2016 Commuter Rail Delay Discussion Thread

GP40MC1118 wrote:1) Make a reverse move at CP3 into the yard.
2) Eastbound on the GJ protecting (at the time) 4 crossings (Mass Ave/Broadway/Binney/Medford St.)
3) Shutting down the Green Line work at Swift.
4) Once at Swift, either going through the layover to Tower A or making a backup move
at Swift to go down the mainline.

All this provided someone signs off on loaded passenger trains in Beacon Park Yard, BET layover and
on the GJ itself.

I'm pretty sure all of this would be remedied if Worcester-BON was an actual service. I don't think anyone thinks it is feasible in its current state.

Now that the Tower 1 problem has been fixed, does anyone know the root cause and resolution details? One article I read suggested a component of some sort had to be replaced and then software needed to be reloaded? Is Tower 1 running on Windows XP? Or MS-DOS?

johnpbarlow wrote:Now that the Tower 1 problem has been fixed, does anyone know the root cause and resolution details? One article I read suggested a component of some sort had to be replaced and then software needed to be reloaded? Is Tower 1 running on Windows XP? Or MS-DOS?

I assume Control-Alt-Delete didn't do the trick.

I'm by no means an expert on this particular system, but looking at the guts of the tower, it felt a little bit old-school to me as an IT professional. I wonder if the logic is really so complex that it couldn't be handled by a few commodity Raspberry Pi's or similar? I know there must be some sort of electromechanical interface for the switches as well, so maybe that's what's consuming so much physical space.

Due to the work at CP4, any move from/to Framingham to the west from Beacon Park has
to make the zigzag move at CP3.

Actually all moves going east at CP3 from/to the Grand Jct have to make backup moves.

D

johnpbarlow wrote:Now that the Tower 1 problem has been fixed, does anyone know the root cause and resolution details? One article I read suggested a component of some sort had to be replaced and then software needed to be reloaded? Is Tower 1 running on Windows XP? Or MS-DOS?

Still working on it....

johnpbarlow wrote:Now that the Tower 1 problem has been fixed, does anyone know the root cause and resolution details? One article I read suggested a component of some sort had to be replaced and then software needed to be reloaded? Is Tower 1 running on Windows XP? Or MS-DOS?

ohalloranchris wrote:

johnpbarlow wrote:Now that the Tower 1 problem has been fixed, does anyone know the root cause and resolution details? One article I read suggested a component of some sort had to be replaced and then software needed to be reloaded? Is Tower 1 running on Windows XP? Or MS-DOS?

I assume Control-Alt-Delete didn't do the trick.

*CLAP-CLAP-CLAP-CLAP....*

Diverging Route wrote:http://www.bostonglobe.com/metro/2016/02/19/computer-issues-blamed-for-commuter-rail-problems/63TqSZ8FwgrzBT8CpvAPBJ/story.html

Well, the article implies that there is a trailer-sized bungalow somewhere outside south station, and that's where the failure occured. dweebly linked to the tweet showing the picture of the control board with PLC failure red light on. This article basically tells us Amtrak had no clue what was wrong, so they reloaded software and swapped components until it started working again. I find it highly unlikely that they don't know what component failed at this point. This Programmable Logic Controller was probably installed when the station was overhauled around 1987, or in the 90's when additional tracks were added. So Time to Failure is 29 years. Bet the state-of-the-art 1898 electro-pneumatic system never had a major outage over its 90 years. The article repeats the drivel about operating the switches manually in the 1940s - maybe in other less busy stations, but not at South Station!

Would be interesting to read up on the brand and model of PLC and desktop software they are using.

Shuttle bus replacing Haverhill Inbound Train 2202 (10:20 am from Haverhill) between Haverhill & Bradford due to Amtrak interference. Train will depart Bradford at 10:42am.

Due to Amtrak 690 whacking a car in Andover this morning

D

rethcir wrote:I'm by no means an expert on this particular system, but looking at the guts of the tower, it felt a little bit old-school to me as an IT professional. I wonder if the logic is really so complex that it couldn't be handled by a few commodity Raspberry Pi's or similar? I know there must be some sort of electromechanical interface for the switches as well, so maybe that's what's consuming so much physical space.

Considering the general skepticism of railroads toward new technology (with good reason) and the 1985 +/- 3 years installation, it was probably state of the art for sometime in the 60s. You wouldn't need that many ARM microcontrollers to drive things and provide a high level of monitoring. Of course, there would probably be a year or two of pain not unlike Thursday while bugs got worked out (although better choice of implementation languages (e.g. Standard ML) would make a difference).

GP40MC1118 wrote:Due to Amtrak 690 whacking a car in Andover this morning

D

Then why are they busing from Bradford to Haverhill?

A few points about Railroad Signalling and Communications.
First and foremost the system must be 100% fail safe. If anything goes wrong it must stop trains and not allow an unsafe move. This puts a burden on any equipment and software that is about the same as nuclear weapons.

The physical interlocking equipment at Tower One is relay based. Not like any relays you have seen anywhere else. The redundant PLC's are connected to the Dispatch center (CTEC) over redundant private data links. The PLC's talk to the relays and the relays are the "Vital Logic" that makes the final decision as to the safety of the requested route.

Second, consider the environment that this equipment must operate in. The same rails that are used to detect a train and transmit cab signalling must carry the return current from the 25 KV electrification without interference.

Forget Mother Nature at your own risk. I've seen lightning strikes on the rails so severe that equipment was literally blown out of the racks inside the CIH (Combined Instrument House). When that happens no allowance can be made for an unsafe route or indication (False Clear).

Being "State of the Art" is not the goal here. Not killing anyone is the goal.

Type 7 3684 wrote:

GP40MC1118 wrote:Due to Amtrak 690 whacking a car in Andover this morning

D

Then why are they busing from Bradford to Haverhill?

The email alert said Ballardvale to Haverhill.

Also, 692 was making local stops at Lawrence and Andover, apparently.