP35 or SD50, then look at an SD60. Which has more thousands of feet of wiring? The cost to build is cheaper in the long run using computers instead of relay logic and wiring. How about the simplicity of doing a mod? A software change is much simpler and faster to do than adding dozens of wires and relays.

The last truly simple locomotive was the GP9. Surely nobody's making the case that this is the ideal locomotive for today!

[daveklepper]

Have any of you had experience with EMD's satilite continuous monitoring of the status of their locomotives on the road to permit any possible possibility of failure to be predicted in advance (if it works?)?

[Randy Stahl]

The SD70s got some pretty big bundles of wire... makes a good target for someone running a sheet metal screw into them. Emd always insisted on individual magnetic switch gear , it was a quantum leap when they installed the motorized switch gear in the sd 40-2 and placed ALL the switches in the front of the locomotive instead of having a front and rear electrical cabinet. There is not a big difference in the amount or length of wire runs between sd40 & sd70
Randy

[Overmod]

Here's an article you've probably read, but that might fill others on this thread in on 'satellite' locomotive monitoring. (I'

http://www.railwayage.com/apr01/loco_health.html

and a link from EMD about a different and slightly later article

http://www.gmemd.com/en/company/news/links/intellitrainreport.htm

that describes some of the specific parameters that EMD's system monitors.

I thought it was interesting that neither EMD nor GE prefer to use 'satellites' for the transmission of the data -- both basically prefer to use some version of data-over-cellular-telephone system. By law, cell-phone towers need to be located in 'non-public' areas when possible, and railroad ROWs are a good place to put them; this helps on-locomotive use.

It also appears that both EMD and GE were smart enough to go directly to TCP/IP data structures so they can use the Internet for all the connectivity other than the mobile wireless link from communication tower to locomotive. That cuts down on the cost and trouble of lots of the software development and connectivity. Note that by using OTS software it would be simple to send text messages to engine crews, parts suppliers, etc. as part of the event detection, as EMD indicated they were starting to do in 2002.

One might think that the OnStar system could be comfortably leveraged as an EMD data... and help... system. This system can handle mixed voice and data, and a reasonable number of operators are all 'in place' and funded via the broad base of automotive application. Wouldn't take much to have a couple of these folks trained in railroad applications, or to establish hotlines, text messaging or e-mail, etc. to forward calls, telemetry data, or emergency information to appropriate places.

Things don't look too promising in terms of railroad valuation of this technology, if the ENSCO experience at Conrail... and non-experience with CSX/NS... is any indication. Note that the advent of differential GPS gives even rudimentary locomotive-information systems great additional usefulness, for PTC as well as train-handling purposes.

I do look forward to hearing railroaders' first-hand experience with the specific devices and systems the builders are using. I worry, however, when I hear someone as competent as Randy talking about '5 VDC or so' as being what computer stuff uses; it's not that complicated to teach people about the meaningful electrical stuff behind TCP/IP, cell communications, onboard processors and sensors, buffer capacity, etc., and evidently that's not being done, let alone done well. Think how much fun Randy could have if he could run a complete load test on an SD-45 by plugging a power-isolated or optical jack into a socket on the locomotive and starting an automated set of tests on the laptop... no need to be a computer geek to start and run the computer to do that, either. (Hell, he could do it on the whole damn power consist, including DPUs, if EMD's network system works the way it should...)

[Randy Stahl]

AWWWW Thats really taking the fun out of it!!!!!
Randy

[Anonymous]

While there seems to be some new and exiting possibilities with all this new technology, I have to admit there is a part of me that erres caution.

Why do quartermasters (navigators) on navy ships still have to learn how to use a sextant? Even with LORAN, GPS, and Radar?

Component specialists are a thing of the past.

A shop used to have a turbo guy, a fuel injection guy, etc....
The new guys did remove and replace, oil changes, etc....

Yes, those things are time, and labor intensive, but let's not forget how to use a sextant.

[ericsp]

QUOTE: Originally posted by jruppert

While there seems to be some new and exiting possibilities with all this new technology, I have to admit there is a part of me that erres caution.

Why do quartermasters (navigators) on navy ships still have to learn how to use a sextant? Even with LORAN, GPS, and Radar?

Component specialists are a thing of the past.

A shop used to have a turbo guy, a fuel injection guy, etc....
The new guys did remove and replace, oil changes, etc....

Yes, those things are time, and labor intensive, but let's not forget how to use a sextant.

I know the chances of all or those failing is incredibly small, however, I would not want to be stuck at sea with no way to navigate. I guess that is the reason why.

[Train Guy 3]

Hmmmm OnStar on EMD locos..... I never thought about that. GM owns the system so that works perfect for putting it to use. Can't wait to see that commercial!