BREAKING NEWS - Amtrak train rear ends freight in Chicago

[ValleyX]

The restricting signal was the warning.  The dispatcher can't be expected to be handing out warnings to every train when he or she has so many trains to deal with, the operating crew members have to know the territory.  If they don't know the territory, they probably should have been running restricted speed before they even encountered a restricting signal.

[Railway Man]

I'm looking forward to finding out how involved the dispatcher(s) were in this acident. There has been some confusion regarding which railroad / dispatcher has authority at this location - NS in Michigan or Metra. I would have to believe that since it's NS owned track it would be a NS dispatcher calling the shots. Anyway, while the dispatcher isn't required to communicate situations like this to a train crew with signal indication being sufficient, I think that this type of situation would have warranted special attention / instructions considering the cargo on the Amtrak train and the congestion in the area: NS DS to Amtrak 321... (after Amtrak 321 respnds) I'm going to line you through CP Englewood two to one to get you around a freight that's eastbound approaching CP Englewood on one. Proceed on signal indication at normal switch speed then restricted speed not to exceed 15mph. I've got a westbound freight holding at CP [whatever the next control point is] so I need you pull up behind that freight and hold. The Amtrak train would then have to repeat those instructions back to the dispatcher that they acknowledge they are going to cross two to one at CP Englewood on signal indication not to exceed 15mph and that they are to hold behind the freight occupying the block.

CC

Not required by the rules on any road I have worked for, not required by CFR 49, not necessary, and in most cases not a good idea, in fact portions of what you think should have been done are a terrible idea.  Some dispatching offices have specifically forbidden all or some of the practices you think should have been used as it implies knowledge the dispatcher may not have or does not have, and the information could be used by a crew to modify the signal indication or otherwise disregard the rules.  The communication you propose creates opportunities for missed repeats and misinterpretations, which decreases safety, not increase it.

If you want to prevent 99.99% of rear-end collisions like this, install CBTC, ETMS, or ATS.  And in fact Metra is doing just that (CBTC overlay).

RWM

[jeaton]

I'm looking forward to finding out how involved the dispatcher(s) were in this acident. There has been some confusion regarding which railroad / dispatcher has authority at this location - NS in Michigan or Metra. I would have to believe that since it's NS owned track it would be a NS dispatcher calling the shots. Anyway, while the dispatcher isn't required to communicate situations like this to a train crew with signal indication being sufficient, I think that this type of situation would have warranted special attention / instructions considering the cargo on the Amtrak train and the congestion in the area: NS DS to Amtrak 321... (after Amtrak 321 respnds) I'm going to line you through CP Englewood two to one to get you around a freight that's eastbound approaching CP Englewood on one. Proceed on signal indication at normal switch speed then restricted speed not to exceed 15mph. I've got a westbound freight holding at CP [whatever the next control point is] so I need you pull up behind that freight and hold. The Amtrak train would then have to repeat those instructions back to the dispatcher that they acknowledge they are going to cross two to one at CP Englewood on signal indication not to exceed 15mph and that they are to hold behind the freight occupying the block.

CC

Not required by the rules on any road I have worked for, not required by CFR 49, not necessary, and in most cases not a good idea, in fact portions of what you think should have been done are a terrible idea.  Some dispatching offices have specifically forbidden all or some of the practices you think should have been used as it implies knowledge the dispatcher may not have or does not have, and the information could be used by a crew to modify the signal indication or otherwise disregard the rules.  The communication you propose creates opportunities for missed repeats and misinterpretations, which decreases safety, not increase it.

If you want to prevent 99.99% of rear-end collisions like this, install CBTC, ETMS, or ATS.  And in fact Metra is doing just that (CBTC overlay).

RWM

I am not sure that it is on line, but I recall reading an NTSB report of a collision between an Illinois Central freight and an IC passenger train south of Chicago near Monee, Illinois.  The accident occured sometime in the 1960s. 

The freight train was southbound on Track 3 near the point where that track ends with switches and crossovers to Tracks 1 and 2.  Train movements at that location were controled by positive signals (red is STOP-period).  Due to fog, visability was poor.  When the freight engineer reached the point where he would have to apply brakes to stop the train at the signal, he was still not able to see the signal lights.  Seeking to avoid an unnecessary stop, the engineer radioed the tower operator or dispatcher controlling the switches and signals to ask if he had the signals allowing him to proceed to crossover to track 1.  Either the dispatcher or the engineer misunderstood the communication and the engineer thought he had an "OK" to proceed.  The signal he was approaching was displaying a STOP indication.  When it became visable, it was too late to stop the train at the signal.  As a result the train proceeded to a point where the engine was actually setting on Track 2 facing the Northbound passenger train.  The passenger train was unable to stop in time and collided with the freight train.

I do not know about the applicable rules of the time, but the NTSB report stated that the engineer must not rely on a radio communication to determine the status of a signal.  Rather, engineers must proceed in a manner so that appropriate and timely steps can be taken to comply with the signal aspect.   In other words radio communication must not be used to determine the aspect of a signal or the occupancy status of track governed by signals. If such a rule was not in place at the time of the accident, it was probably added to the IC's book.

[edblysard]

GCOR has just such a rule, and I believe NORAC has one also.

2.12 Fixed Signal Information

Employees must not use the radio to give information to a train or engine crew about the name, position, aspect, or indication displayed by a fixed signal, unless the information is given between members of the same crew or the information is needed to warn of an emergency.

What I was suggesting was that if the dispatcher was re-routing the Amtrak from its normal route, the dispatcher should have made a call to the train, warning him of traffic ahead.

I understand the signal system there and the aspect of the signals conveys both route and speed, but a simple call from the dispatcher along the lines of "Amtrak number 8, I will cross you over at control point STO 003, proceed at restricted speed looking out for traffic ahead" might have prevented this altogether.

 

Either way, I would hate to be in that engineers position right now.

I'm looking forward to finding out how involved the dispatcher(s) were in this acident. There has been some confusion regarding which railroad / dispatcher has authority at this location - NS in Michigan or Metra. I would have to believe that since it's NS owned track it would be a NS dispatcher calling the shots. Anyway, while the dispatcher isn't required to communicate situations like this to a train crew with signal indication being sufficient, I think that this type of situation would have warranted special attention / instructions considering the cargo on the Amtrak train and the congestion in the area: NS DS to Amtrak 321... (after Amtrak 321 respnds) I'm going to line you through CP Englewood two to one to get you around a freight that's eastbound approaching CP Englewood on one. Proceed on signal indication at normal switch speed then restricted speed not to exceed 15mph. I've got a westbound freight holding at CP [whatever the next control point is] so I need you pull up behind that freight and hold. The Amtrak train would then have to repeat those instructions back to the dispatcher that they acknowledge they are going to cross two to one at CP Englewood on signal indication not to exceed 15mph and that they are to hold behind the freight occupying the block.

CC

Not required by the rules on any road I have worked for, not required by CFR 49, not necessary, and in most cases not a good idea, in fact portions of what you think should have been done are a terrible idea.  Some dispatching offices have specifically forbidden all or some of the practices you think should have been used as it implies knowledge the dispatcher may not have or does not have, and the information could be used by a crew to modify the signal indication or otherwise disregard the rules.  The communication you propose creates opportunities for missed repeats and misinterpretations, which decreases safety, not increase it.

If you want to prevent 99.99% of rear-end collisions like this, install CBTC, ETMS, or ATS.  And in fact Metra is doing just that (CBTC overlay).

RWM

I am not sure that it is on line, but I recall reading an NTSB report of a collision between an Illinois Central freight and an IC passenger train south of Chicago near Monee, Illinois.  The accident occured sometime in the 1960s. 

The freight train was southbound on Track 3 near the point where that track ends with switches and crossovers to Tracks 1 and 2.  Train movements at that location were controled by positive signals (red is STOP-period).  Due to fog, visability was poor.  When the freight engineer reached the point where he would have to apply brakes to stop the train at the signal, he was still not able to see the signal lights.  Seeking to avoid an unnecessary stop, the engineer radioed the tower operator or dispatcher controlling the switches and signals to ask if he had the signals allowing him to proceed to crossover to track 1.  Either the dispatcher or the engineer misunderstood the communication and the engineer thought he had an "OK" to proceed.  The signal he was approaching was displaying a STOP indication.  When it became visable, it was too late to stop the train at the signal.  As a result the train proceeded to a point where the engine was actually setting on Track 2 facing the Northbound passenger train.  The passenger train was unable to stop in time and collided with the freight train.

I do not know about the applicable rules of the time, but the NTSB report stated that the engineer must not rely on a radio communication to determine the status of a signal.  Rather, engineers must proceed in a manner so that appropriate and timely steps can be taken to comply with the signal aspect.   In other words radio communication must not be used to determine the aspect of a signal or the occupancy status of track governed by signals. If such a rule was not in place at the time of the accident, it was probably added to the IC's book.

[Railway Man]

GCOR has just such a rule, and I believe NORAC has one also.

2.12 Fixed Signal Information

Employees must not use the radio to give information to a train or engine crew about the name, position, aspect, or indication displayed by a fixed signal, unless the information is given between members of the same crew or the information is needed to warn of an emergency.

What I was suggesting was that if the dispatcher was re-routing the Amtrak from its normal route, the dispatcher should have made a call to the train, warning him of traffic ahead.

I understand the signal system there and the aspect of the signals conveys both route and speed, but a simple call from the dispatcher along the lines of "Amtrak number 8, I will cross you over at control point STO 003, proceed at restricted speed looking out for traffic ahead" might have prevented this altogether.

It might have, but why not just expect people to obey the rules?  And if you don't think they can obey the rules sufficiently consistently, then overlay a positive train control system.

I'm not comfortable with the idea of going down this path of suggesting the dispatcher "ought" to have done something.  The regular practice of advising trains that the track may be occupied may lead to people relying upon that information, or the absence of that information, instead of obeying the rules.  And then, when the next collision happens, the train crew says "well, normally the dispatcher tells us that the track is occupied but this time he/she didn't, so we didn't think there was anything there."  The train crew would have a point.

RWM

 

[Dutchrailnut]

Norac's equivalent to Ed's G.C.O.R rule is :

712. Signal Indications

Dispatchers or Operators must not advise the aspect, name, or indication of any fixed signal, and crew members must not request this information. Crew members may use the radio to communicate a fixed signal to other members of the same crew.

Except as provided in Rule 241, radio communication may not be used to convey instructions that would have the effect of overriding the indication of a fixed signal. Radio communication may only be used to impose a more restrictive action than the indication of a fixed signal.

(The rule 241 exception is to being talked by a red signal)

[edblysard]

Railway man and Dutchrailnut...

Part of my concern is that I work in essentially "dark territory" for the most part, under RTC, radio traffic control.

We do have to enter CTC on occasion, down here and working under GCOR, the dispatchers do advise trains of traffic ahead.

 

I believe it has to do with the fact that in the North East Corridors, the signal aspect is used both for route indication and speed, while under GCOR, timetables generally set speed, and signals advise of routing, or grant authority to occupy a given portion of the railroad.

 

Our dispatchers routinely ask trains for their location, and routinely advise of traffic ahead.

[Railway Man]

Railway man and Dutchrailnut...

Part of my concern is that I work in essentially "dark territory" for the most part, under RTC, radio traffic control.

We do have to enter CTC on occasion, down here and working under GCOR, the dispatchers do advise trains of traffic ahead.

 

I believe it has to do with the fact that in the North East Corridors, the signal aspect is used both for route indication and speed, while under GCOR, timetables generally set speed, and signals advise of routing, or grant authority to occupy a given portion of the railroad.

 

Our dispatchers routinely ask trains for their location, and routinely advise of traffic ahead.

Some railroads are speed signaled, some are route signaled, and thanks to mergers, most are now a hybrid of both.  GCOR is silent as to route signaling vs. speed signaling, and as far as I can tell NORAC is too.

Many dispatchers advise of trains ahead when authorizing a train to pass an absolute signal (Rule 9.12.1), or advise in single-track territory to expect a meet, as a convenience to the train crew so they know that they'll have to lay off crossings.  I don't have a problem with this so long as it's an experienced and cautious dispatcher who understands that this can easily cross the line from a smart practice to an unsafe practice.  Where I frown is when it's suggested that the dispatcher has responsibility to do this, because it is the crew's responsibility to obey restricted speed and signal indications, and the absence of instructions by the dispatcher might lead a careless crew to make a habit of ignoring restricted speed and disregarding signal indications.  It's a fine line; but I think we have to be on guard to avoid crossing it.

RWM

[Chris30]

Quote by: Railroad Man

Where I frown is when it's suggested that the dispatcher has responsibility to do this

Responsibility depends on the situation and whether the situation can be described as an emergency situation. Now, I don't work for a railroad and I'm not going to pretend that I know the rulebook, but I do know what rule #1 is - safety. It's probably listed before the table of contents in the rule books! I think that the NTSB information and the posts on this subject regarding this crash have made it clear that it's the train crews responsibility to follow the signal indication given. 

For what it's worth, I've used CTC based dispatching simulators for about ten years now. I've seen a few posts suggesting that the dispacther can't watch everything at once. That's true. Generally, in my experience using dispatching simulators the higher the priority on a train the more attention it gets. Amtrak is the top priority (maybe). If I'm the dispatcher in this situation and I make the decision to route a loaded passenger train through a congested freight yard area and into an occupied block with a stopped train ahead I'm going to give the engineer some verbal instructions to verify that the train crew understands the signal indication given. Then, I'm going to keep an eye on the situation. It's not just about responsibility, it's also about conscience. One engineer in the cab right? Who are they calling signals to? Ed already made it  clear that the luggage, er, dead-heading employee can't be involved. The conductor in the passenger cars simply notes the signal indication as read over the radio by the engineer. This is why I raised the question on an earlier post about being curious to the actions of the dispatcher in regards to this accident.

CC

[Railway Man]

Quote by: Railroad Man

Where I frown is when it's suggested that the dispatcher has responsibility to do this

Responsibility depends on the situation and whether the situation can be described as an emergency situation. Now, I don't work for a railroad and I'm not going to pretend that I know the rulebook, but I do know what rule #1 is - safety. It's probably listed before the table of contents in the rule books! I think that the NTSB information and the posts on this subject regarding this crash have made it clear that it's the train crews responsibility to follow the signal indication given. 

For what it's worth, I've used CTC based dispatching simulators for about ten years now. I've seen a few posts suggesting that the dispacther can't watch everything at once. That's true. Generally, in my experience using dispatching simulators the higher the priority on a train the more attention it gets. Amtrak is the top priority (maybe). If I'm the dispatcher in this situation and I make the decision to route a loaded passenger train through a congested freight yard area and into an occupied block with a stopped train ahead I'm going to give the engineer some verbal instructions to verify that the train crew understands the signal indication given. Then, I'm going to keep an eye on the situation. It's not just about responsibility, it's also about conscience. One engineer in the cab right? Who are they calling signals to? Ed already made it  clear that the luggage, er, dead-heading employee can't be involved. The conductor in the passenger cars simply notes the signal indication as read over the radio by the engineer. This is why I raised the question on an earlier post about being curious to the actions of the dispatcher in regards to this accident.

CC

I give you credit for your enthusiasm and interest, and for having thought through the scenario.  So don't feel bad that in taking these actions you would have committed a rules violation and an unsafe act.  The dispatcher has no idea what the signal indications are in the field.  Nor does the dispatcher know where the other train is, he merely has an indication on his CTC machine that there is something shunting the track in the block which might be a train.  Or it might be a broken rail, who knows?  The moment the dispatcher opens his mouth and "instructs" the crew or "makes sure they understand the signal indication" he's violated the rules as well as totally baffled the crew.  The dispatcher doesn't run the train, the engineman does.

And I'm curious, how do you propose the dispatcher "keep on eye" on the situation?  Depending on where the track cuts are located, he may not even have separate track lights in that block on his console.  All he is going to see is an occupancy in the O.S. when the following train hits it and another occupancy beyond it which may already be lit by the first train!  He knows not the speed of either train, the location of either train, the signal indication either train is seeing, or the sight distances.  By the way, a "green signal" on a dispatching console only means the signal system will accept movement in that direction when the signal system decides it is safe to do so.  The signal in the field that shows green on the dispatcher's console could well be showing red to the train crew looking at it.  That's why it's called an automatic block system.  All the dispatcher does is request priority of movement and route.  The signal system decides what will really happen according to an internal system of logic that is designed on fail-safe principles. 

Lining a train onto an occupied track is a very common, very ordinary condition; it's not special or unusual.  What is unusual when the engine crew of a train being lined onto an occupied track fails to proceed at restricted speed and collides with another train.  The presumption that this scenario is high-risk, unusual, requires extra attention, etc., is incorrect.  Would you have the dispatcher worry that a train won't adhere to a red signal?  A flashing yellow?  A Y/Y?  What's different about any of those?  Failure to adhere to any one of them can just as easily lead to the same or worse outcome.

I think you would find real dispatching to be about as different from the simulation as the ocean is from a bathtub.  I mean no put-down by that as you would have no way of knowing.  The typical training regime for a new dispatcher until they have the ability to work on their own is 1,000 hours of training one-on-one with a highly experienced dispatcher at their side, plus 6-8 weeks classroom work, plus additional one-on-one time for each new console.  I hope you will agree that the "learn at home version" captures only a tiny fraction of that. 

RWM 

[spokyone]

RWM. You make some good points. In the event that I posted on the previous page, I do not know what the signal aspect was. As a passenger, I was happy to know that the ds advised us of an occupied siding, and that siding was barely long enough to accomodate both trains.
 RWM Do you work in dispatching?

[Railway Man]

RWM. You make some good points. In the event that I posted on the previous page, I do not know what the signal aspect was. As a passenger, I was happy to know that the ds advised us of an occupied siding, and that siding was barely long enough to accomodate both trains.
 RWM Do you work in dispatching?

No, and I thank God at least once daily.

RWM

[edblysard]

Here is what was drummed into my head during the first year out here....

 

Authority to occupy a track is granted by signal indication....protection is provided through timetable and rulebook.

 

In other words, the dispatcher can grant you authority to occupy track, verbally or with a signal indication, but they can not provide protection, it is up to you to "go looking".

[Railway Man]

Here is what was drummed into my head during the first year out here....

Authority to occupy a track is granted by signal indication....protection is provided through timetable and rulebook.

In other words, the dispatcher can grant you authority to occupy track, verbally or with a signal indication, but they can not provide protection, it is up to you to "go looking".

Amen!

[Krazykat112079]

Did anyone happen to catch what was in the Hyundai container that got torn open?  Just curious.

[Railway Man]

Did anyone happen to catch what was in the Hyundai container that got torn open?  Just curious.

Lead-based paint.

[zardoz]

If you want to prevent 99.99% of rear-end collisions like this, install....ATS....

ATS is a joke and is very easy to override.  Mostly useless (although it is helpful in dense fog).

[DMUinCT]

I posted Amtrak Photos that were downloaded to me under "Who Thinks Genesis #8 is totalled?" 

[CShaveRR]

If you want to prevent 99.99% of rear-end collisions like this, install....ATS....

ATS is a joke and is very easy to override.  Mostly useless (although it is helpful in dense fog).

With ATS, if the engineer had acknowledged the signal, and still misinterpreted it, the result would have been the same.

[Railway Man]

Points in favor of CBTC or ETMS, then.  To defeat those you have to work very hard at it.

RWM

[wabash1]

Points in favor of CBTC or ETMS, then.  To defeat those you have to work very hard at it.

RWM

why electronics why not knowing what your doing. if this man knew his territory he would not have speeded up its obvious that he was not current on this territory( leaves amtrak resposible)  and some statements earlier someone said why not have the dispatcher tell him, its not his job, I run through major hubs and run along with amtrak also, ive had dispatchers tell me the signal is comming up, and that i will be following one, its up to me to find his rear end, and at night and if its a black tank car it is hard to see with out a marker( now before you know it alls get started no you dont need a marker in yard limits as a cut of cars is not a train) and not all eot flash til moving or air is released.  what is not been said if it was abs or ctc.  if the signals were abs then it could have been anything displayed and been restricting, in abs in yard limits ( if this was the case) any signal other than clear is considered restricting.

[Semper Vaporo]

Maybe that kind'a explains why on the radio I hear a dishbasher say, "What'da'ya mean, you see a red signal indication?  You should have already left and been half way across the state!"

[chefjavier]

Did report of NTSB came out yet? I would like to know the debate to who's is right.

[Railway Man]

Check back this time next year.  Typical time for the NTSB report to appear is 9-24 months after the accident.  For example, the most recent rail accident report released was of WMATA derailment January 7, 2007, release date October 16, 2007. 

RWM 

[chefjavier]

Check back this time next year.  Typical time for the NTSB report to appear is 9-24 months after the accident.  For example, the most recent rail accident report released was of WMATA derailment January 7, 2007, release date October 16, 2007. 

RWM 

Thanks!