henry6The more I hear about this incident in particular, and the WMATA in general, the I fear that the software has reached an age of senility....like "Vyger" in the first Star Trek movie...out working for so long it forgot why it was out there and what it was supposed to be doing. All kidding aside, though, is there a point when software will go soft, suffer from spasmadic loss of memory or have complete black outs if ony because of its age? Or don't we really know yet?
The more I hear about this incident in particular, and the WMATA in general, the I fear that the software has reached an age of senility....like "Vyger" in the first Star Trek movie...out working for so long it forgot why it was out there and what it was supposed to be doing. All kidding aside, though, is there a point when software will go soft, suffer from spasmadic loss of memory or have complete black outs if ony because of its age? Or don't we really know yet?
Rgds IGN
Local media today is reporting that the train operator has been terminated for passing the red signal. It's sad for her since she had been an employee since 1976 and a train operator since 1999. I think this speaks more to Metro's training program, or lack thereof, than anything else.
http://wtopnews.com/?nid=25&sid=1911122
We are being given to understand now that the operator was instructed by Central Control to pass the red signal. In some other systems this causes a mechanism to stop the train in its tracks, so to speak, but Metrorail evidently doesn't have that feature.
In addition, there have been several conflicting media reports that Metrorail has or has not been running in "manual" mode since the June, 2009 accident. I, for one, did not know they had enough qualified operators to run regular service in "manual" mode 100% of the time in revenue service, since they are chiefly door operators.
Red Signals are RED for a purpose....to stop the movement of the trains viewing the signal.
If 'automatic' operation of the train caused it to pass the RED signal, the software controlling the 'automatic' operation has failed.
If the train was being 'manually' operated, then the operator of the train has failed in his responsibilities.
Red signals are not to be passed until permission is received from the party that actually controls the signal.
Never too old to have a happy childhood!
We still do not know what really happened. As a commuter on Metrorail for several years I have never experienced routing onto a pocket track that then stopped. In my experience riding lots of rush and non-rush trains the trains always come "out" of pocket tracks empty for skip-train service and never go "in" with passengers except when "single-tracking" at which time they come out the other end on the "other" track. This only happens when something is in the way (train with malfunctions or a sick passenger, the latter being much more frequent).
Maybe Central and the Operator expected this to be a single-tracking situation when it really wasn't.
From Trains News Wire (links in the article are probably subscriber only):
Report: Washington Metro train was automatically derailed to prevent crash Published: Monday, February 15, 2010 WASHINGTON — A safety system deliberately derailed a Metro Red Line train leaving the Farragut Station Friday after the train entered the wrong track, the Washington Post has reported. The 10:15 a.m. derailment trapped 345 passengers underground aboard the train for an hour and snarled the morning commute.The derailment marks the fourth open National Transportation Safety Board investigation into Metro wrecks. The most high profile incident, which occurred last June 22, left nine people dead, and involved a supposedly failsafe control system that's designed to prevent two trains from occupying the same track space simultaneously.In Friday's wreck, sources told the Post at least two human errors appear to be involved. First, controllers in Metro's downtown control center routed the train into a "pocket track," a third track between the two main tracks that's used for crossover movements. Second, the Metro operator apparently passed a red signal on the main track."Part of the investigation will be to determine how the train came to be on the pocket track," said Metro spokeswoman Lisa Farbstein.As the train attempted to proceed northward out of the pocket track, the system detected a movement that should not have been occurring, and automatically derailed the train. That move may have prevented the train from entering into the path of southbound trains.
James
With possibile soft ware problems maybe the operator expected the train to stop at the red signal and when he realized it wasn't hit the emergency stop but for whatever reason the train hit the derails. Any info how fast the train hit the derails.
> is there a point when software will go soft, suffer from spasmadic loss of memory or have complete black outs if ony because of its age?
Basically, yes. Software is rarely "static". It goes through revisions. In the normal creation of software, errors (bugs) are found and they are characterized at four (sometimes five) levels:
1. Cosmetic error or typo in the user interface, data log, etc.
2. Error that can be worked around. For example, the print button might not work, but Alt-P does.
3. Something that just plain does not work, and no work around. (But does not crash system)
4. Something that crashes system.
Sometimes a Sev 0 (Severity 0) "error" is the addition of a feature. This is not "bad", it's just a means of using the bug tracking system to document someone re-reading the spec, or a customer request.
In development, the software department does a build on a regular basis. This means every piece of code is recompiled, not just the stuff that got modified (a "make"). Regular means daily or weekly, it is common for a project to have more than a few hundred builds. Often, only Sev 3 and Sev 4 bugs get fixed, but, it is important to fix the Sev 2 bugs (a "Silver build") and even the Sev 1 bugs (a "Gold" build).
Here's why. Every time you change something, you change everything else a little bit. Correcting typos often lengthens or shortens the text, and moves code over one byte or five. Typically, 50 typos cause one Sev 2 error. Fixing 10 or 20 Sev 2 errors causes one Sev 3. Fixing 10 Sev 3 errors causes one Sev 4 (crash). These are all theoretical, experiential numbers, your mileage will vary.
Software runs on hardware. Hardware rarely stays the same, and, initially, hardware failures are individual pointwise events. Eventually, you will have two or three events at the same time. The software has never seen that combination.
There are individual quirks in software. On an automatic train, it is possible, at a few spots in the system, to identify "weekday", "Saturday" and "Sunday" software, just from the unusual behavior and timing observed. Trains get old and "sticky" (they stop a foot short) or slippery (a foot long). Add it all up and you get an incident.
RIDEWITHMEHENRY is the name for our almost monthly day of riding trains and transit in either the NYCity or Philadelphia areas including all commuter lines, Amtrak, subways, light rail and trolleys, bus and ferries when warranted. No fees, just let us know you want to join the ride and pay your fares. Ask to be on our email list or find us on FB as RIDEWITHMEHENRY (all caps) to get descriptions of each outing.
Most likely this derail is a permanent derail; the typical configuration of a pocket track is to have a derail at both ends. Most likely a pocket track will be the length of the longest train (8 cars for WMATA?)
I'm more familiar with BART, similar to WMATA. Pocket tracks are used frequently, so an operator should not be surprised if they wind up on one. Most likely, the train was automatically routed to the pocket track, inconvenient, but not dangerous (if the derails were up, or the pocket track was already occupied, the train would have automatically stopped before entering.) Mis-routing is not rare, and, if the operator is quick enough to stop the train, he can ask for a correct route. You don't have to be close enough to see the switch, you can infer a mis-route from speeds. If you are at a point where you expect 70 mph and have 36 or 27, you can guess that you are doing something different. But if the operator does not catch it quick enough, he just goes to the other end and pulls it back until dispatch says "clear" (from looking at a track circuit), then you go back and do it right. (You can't "back up" these trains - you don't slam it in reverse and go a thousand feet using your mirrors.)
Like automatic switches, automatic derails can be manually cranked or moved.
My initial thoughts are that this train was automatically misrouted to the pocket track, the dispatcher could not get a clear automatic route back to mainline, the operator was given manual orders to proceed, was given permission to run the red signal, looked out and saw a good switch route, and proceeded over the derail. I believe WMATA uses married pairs, which is why he left two cars, and uncoupled and took four good ones out the other end of the pocket track?
Was this a manually installed temporary derail, the kind the Mechanical and/or MofW personnel use from time to time to protect their on track activities....or was it a derail that is interlocked with the signal system?
We do not know, yet, just like we do not know why nine people died last summer. I think this extreme transparency is a new tactic to try to understand Metrorail's problems (remember General Manager Cato is quitting).
I must emphasize that operators do not control the train in this territory. It runs by itself. If the train passed a signal (which only says "Go," "Go Slowly," or "Stop") it would be too late for the operator to react.
The train did, indeed hit a derail. This is a two-way "pocket" track UNDERGROUND. These are used all the time for skip-train service and "single-tracking." Skip-train is why savvy commuters like me wait for the third Orange Line train at Metro Center because it comes out of a pocket track and its relatively empty. "Single-tracking" is how Metro performs maintenance and it is exactly what you think it is. Sometimes they use the crossover but in some locations they have to use the pocket track as the crossover. I had originally assumed that most of these 'pocket tracks' are actually stub-ended tunnels underground (but there are actually both types in use).
Let's play "jump to conclusions." Given the insane maintenance worker problems recently (involving deaths), it would appear that the derail was not known to be in place by Central Control. Exactly why a train with passengers would ENTER a pocket track is the second mystery. They *never* do that except in single-tracking situations where the crossover can only be done through the pocket track.
Metrorail needs to become more transparent. I still cannot believe we do not know for sure why nine people had to die last summer and exactly how ARINC is getting money to design and install a system "never before implemented."
Metrorail is in a very sad state today.
Controls are always on the right side.
Operators control the doors from BOTH sides. There is a button in a cabinet on the left side.
The operator does not make the train go. The train goes after the doors close.
The accident was underground.
Today's Wash Post has a diagram of the derailment. Evidently for some reason after leaving the Farragut north station; central control routed the train onto the pocket track and the train proceeded to the far north end and hit the derailers.
This pocket track is a full pocket track with outlets at both ends to both tracks. Evidently there are automatic derails at the north end that retract when route is clear.
The investigation will have to answer these several questions and more
1. Why was a revenue train routed onto the pocket track? Control center has to answer that one.
2. Since the train was routed onto the pocket track did the south end pocket track derails (if installed) retrack as designed?
3. What is the usable length of this pocket track?
4. Are the signals approaching any switch speed only indications or both speed and routing info? Anyone who has ridden it know? Does metro signals give a leave station signal only are they a combination signal?
5. Was the train in automatic or manual control?
6. What are the procedures for closing the doors? Operator evidently was on left side as the station is a center platform station.
7. What side of the operating cabin are the controls?
8. Once the doors are closed can the operator start the train forward with out returning to the operator's chair? That is a feature on many subways I have ridden.
9. Could the operator have missed seeing the signal if operator not in seat?.
10.. Was it snowing?
11. Since there is a lot of snow on the ground did the operator not realize he/she was on the pocket track?
12. Had pocket track been used since snow and if not were rails coverd by snow?
13. Is pocket track automatic or manual control only?
14. If operator was thinking train on main track did he ignore red signal on end of pocket track until too late?
15. Did operator apply emergency brakes?
16. Toward which track was the switch at the north exit of the pocket track aligned?
17. Was the main track switch that the north pocket track exit switch was aligned toward remain aligned for the main track?
18. If the main track switch was aligned for main track did not the derails operate as designed to prevent train from entering a main track and destroying the switch?
19. Was train moved after derailment?
20. After about hour passengers were moved to trailing cars which were cut from married pair at front and returned to the station?
As we can see all these facts and probably more will need to be answered before any cause can even be proposed.
TomDiehl BNSFwatcher Methinks it is time to privatize most, if not all, municipally-owned transit systems. "Politically-Correct" hiring is not doing the job and is a peril to public safety. DC Metro is a case-in-point! Want more info? Text me, while you are on duty!!! Hays Methinks you need to research your remarks a little better. The operator that caused the accident while texting in California was an employee of a contractor, not a transit authority employee. The "privatize" BS in most cases lowers costs temporarily while the contractor secures the position, and is lowered by hiring marginal employees at best. As soon as they're secured, the price quickly creeps up. Then you have a situation of higher cost and lower qualifications.
BNSFwatcher Methinks it is time to privatize most, if not all, municipally-owned transit systems. "Politically-Correct" hiring is not doing the job and is a peril to public safety. DC Metro is a case-in-point! Want more info? Text me, while you are on duty!!! Hays
Methinks it is time to privatize most, if not all, municipally-owned transit systems. "Politically-Correct" hiring is not doing the job and is a peril to public safety. DC Metro is a case-in-point! Want more info? Text me, while you are on duty!!!
Hays
Methinks you need to research your remarks a little better. The operator that caused the accident while texting in California was an employee of a contractor, not a transit authority employee. The "privatize" BS in most cases lowers costs temporarily while the contractor secures the position, and is lowered by hiring marginal employees at best. As soon as they're secured, the price quickly creeps up. Then you have a situation of higher cost and lower qualifications.
I can't say I agree with either of you guys, however, I have noticed that there is always an increase in efficiency and lowers costs whenever a transit system changes hands in either direction: from private to public or public to private. Sooner or later it falters and there is another shuffle back to whence it came; and then it cycle through again. Evidently, private stockholders get a property, fix it up, run it, don't reinvest, run it into the ground, then sells or turns over to public entity. Public entity gets it, fixes it, has marginal success, sleeps on it, becomes political football, then is handed off to a private concern and the cycling continues. The one fact that has to be considered through all of it is that it appears transit services are needed and used the service is actually maintained and provided, then slackens off as service deteriorates and picks up again when service is better. To get rid of it would be suicidal to society and possibley even more costly than without its benifets. At present the new Governor of New Jersey is threatening to pull the plug on much of NJT'S funding without looking at the cost of not providing the service. Because of the interactivity of lines and services, it would be difficult, costly, and probably not operationally efficient should services be shut down, eliminated, or curtailed too much...and I'm talking rail, light rail, and bus vs. private vehicles or no vehicles (those who are taken out of the work force due to not being able to travel to and from jobs. In an unemployement benifet case on Long Island some 30 or 40 years ago a judge denied benefits to a person because there were jobs up to two hours distant accessable by public transit. I feel this judge has sort of set a standard of service and distance which a transit system must provide in whole or in part). Transit, we've got a problem.
Derails are installed at many locations to protect main track movements and prevent even bigger derailment or collision situation.
What I would like to know is how the operator got into the Pocket Track in the first place? Did the Metro Dispatching system line his equipment there on purpose or in error. If it was lined on purpose, obviously the operator was not informed of the purpose and did not communicate with the proper authorities to get permission to exit the Pocket Track. If a train (railroad or transit operation) gets to a place it doesn't belong....they must, repeat MUST contact the party(s) in control of the territory to obtain proper protection and authority for their next move(s).
The first thing to remember is the derails are put in place on ALL rail operations for the purpose of protecting the 'greater good'. The second thing to remember is that derails do a very good job.
Knowing the name of a device, knowing the definition of the device, and knowing the application of the device are three things journalists are not taught to put together, especially when it comes to railroading. Yes, the divice is calle a "derail" and by definition it is designed to derail a train, remove it from the track. But the application is that it is a device usually found before the foul point (where trains, cars, or locomotivis would be likely to strike each other while on seperate tracks) on sidings to protect cars, trains, or locomotives from going beyond the fould point and/or enter a main track. It is manually or elctrically controlled by the train crew or dispatcher/operator. Yes, in effect, it is designed to purposely derail a train. But is that what really happend here? Did Metro actually set out to purposely derail the train? I really don't think so. Opinionated theory only: the train somehow was switched into the so called pocket track; the motorperson, realizing the wrong route, stopped the train and attempted to back out; apparently the motorperson was not aware of the derail was still in place and thus the derailing. The whys and hows and faults and communications and miscommunications and judgements and misjudgements are all facts which will be determined and are not part of my opinion.
Saw this headline and story on DRUDGE this afternoon, was stunned that the METRO Authority would admit this and allow such a charge to get out. Their lawyers and risk managers must not have made it int to work, due to the snow storm.
WTOP.com
"Source: Metro train intentionally derailed
The train operator entered a section of the tracks known as the pocket track - an area that's located between the two main line tracks.
The operator was not supposed to be on the pocket track. When the train tried to get back on the main line, a piece of equipment known as a derailer took the train off the tracks intentionally.
The derailer took the train off the tracks because it was about to run a red signal -- something that could have caused a collision.
The National Transportation Safety Board is launching an investigation into the derailment. Metro officials also notified the Tri-State Oversight Committee.
About 350 passengers were on board when the front two cars of the train popped off the track as it headed toward Shady Grove.
Three people suffered minor injuries, D.C. Fire Chief Dennis Rubin says. One person was taken to George Washington University Hospital.
The Farragut North Metro station was closed for about two hours after the 10:15 a.m. derailment. The station reopened around 12:10 p.m. and trains are running on both sides of the track.
People should expect extensive delays on the Red Line.
Trains will be restricted to a speed of 25 mph between the Dupont Circle and Farragut North stations while Metro officials investigate the incident.
Metro spokesperson Lisa Farbstein says the passengers in the first two cars of the train were moved to the back four cars.
"The customers who were on this sort of mini-train were taken to the platform where they were able to exit the train," Farbstein tells WTOP.
The train derailed on the first day back to work for thousands of federal workers after the region's snow storms.
Dana Mozie was a passenger on the derailed train.
"In my head my alarms went off," he tells Channel 7. "We just left the station. Why are we stopping so fast?"
Soon afterward, he says passengers heard an announcement from Metro: "We just derailed."
"I think folks were actually in shock. For at least a good 15 to 20 minutes, no one said anything. No one smiled. No one even moved. No one brought out a telephone."
Mozie says the operator then came on and said, "I apologize."
But after all the excitement, Mozie says passengers were able to breathe a collective sigh of relief. They later joked a bit and traded Metro stories.
Speaking outside the station, passenger Wendy Bogart, 51, of Washington says as the train arrived at the station the lights and fan went off, but the train operator said "Don't panic... We have to re-energize the third rail." She says she felt nothing.
WTOP's Adam Tuss contributed to this report.
(Copyright 2010 by WTOP. All Rights Reserved.)
The operator was not supposed to be on the pocket track. When the train tried to get back on the main line, a piece of equipment known as a derailer took the train off the tracks intentionally. "
ARTICLE linked here: http://www.wtop.com/?nid=25&sid=1887376
Our community is FREE to join. To participate you must either login or register for an account.