Network News

X My Profile
View More Activity
Share Stories  |  Traffic  |  Columns  |  Q&A     |  Get Gridlock:    Twitter |    Facebook  |     RSS   |  phone Alerts

Metro's Continuing Problem With Train Controls

Should Metro, and some other transit systems, rethink the use of automatic train controls? That was an issue raised by Rep. Brian Bilbray (R-Calif.) during a congressional hearing Tuesday on the June 22 Metrorail crash.

But whether trains are run automatically or manually, they still need a protection system that prevents them from getting too close to each other. Right now, Metro doesn't have a system we can trust completely.

On the Red Line, the track circuit that flickered on June 22 and hid the location of the train that was struck continues to malfunction, even after critical equipment was replaced. Metro and the NTSB don't know why. Metro Board Chairman Jim Graham said during the congressional hearing that this mystery must be solved before normal operations can resume.

Until Metro is sure that all trains can be detected by its electronic protection system, it will send only one train at a time through the crash zone between Takoma and Fort Totten. Next time you're riding, look at a Metro map to see how long a stretch of track that is. It's one of the longest gaps between stations in the entire system. Trains would normally be able to reach 59 mph between them.

Sending only one train through at a time backs up the entire line. But if Metro operated as usual, it would risk duplicating the conditions that led to the June 22 crash.

All Lines Affected
Metro has tested all of its track circuits and found that only the mystery one between Takoma and Fort Totten is malfunctioning. But that doesn't mean train control issues are confined to the Red Line, or that they have been addressed by switching the trains from automatic to manual control.

As the National Transportation Safety Board has pointed out, the train protection system is the same, whether the trains are operating in automatic or manual. The system has the ability to maintain a safe distance between the trains -- stopping one train, if necessary -- even with a human at the train controls.

That's what made the NTSB's safety recommendation on Monday so urgent. If another track circuit were to malfunction, and a train became invisible to the electronic defenses, there's no more guarantee today than there was a month ago that the humans in the control room would detect that. And there is no real-time electronic backup. That's what the NTSB wants Metro -- and any other transit system that lacks one -- to acquire as quickly as possible.

But there is no quick fix, and no amount of instant money can create one. Only a few companies are capable of creating this fix, and Metro is just starting its talks with them.

By Robert Thomson  |  July 15, 2009; 9:05 AM ET
Categories:  Metro , Safety  | Tags: Dr. Gridlock, Metrorail, Red Line crash  
Save & Share:  Send E-mail   Facebook   Twitter   Digg   Yahoo Buzz   StumbleUpon   Technorati   Google Buzz   Previous: Congressional Committee Examines Metro Crash
Next: Traffic Alert: Backup on Northbound I-395


Of course, with its perfectly staight lines and 45 degree angles, the Metro diagram is more of a work of art than a map to scale. Take a look at the bus maps for a better idea:

The distance between Takoma and Ft Totten is about the same as the distance between all of the other station pairs north of Brookland.

Posted by: Chris737 | July 15, 2009 10:27 AM | Report abuse

I'm not sure operating the trains in manual mode would have made a difference in this accident. Apparently the trailing train had a clear track signal (green light) and was going at the approved speed (59 MPH). The stopped train was just around a curve and not visible until it was too late for the approaching train to stop.

Perhaps the answer is to slow down appropriate speeds in proximity to blind curves unless and until a fail safe train detection system can be developed and installed.

Posted by: imareader | July 15, 2009 11:02 AM | Report abuse

I would agree. First of all, why is there a fence between the Metro line and the CSX track? Wouldn't it make more logical sense to have the fence on the outside of the CSX track? That would improve sight distance.

Second, the system needs to be made fail-safe. Not fail-proof, but fail-safe. There is a difference. If a railroad crossing gate breaks, the battery backup kicks in and the gates go down. In otherwords, the presence of a train does not cause the gates to go down, the lack of a train causes the gates to hold up and the default is for them to be down. So if they break...they have failed, and the system is not fail-proof. But, while it is failing, the gates are down so hopefully people aren't on the tracks if a train comes. That is fail-safe.

What I don't understand is why the a broken signal doesn't automatically stop all trains. Why doesn't a "fluttering" signal stop all trains? Why is the indicator light for a fluttering signal so small that it is barely noticible to control personnel? I believe Automatic train controls are great, but there needs to be a fail-safe, redundant, backup system in place to oversee the ATC.

How many times do you see car crashes because the main road and the side street both have a green light? Almost never. That is because in every signal controller cabinet, there is a completely seperate piece of equipment called a "conflict monitor". The conflict monitor "watches" the signal controller. If the signal controller were to malfunction and attempt to display a green indication to a main road and cross street at the same time, the conflict monitor over-rides the controller and forces the signals to flash in red on all streets. The flashing operation will get the attention of people, so the authorities know the signal failed. But having the lights flash red is fail-safe, since it never did display a green to cross traffic at the same time as the main street.

Metro needs a completely seperate system as well, using technology other than track circuits. This could be as simple as outfitting every Metro car with an RFID tag or transponder, and putting sensors at the ends of each block. It could even be done by laser beam, where the train breaks the beam. Anything to detect the presence of a train. If one system disagrees with another....the track circuit says there is no train but the RFID sensors never reported the train leaving the block....then all trains must stop and operate in manual at a slow enough speed to ensure that the operator can stop if there is an obstruction. Engineers would be needed to determine what the appropriate manual speed should be on any particular track segment.

Posted by: thetan | July 15, 2009 6:34 PM | Report abuse

Great post, thetan -- I don't quite understand why the "engineers" at Metro never grasped this until it was too late.

Your solution to use a laser or RFID tag is a little more complicated than it needs to be, though -- just use a simple electrical or mechanical sensor of the type that railroads have been using for almost 100 years. Problem solved...

Posted by: stuckman | July 15, 2009 6:47 PM | Report abuse

The comments to this entry are closed.

RSS Feed
Subscribe to The Post

© 2010 The Washington Post Company