Metro Preparing to Bring Back Automatic Train Operation for 1st Time Since Deadly 2009 Crash

News4 has learned that a presentation to switch trains back to a computer-controlled automatic mode is on the Metro board meeting agenda for this week.

The plan is to have the Red Line switch to automatic train operation by the summer and the entire system by the end of the year.

However, Metro was designed to operate in automatic mode — and is currently among only a handful of large transit systems that aren’t automated.

Metro says ATO should improve everything from on-time performance to wear and tear on trains and even energy consumption because it will ensure a much smoother ride.

https://www.nbcwashington.com/news/loca ... h/3293031/

I hope it clears everything. ATO is desperately needed as it will eliminate one variable from safety considerations -- the operator.

Having moved here in 2012, I've never known ATO-run trains and I cannot wait. According to WaPo's story, the ATO return will include automatic door open/close. https://www.washingtonpost.com/transpor ... m-transit/. IF Metro manages to pull this off, I think it's a pretty big feather in its cap (to be clear, it's something that is also very, very long overdue).

Just hope that no passenger ever gets caught as happened on MBTA red line.

JDC
Having moved here in 2012, I've never known ATO-run trains and I cannot wait. According to WaPo's story, the ATO return will include automatic door open/close. https://www.washingtonpost.com/transpor ... m-transit/. IF Metro manages to pull this off, I think it's a pretty big feather in its cap (to be clear, it's something that is also very, very long overdue).

The Washington Post article does not state that doors will both opened and closed automatically, only opened automatically. WMATA has never closed doors automatically, though the feature to do so was fitted in the 1, 2 ,3 and 4k cars. not sure if was fitted in the 5 and 6k cars. The 7k cars can not automatically close doors.

I will also note, not having automatic door closing is a fail safe feature in the event of an automatic mode runaway or an incapacitated train operator.

Safety commission monitoring Metro’s move toward self-piloting trains

https://www.washingtonpost.com/transpor ... perations/

“I hope Metro understands how complicated a return to automatic train operation is going to be,” said Robert Lauby, a commissioner with the Washington Metrorail Safety Commission, the agency charged with monitoring rail system safety. “It’s not just a matter of flipping a switch and turning the system back on.”
...
-30-

The big issue in days past was poor sensing/stopping at the correct spot. What, if anything, has changed in that regard in the last decade? I seem to recall the 1000's were the best performing, and later models were worse.

I've speculated how I would resolve the sensing problem. What came to mind is technology far newer than the 1000's. There are off the shelf laser rangefinder/reflectors for industrial applications. If there was an appropriate reflector at the far end of each station's platform, the train would pick it up as it enters the platform, and shed velocity while the distance decreases until the train is at the correct spot.

Yes, you have to add a sensor+ to each A car {Do we still call them that?} and a reflector at each platform; the latter trivial, the former not so, but if it resolved a decades-old issue....

Sand Box John wrote: ↑Wed Mar 08, 2023 10:26 am JDC
Having moved here in 2012, I've never known ATO-run trains and I cannot wait. According to WaPo's story, the ATO return will include automatic door open/close. https://www.washingtonpost.com/transpor ... m-transit/. IF Metro manages to pull this off, I think it's a pretty big feather in its cap (to be clear, it's something that is also very, very long overdue).

The Washington Post article does not state that doors will both opened and closed automatically, only opened automatically. WMATA has never closed doors automatically, though the feature to do so was fitted in the 1, 2 ,3 and 4k cars. not sure if was fitted in the 5 and 6k cars. The 7k cars can not automatically close doors.

I will also note, not having automatic door closing is a fail safe feature in the event of an automatic mode runaway or an incapacitated train operator.

John is correct - I was typing too quickly and added 'closing' without thinking when I wrote 'opening'

farecard

The big issue in days past was poor sensing/stopping at the correct spot. What, if anything, has changed in that regard in the last decade? I seem to recall the 1000's were the best performing, and later models were worse.

I've speculated how I would resolve the sensing problem. What came to mind is technology far newer than the 1000's. There are off the shelf laser rangefinder/reflectors for industrial applications. If there was an appropriate reflector at the far end of each station's platform, the train would pick it up as it enters the platform, and shed velocity while the distance decreases until the train is at the correct spot.

Yes, you have to add a sensor+ to each A car {Do we still call them that?} and a reflector at each platform; the latter trivial, the former not so, but if it resolved a decades-old issue....

π is used to make automatic station stops. Two things need to be functioning properly for automatic station stops to work:

• The 7 marker coils, 1 at 176' from center of platform and 3 pairs at 484', 1,200' 2,700' from center of platform must be function properly.
• The lead pair of cars need to know what its wheel diameter is.

Wheel rotation counts is what is used by the train to know how far it is from the platform stopping point.

5 of the 7 marker coils are passive and tell the train how far it is from the center of the platform. the 5th passive marker coil at 2,700' is the 'grade change annunciator' along with the one at 484', The 2 other marker coils are variable frequency and are used to tell a train less then 8 cars long to short, center or long stop at the platform.

The problem with your laser rangefinder schema is it can not see around the vertical and horizontal curves 75' beyond the ends of the platform and tell the train that it needs to compensate for vertical curvature. Weather would also effect laser range finding.

My approach is solely for the interval starting when the the lead car is approaching the leading end of the platform; that's where precision starts to be important.

farecard
My approach is solely for the interval starting when the the lead car is approaching the leading end of the platform; that's where precision starts to be important.

How is a train going to know when to begin decelerating before reaching the end of the platform? Those 3 marker coils provide a level of redundancy.

I will also note, when all of the parts of the automatic station stop system are properly maintained and celebrated it has an almost nonexistent failure rate.

Sand Box John wrote: ↑Thu Mar 09, 2023 2:14 am
How is a train going to know when to begin decelerating before reaching the end of the platform? Those 3 marker coils provide a level of redundancy.

I wasn't suggesting junking the existing system, rather supplanting it.

The coil scheme may offer redundancy, but they {and wheel counting} did not offer the needed precision for 8-car trains. Some years ago, WMATA paid for an external study; even after that, misplacement remained a problem.

farecard
I wasn't suggesting junking the existing system, rather supplanting it.

The coil scheme may offer redundancy, but they {and wheel counting} did not offer the needed precision for 8-car trains. Some years ago, WMATA paid for an external study; even after that, misplacement remained a problem.

Proving the the route cause was improper maintenance and not doing routine calibration often enough.