The Manhattan Project
Controversy Over Rhee Now to Include Sex, Congress
Report: CCDC Split From UDC Is "Critical To Success"
Increased Urban Walkability For The Win
D.C. United Icon Ben Olsen To Retire
The National Transportation Safety Board today issued what it's calling an "urgent safety recommendation" to the Washington Metropolitan Area Transit Authority, based on the preliminary findings of its investigation into the fatal June 22 Red Line crash. Despite the fact that the investigation has not yet concluded and the NTSB has yet to say officially what they believe was the cause of the accident, they have gone ahead and told WMATA that the transit agency needs to adopt "enhanced safety redundancy" of its train control system. Here's what they said:
A recent accident on the Metrorail's Red Line between the Fort Totten and Takoma stations has shown that WMATA's train control system is susceptible to a single point failure because it did not fail safe and stop a train when detection of a preceding train was lost. The urgent safety recommendation issued today calls for WMATA to evaluate track occupancy data on a real-time basis in order to detect losses in track occupancy and automatically generate alerts to prompt such actions as immediately stopping train movements or implementing appropriate speed restrictions to prevent collisions.
The Safety Board made a second urgent recommendation to the Federal Transit Administration (FTA) urging the agency to advise all rail transit operators with train control systems capable of monitoring train movements to evaluate their systems for adequate safety redundancy.
WMATA has since responded with its own official statement, in which the transit agency argues that "that there are currently no systems available commercially that could provide the Metro system with the kind of alerts that the NTSB has recommended." In other words, someone needs to build such a backup system, specifically designed for Metro, and then it'll need to be paid for and installed. That's a big undertaking, one that could easily take over a year. In the meantime, WMATA says, Metrorail riders should get used to trains running in manual mode.
We will continue to operate in manual mode until a suitable backup, designed specifically for our rail system is developed.
No one's even able to speculate on a timeline for such a project yet. All we can say right now is that for the foreseeable, and likely lengthy future, Metro trains will be operated manually and will be stopping at the end of the platforms.
add to stumble upon
add to reddit
tweet this
post to facebook
in otherwords, prepare yourself for some whiplash.
and learn to love the smell of scorched brakes...
One day I'll be telling my children, "Back in my day, Metro trains were operated automatically ... yeah, with computers and such."
Can someone explain the stopping at the end of the platform to me?
Because there have been several incidents where train operators forget they're running an 8-car train, stop at the 6-car mark, and open the doors, while the last car or two are still in the tunnel.
Thank you!
You know, driver wakes up, or looks up from texting, and the train isn't at the right place on the platform.
R.I.P. Metro! You've got only yourself to blame.
Failure of equipment manufactured by a private company is Metro's fault? Please explain.
If the transmission in your Chevy dies, do you blame Chevy or the company that made the transmission?
That depends. Did someone inspect my car a year before and warn me about driving it and I went ahead and drove it anyway to save some money on my taxes?
There's an app for that. It's called GPS and a cellular connection. Monitoring metro trains on a track is significantly easier that a fleet of trucks throughout a city. I'm sure there are off the shelf products that could be used tomorrow. Possibly free off the shelf products. This handles the above ground tracks.
For below ground tracks, it'd probably take a little more ingenuity, but a similar system could be implemented fairly easily. Shoot, pretty soon a lot of underground stations will have cellular service, so you could just use those as markers (like google maps does on a phone without GPS). Again, these trains are on tracks, so this isn't something that is difficult to put together.
But it's Metro. Even the basic things are impossibly difficult for them.
You really want to trust your safety to the reliability of an underground cellular network?
Actually, GPS is not up to the task. This is a far more complex system than the ones used to track buses and delivery vehicles, which themselves are expensive proprietary systems. A safety fail-safe system for Metro would have to be far more accurate and faster than any existing GPS application. Think about how often the GPS system in your car loses the signal. The time that it would take to re-acquire a signal would be more than sufficient to allow a collision. The alternative would be to stop all trains as soon as any one train loses a GPS fix. This would make the system far less reliable. In addition, GPS will not work in tunnels or under bridges.
No existing cell phone technology can fill the gap, either. Even if there were, it would have the same failings as GPS: unreliability and occasional signal loss. If the Nationals won a game, and everyone called a friend to tell them the good news, do you want the Metro system to shut done because the cell system is overloaded?
i spoke to a Metro driver and he told me, "now i get up around whenever, i used to get up on time. but that old man, he's a real motherf#*%er, gonna kick him on down the line"
Absolutely ridiculous. Just exactly what's going to happen when an operator causes an accident on manual? Is the NTSB going to require that all trains be pulled by laser-guided donkeys until a suitable backup backup can be found?
If only we had the NTSB dictating highway safety... we could all be driving cars insulated by a four-foot-thick cushion of bubble wrap.
And we're going to get to hear those platform announcements about too! Does knowing that the trains are stopping at the end of the platform (which seems like an obvious, sensible idea) matter to anyone on the platform?
I've seen people get PISSED when they are strategically positioned where the doors would "normally" be located, only to have to move their butts down the platform when the train blows by them on its way to the end of the platform. It's quite entertaining.
Sheesh Metro, just stop at the end of the platform from now on! I get nervous when people start shoving toward the train when it doesn't stop exactly in front of them. I don't want to get pushed onto the tracks and die!
It's not just an annoyance. If you are at the back of the platform and the train blows by, you are basically not getting on that train if it's rush hour since everyone is then clustered around the same car. Since this changed, I have to walk halfway down the platform for no reason other than to ensure I actually make it on the train. A big deal? No. But at some point I had to realize what was happening and adjust.
Single point failures can be eliminated by doubling the sensors, or even tripling. At each zone entry and exit point, two or more sensors would monitor passing trains. At any point, if the two sensors do not send the same signal, then an error is logged and lights can turn red. To truly be effective, the redundant sensors must communicate through separate wires to separate computers then to separate lights/signaling. Only then can single point faults be eliminated throughout the system. This obviously is prohibitively expensive; double the cost for two of everything. All of this technology exists and is in use by Metro today.
I still fail to understand why massive equipment upgrades are necessary to have a system that is smart enough to identify failures.
The sensors send data to a computer somewhere. It should not be rocket science to have the software on that computer let someone know when a sensor stops sending data.
Every automobile made for the last 20 years lets you know when a sensor stops working and turns on the "check engine" light. It's mind boggling that the software that manages Metro trains in automatic mode wasn't designed to perform such a basic diagnostic from the beginning, but it shouldn't require much to upgrade the software to do this now.
You don't need redundant sensors. You just need to let someone know when one stops transmitting data, or transmits data that is not expected.