On Board Video Of a Head On. (Well, Close Enough)

[cordon]

Sorry, the middle one should end in "htm" instead of "pdf."  I went back and edited it.  Good catch - thank you.

I spent about 40 minutes searching for a more complete accident report.  Haven't found any yet.

With respect to the radio, What I'm thinking of is one train calling the other when they are a few miles away from the scheduled meet.  Or, if a crew has no hint of a meet until they see the first diverging signal, in this case the yellow over yellow at the beginning of the video, are they in the habit of calling to establish contact with the crew of the other train at that time? 

I'm thinking that maybe a radio call at that time might have alerted the SB crew when there was still time and distance available to stop their train before fouling the diverging route.

 

[Semper Vaporo]

...

http://www.ntsb.gov/publictn/2006/RAB0605.pdf

http://www.ntsb.gov/publictn/2007/RAR0701.pdf

http://www.ntsb.gov/publictn/2006/RAB0604.pdf

...

 

cordon: your middle link is "not found". 

ericsp: the last line ("Full story here") in the the link in the link in your second post is also dead. 

The other link refered to here is just a synopsis of several accidents and other reports.

I searched the NTSB site for "Kismet" and found only one document and it was about aircraft accidents (and it was Kismet Island and the whole report was severely censored, lots of blacked out text).

Does anyone know of a completed report available?

 

[ericsp]

I doubt the crew of the northbound tried to contact the crew of the M RICBAR. I have seen BNSF hold a train on the main with the second train going into the siding a few times. I have also noticed that many times the stopped train will leave its brights on. Just yesterday I saw a BNSF sitting in a siding with its lights on bright.  The FRA synopsis (http://www.fra.dot.gov/downloads/safety/120706trainaccsum.pdf) lists the M RICBAR's speed as 22 mph. If I remember correctly, the curve and rolling hills would have obstructed view until just a few hundred feet before the siding. Therefore, I doubt the northbound crew thought anything was wrong until the M RICBAR pasted its signal at the end of the siding.

If I remember correctly, the report said that the sun's position had no role in the wreck.

[cordon]

Yes, I read it.  The final report may have some additional information.  For example, we don't know yet whether or not the crew of the camera train tried to communicate by radio with the crew of the train that ran the signal.

Here are three other accidents with the common factor of happening early in the morning.

http://www.ntsb.gov/publictn/2006/RAB0605.pdf

http://www.ntsb.gov/publictn/2007/RAR0701.htm

http://www.ntsb.gov/publictn/2006/RAB0604.pdf

The first two are eerily similar to the accident we have been talking about.  In the first one the crew apparently drove for over an hour without taking any actions.  Curiously, the report does not say anything about an alerter (unless I missed it).

 

[ericsp]

If you will look at my second post (first page), you will find a link to the thread about the release of the report. There is no need to speculate, the investigation found the crew of the opposing train to be at fault. Has anybody read that?

[Poppa_Zit]

How to view this scintillating video:

Go to Google.

Click on "videos".

Use search terms "head-on" and "train".

At least one version will pop up. Last night, there were FOUR on YouTube.

PZ 

[BaltACD]

I know it can be...we had a engineer who could sleep, and I mean absolutely sound, snoring, head down drooling sleep, and blow the horn for every grade crossing...he had worked the same run for years, and the track condition kicked in his reflexes...the alerter would go off, and he would move the throttle a notch up and then back down, or press the button( flick the wisker), with out ever being awake.

Ironically, he died at home, in his sleep, of a massive aneurism.

 

But to answer the question...looking at the smoke coming from the wheels, I would guess, and it is only a guess, they plugged it about the same moment they passed the last signal post.

HHMMM, did the DS not notify the crew of the meet, so the red was totaly unexpected? Admitilaly, all crews chould always expect a red red, but come on! Why did the crew overun the signal, noones answerewed that! DS error?

Unless the Dispatcher took down the signal for the approacing train (and there is a normal 8-10 minute time out between taking down a signal and being able to change a route and re-line a signal - additionall before taking down a signal that has been lined for a train that is in a position to accept the signal, the dispatcher is required to contact the crew and get a confirmation from them that they can make a safe stop without passing the signal to be taken down.) the Dispatcher is in the clear.  The non-camera train is at total fault. 

While given the hour of occurence, a sleeping crew is the most likely cause, other causes that have occured in similar accidents have ranged from the crew no knowing 'exactly' where they were; being engrossed in a converstion within the cab and not paying attention to the wayside signals.

[Railfan1]

Does anyone know where the video can be currently viewed?  I didn't get to this post in time to see the video.

I'm pretty sure if you go to Youtube, it'll be there somewhere.

[CNW 6000]

Thanks for the explanation again Ed.  That last one makes everything perfectly clear.

[ChuckCobleigh]

No, opposing train crew error.

The crew on the video train already knew of a meet...they had several restricting signals slowing them down in preparation to taking the siding.

 

etc 

 

Thanks, Ed...another detailed explanation of how the signaling works towards preventing this sort of "meet" on the rails.  Of course, a system can be designed with all of the safeguards in place and still rely on the human factor for the final step.  A better (from a performance standpoint) solution is positive train control, I guess, but that is not likely fail-safe either; it's just an added redundancy to the human factor, at an additional cost.

I'm just glad you've got the time and interest to share with us on the forum; we're too impatient to wait for the next magazine piece. 

[OldBNfan]

Does anyone know where the video can be currently viewed?  I didn't get to this post in time to see the video.

[edblysard]

No, opposing train crew error.

The crew on the video train already knew of a meet...they had several restricting signals slowing them down in preparation to taking the siding.

The only reason they would be getting these signals would be a train ahead, be it moving in the same direction they are, and the system is slowing them down because the other train is still in a block close enough to trigger the system, or a meet

Because most crews are quite familiar with their territory, I would bet they already figured out a face to face meet was on...

This looks like a ABS system with a CTC overlay.

The opposing train would have had signals slowing them down to a stop.

The system is automatic; it will not show a green, followed by a red unless there is a problem, like a broken rail.

Red signals are not unexpected; the system is designed to progressively restrict your speed and movement.

The control operator can not arbitrarily change a signal aspect; the system will not allow him to.

In fact, if he does have to change a signal, under normal circumstances he has to time out that signal, reset it, and wait ten minutes.

This prevents him from making mistakes.

It works on a failsafe system, so for the opposing train to have an absolute red at the last signal, they had to have at the least two signals more restrictive than green before they arrived at this point.

The system can not tell the control operator where in a block, or how fast a train moving...only that a block is occupied.

From his desk, all the control operator would have seen is the opposing train enter the last block.

He would have no idea if the train in that block was sitting still at the end of the block, or if it had just entered the block at the far end...only that he was in the block.

With the turnout lined for the siding, the opposing train would have been facing an absolute stop signal at the signal post just past the turnout.

Meets where trains are looking at each other this close are normal, happens hundreds of times a day, no more odd than you pulling up to a stop light in your car, and if everybody follows the rules of the road, nothing happens.

The other crew, for what ever reason, ignored the signals they had.

When they ran past the absolute signal, (the one on the left of the screen, right about where the other crew bails) there was what, two or three seconds before impact?

What time did the control operator have to warn the other crew?

Note the signal we are facing goes red over red the instant the other train violates our block...about a two or three seconds response time left before impact.

 

All the control operator would have seen on his monitor was the opposing train enter the last block before the switch, and our train enter the block on this side of the switch.

He would have no idea where in the block these train were, and because he had to have lined the turnout while our train was at least two blocks away from the last block, everything would look normal on his monitor.

The system makes it very hard for the control operator to line a switch with the train in the same block as the switch...on purpose.

This keeps an operator from lining a switch under a train, or making a decision thats wrong...he couldn't have lined the switch with either train in the last block, so he had to have had it all line up while both of them were several blocks away from this point, again, normal operating practice.

Once he lined the switch, the ABS system would have set all the signals for a meet, slowing the opposing train with successively restrictive signals down to a absolute stop, and informing our train they were going to take a siding, which the last signal they see, just before the switch, confirms...its indication is diverging route, at timetable speed for the turnout, prepared to stop at the next signal, which would be at the far end of the siding.

My guess is the opposing train was long enough to foul the block behind, and foul the switch at that end, so the system would have read two blocks occupied by this train, and would have signaled to stop our train in the siding because the other trains rear is hanging out past the switch at the far end.

 

Here is my guess, and it is only a guess.

The opposing train crew knew there was a meet, they would have gotten a signal other than high green somewhere behind this point, and meets are common there, (that's why there is a siding there in the first place), so they would have slow rolled it, trying to time it so they roll up on the far end (this end) of the siding just as the other train (us) clears the switch.

Probably have done this a few hundred times before, its normal for crews to not want to stop unless they have to.

 

Our crew would know there was a meet, for the same reason, they were getting signals other than high green, and like the other crew, were used to meeting trains here...nothing out of the ordinary.

Most railroad working under GCOR use signals for routing purpose, then speed control, because the density of trains does not require them to run trains on each other tails, so the main concern is directing trains to their destination, not spacing them out like dense eastern roads have to.

 

From the video, you can hear and see our train with dynamic brakes slowing (at least it sounds like it, and I hear a air brake application happening also) , prepared to take the siding as he comes around the curve...he sees the signal directing him into the siding, prepared to stop at the far end of the siding, and he sees the other train.

He assumes the other train is standing still, after all, he meets trains here all the time.

It isn't until he is a few hundred yards away and the other train has not dimmed his headlight that we realize he is still moving, and by then it is too late for us to take any preventive action.

GCOR rules tell engineers of trains stopped in sidings or on the main to dim their headlight, extinguish their ditch lights, as a visual way to inform other trains they are standing still, but, just like the auto driver on the way to work with his high beams on, sometimes they forget...not often, but it happens.

Nothing really out of the ordinary in the video right up to the last few seconds, so there is nothing to warn our crew that an accident is going to happen.

 

Again, as a guess, the other crew was slow rolling towards the meet, nodded off, woke up at this end of the siding, realized they were never going to stop in time, plugged the train and bailed out...it looks like both guys jump at the same time, but I am not sure, the video is poor...the guy you see jumping is either holding a big bag, (grip) or is pushing the other crew member ahead of him, and I doubt he would have taken time to grab his bag.

 

All the control operator would have seen is the block occupancy showing the last main block occupied, and the block our train is in occupied, with the turnout indicator showing our train routed into the siding, perfectly normal.

Odds are he never spoke to either crew, there is no need when you are proceeding on signal indication...the only reason he would talk to either crew is if something was wrong with the system, or something out of the ordinary showed up and he had to flag them through a signal.

It is not uncommon for crews to run hours on hours without ever speaking to an operator...I routinely take a train from our North Yard all the way to Barbors Cut with out speaking to anyone on the radio, all on signal indication.

[snagletooth]

I know it can be...we had a engineer who could sleep, and I mean absolutely sound, snoring, head down drooling sleep, and blow the horn for every grade crossing...he had worked the same run for years, and the track condition kicked in his reflexes...the alerter would go off, and he would move the throttle a notch up and then back down, or press the button( flick the wisker), with out ever being awake.

Ironically, he died at home, in his sleep, of a massive aneurism.

 

But to answer the question...looking at the smoke coming from the wheels, I would guess, and it is only a guess, they plugged it about the same moment they passed the last signal post.

HHMMM, did the DS not notify the crew of the meet, so the red was totaly unexpected? Admitilaly, all crews chould always expect a red red, but come on! Why did the crew overun the signal, noones answerewed that! DS error?

[Poppa_Zit]

now THAT was cool! the train approaching the camera train appeared to not stop and wait for the camera train to take the siding. the video showed the other train advancing fowards while the track was set for the siding. basically, i'm saying the other train was at fault for not stopping at the designated place. through out the entire video, it shows no other alternate tracks, so the video train couldn't have been at fault.

Thanks for clearing it up for us.

[edblysard]

I know it can be...we had a engineer who could sleep, and I mean absolutely sound, snoring, head down drooling sleep, and blow the horn for every grade crossing...he had worked the same run for years, and the track condition kicked in his reflexes...the alerter would go off, and he would move the throttle a notch up and then back down, or press the button( flick the wisker), with out ever being awake.

Ironically, he died at home, in his sleep, of a massive aneurism.

 

But to answer the question...looking at the smoke coming from the wheels, I would guess, and it is only a guess, they plugged it about the same moment they passed the last signal post.

[ValleyX]

If the brakestand is set up for trail or tow, it deactivates the alerter.  Shutting it off in the cabinet will not work if the unit is set up for lead (controlling) operation.

As for dumping the air, obviously not nearly soon enough.

I firmly believe that the alerter reset can be hit while in a semi-comatose state, sleepwalking, so to speak.

[Anonymous]

 

Watch it again, you can see the opposing train trip the signal system and our signal aspect changes from red over green, (diverging route) to red over red, as it should once someone ocupy the circut, which the opposing train does just as he passes the signal mast on the other side of the switch.

I thought I was seeing, at first, a green aspect on that signal the first train passed, but when the camera got closer, it looked like red over red.  I did not start/stop to check it out, but just concluded that it was red all along.  Your explanation that it changed by the violation of the oppsing train makes sence.  But even if that signal did go red right in their face, would they not have been required to see it and stop, even if they could not stop short of it?  Don't they have to watch and react to every signal right up to the point where they pass it and can no longer see it?

That brings up another question.  Wouldn't the dispatcher see that the first train passed a red board and communicate the error to the crews of both trains?  When the first train got the red signal in their face, did that mean the other train was passing a red board at that moment?  Wouldn't the dispatcher have seen this whole thing unfolding? 

after watching the video many times.. as fast as that all happend..there was NO WAY anything could have been done in time by anyone but the crews.. the meet was "perfict" so to speak for a headon in the middle of the plant.. the signal on the train with the cam went from medium clear to red about 2 seconds befor the locomotive passed it..that is not enought time to stop unless he was going 1mph.. and he was going alittle faster then that.. and as far as the dispatcher seeing it unfolding..he saw it unfold as soon as the system went "out of corrisondance" as the derailing trains destoryed the wireing and track circits that ran the system at that interlocking.. if the train with the cam on it was a few miles away from the plant.. then there might ahve been some time to stop both trains in time but there was nothing anyone could have done to warn anyone in time for this situation..

csx engineer 

I do not know what the conclusions of the investigation were, but I wonder where the train opposing the camera was when they set the brakes.  It raises one other question, but I won't speculate before others comment.  I think it is interesting to understand exactly how these things happen, and what options are available as time runs out.

I suspect you are wondering why the alerter didn't wake them up...

It can be disabled with a breaker in the rear cab wall, in front of the electrical cabinet.

You have to have a way to disable it so you can move the locomotive as dead in tow or for maintenance purposes.

Get caught with it turned off and your fired outright, no questions allowed nor any argument entertained.

Yes, the black box would note the alerter was off...you can be sure both the engineer and conductor, if they survived, won't work in the industry again.

I am not familar with the alerter, but since you mentioned it, I am curious about what role it might have played.  What I was wondering about in the earlier post was the amount of time between the point where the opposing crew put the brakes into emergency and the point where they ran past the stop signal.  

[edblysard]

 

Watch it again, you can see the opposing train trip the signal system and our signal aspect changes from red over green, (diverging route) to red over red, as it should once someone ocupy the circut, which the opposing train does just as he passes the signal mast on the other side of the switch.

I thought I was seeing, at first, a green aspect on that signal the first train passed, but when the camera got closer, it looked like red over red.  I did not start/stop to check it out, but just concluded that it was red all along.  Your explanation that it changed by the violation of the oppsing train makes sence.  But even if that signal did go red right in their face, would they not have been required to see it and stop, even if they could not stop short of it?  Don't they have to watch and react to every signal right up to the point where they pass it and can no longer see it?

That brings up another question.  Wouldn't the dispatcher see that the first train passed a red board and communicate the error to the crews of both trains?  When the first train got the red signal in their face, did that mean the other train was passing a red board at that moment?  Wouldn't the dispatcher have seen this whole thing unfolding? 

after watching the video many times.. as fast as that all happend..there was NO WAY anything could have been done in time by anyone but the crews.. the meet was "perfict" so to speak for a headon in the middle of the plant.. the signal on the train with the cam went from medium clear to red about 2 seconds befor the locomotive passed it..that is not enought time to stop unless he was going 1mph.. and he was going alittle faster then that.. and as far as the dispatcher seeing it unfolding..he saw it unfold as soon as the system went "out of corrisondance" as the derailing trains destoryed the wireing and track circits that ran the system at that interlocking.. if the train with the cam on it was a few miles away from the plant.. then there might ahve been some time to stop both trains in time but there was nothing anyone could have done to warn anyone in time for this situation..

csx engineer 

I do not know what the conclusions of the investigation were, but I wonder where the train opposing the camera was when they set the brakes.  It raises one other question, but I won't speculate before others comment.  I think it is interesting to understand exactly how these things happen, and what options are available as time runs out.

I suspect you are wondering why the alerter didn't wake them up...

It can be disabled with a breaker in the rear cab wall, in front of the electrical cabinet.

You have to have a way to disable it so you can move the locomotive as dead in tow or for maintenance purposes.

Get caught with it turned off and your fired outright, no questions allowed nor any argument entertained.

Yes, the black box would note the alerter was off...you can be sure both the engineer and conductor, if they survived, won't work in the industry again.

[Big Ed]

BNSF 997 was the trailing motor on that M-RICBAR southbound Kismet collision which had 7 units (train w/o the camera) and its cab was mangled up pretty good with all the run in similar to what Jeff & Brian explained.

Not sure if it's back in service or not.

[Anonymous]

I read an account of a wreck at Waco, Co., on the Salida Division of the D&RGW, where a steam locomotive on a passenger train was being ridden by a traveling engineer to learn why the train had been having trouble making the time.  Under the encouragement of the traveling engineer, they were moving at least as fast as normal on straight track.  The fireman told the engineer that he would not make the curve at Waco unless he set some air.  The traveling engineer overruled the fireman, so he simply stepped through the gangway, climbed down the steps, and got off.  Can you imagine the conviction it would take to make the decision to step off a speeding passenger train on the gamble that it won't make a curve that is not even in sight yet?

The fireman won the bet and survived getting off at speed.  The three-cylinder 4-8-2 did not make the curve.  Instead, it leaped across the Arkansas River, killing the engineer, the traveling engineer, and several passengers.

[CNW 6000]

Those look like they'd hurt!

[blhanel]

The third was the rear ender we had by Blairstown, IA a few years ago.  It occured about 28MPH IIRC.  No one was killed, but the engineer is still off hurt.  The conductor had minor injuries and I've since worked with him.  The fireman (trainee) jumped and was also injured.  The fireman resigned afterwards, once was enough for him.  The engineer and conductor thought he was dead because they couldn't find him after the resulting derailment.  He had been able to walk thru a field to a highway and got a ride into town and to medical help. 

A picture of the collision point on that one...

The trailing engine on their train faired worse than the one they were in- several of their autoracks successively battered it pretty good.

 

[Chris30]

The video is scary because everything looks normal from the train with the camera view until almost the very end. Perhaps if the curve wasn't there the crew on the train with the camera might have seen what was going on (headlights not dimmed, etc.) sooner and the outcome somehow different.

As for the debate about riding it out or jumping I don't like either option. Maybe a James Bond like ejection seat with a parachute would be be a better option.

CC

[jeffhergert]

  I think the decision for jumping would depend on, A. Being at a speed that allows you to survive the jump itself and B. Being able to jump far enough away from the impact that the debris doesn't kill you.

 Three collisions come to mind.  One was a 17mph rear ender at Clinton in the IMRL where the crew died.  They struck the rear end of a local and the last car, an empty tank rode up over the cab.  I don't remember exactly where the conductor ended up but the engineer's body was hanging out his side window.  The ice bucket on the floor at the back of the cab was touched.

 The second was on the IAIS about 20 or so years ago near Altoona, IA.  Still operating under time table and train orders, the eastbound train had an order to wait at Altoona until 1215PM.  They overlooked the order and struck head on a westbound local about 12noon.  (This has been a while so the times may not be exact.  Suffice it to say the westbound could've made it to Altoona to clear the time in the wait order by the required 5 minutes.)

 The crew on the eastbound saw the local approaching on a curve and the crew jumped and survived.  The westbound, running long hood forward, stayed on and were killed.  I've often wondered if the engine running backwards on the local contributed.  Possibly the crew didn't see the other train in time to jump.

 The third was the rear ender we had by Blairstown, IA a few years ago.  It occured about 28MPH IIRC.  No one was killed, but the engineer is still off hurt.  The conductor had minor injuries and I've since worked with him.  The fireman (trainee) jumped and was also injured.  The fireman resigned afterwards, once was enough for him.  The engineer and conductor thought he was dead because they couldn't find him after the resulting derailment.  He had been able to walk thru a field to a highway and got a ride into town and to medical help. 

 Three diffenerent accidents with different results.  You just have to go with your gut feelings on the spur of the moment and hope you choose wisely.  I would still say on a modern engine, riding it out would usually be the best option.

Jeff    

[Anonymous]

  I doubt that there will ever be an official policy or rule to cover the choice of riding it out or jumping.  It would add more liability on the company if some one died following that policy or rule. 

  In case of doubt or uncertainty, take the safe course.  Getting on or off moving equipment is prohibited except in emergencies. 

  There's the policy, take your choice. 

            Jeff

The liability concern makes a lot of sense, and I guess the "take the safe course" rule would cover it alright.

[CPRail modeler]

now THAT was cool! the train approaching the camera train appeared to not stop and wait for the camera train to take the siding. the video showed the other train advancing fowards while the track was set for the siding. basically, i'm saying the other train was at fault for not stopping at the designated place. through out the entire video, it shows no other alternate tracks, so the video train couldn't have been at fault.

BTW, what was the time (in the video) that the engineer jumped from the other train?

[Anonymous]

  The safety meeting was about not getting into stuation like this.  The railroad tries to teach you NOT to get into situations where you have to choose between riding it out and jumping. 

  I think it depends on the situation, but the safest thing would probably be to get on the cab floor as close to the back wall as possible and try to ride it out.  Sadly there will be situations that will most likely be deadly no matter what you do.  The best thing is to not get into them at all.

  At the meeting, they slowed down the video and the last thing you see before it goes black is the front windshield shattering.  The audio continues a bit longer, but as stated before the microphone is in a different location so may have survived a bit longer.

Jeff  

Jeff,

I understand that it would be preferable to try to prevent collisions, but I just wanted to know if railroads provided any advice on what to do if a collision was imminent.  The crew on the video train, for instance, did everything proper to prevent a collision, but nevertheless, were confronted with one.  I have read accounts of collisions where crew members either stayed on the engine or jumped, with every imaginable result, and many were the opposite of what one would expect.  It is interesting because it is only railroad transport that offers the often attracive option of jumping because the same momentum that makes a train hard to stop, also can give a long warning before a collision. 

[jeffhergert]

  I doubt that there will ever be an official policy or rule to cover the choice of riding it out or jumping.  It would add more liability on the company if some one died following that policy or rule. 

  In case of doubt or uncertainty, take the safe course.  Getting on or off moving equipment is prohibited except in emergencies. 

  There's the policy, take your choice. 

            Jeff

[mikeyuhas]

I find myself wondering:

a. How the camera survived the impact, when most of the locomotive was destroyed.

b. Since the camera somehow did survive (along with post-impact audio (in tree68's link)), where is the post-impact video--it seems as though the actual wreck was edited out.

 

Zardoz,

You may have answered your own question -- clearly, the camera itself didn't survive (or if it did, the connection between it and the recording box was severed). The microphone still picked up the sounds of impact long after the camera itself was toast. This proves that the digital recorder electronics "black box" is bulletproof, and possibly the location of the recorder in the locomotive played a role in its survival.

  

[jeffhergert]

  The safety meeting was about not getting into stuation like this.  The railroad tries to teach you NOT to get into situations where you have to choose between riding it out and jumping. 

  I think it depends on the situation, but the safest thing would probably be to get on the cab floor as close to the back wall as possible and try to ride it out.  Sadly there will be situations that will most likely be deadly no matter what you do.  The best thing is to not get into them at all.

  At the meeting, they slowed down the video and the last thing you see before it goes black is the front windshield shattering.  The audio continues a bit longer, but as stated before the microphone is in a different location so may have survived a bit longer.

Jeff