GO Transit: Service thread (including extensions)

[max]

How would the locomotive know that?

Dan

Pardon the pun, but it's not my wheelhouse so find your salt and administer the appropriate dosage. My thought was that if track circuits are being used, the voltage being used to detect train presence would be measured to detect the locomotive is still on the tracks. Otherwise, the locomotive itself could provide a signal from one wheel set and detect it on another. Automagic detection of which situation a locomotive is operating in would probably be desired and would have to be certain not to trigger or disrupt track circuits as they are safety critical.

Admittedly it would add cost, complexity, and is yet another system that could fail leading to trains not moving when they should which are all possible reasons why such a thing hasn't been developed.

 

[T3G]

Another baffling communications issue from Metrolinx.

A train today was held at Clarkson for 7 minutes. At first, the CSA claimed that we were waiting for a signal (traumatic flashbacks to Monday morning...), but after we left he said that they'd had to fix a minor brake problem.

It was a minor delay, thankfully, but for what possible reason did they change the story?

 

[lastcommodore]

Repairs still ongoing (shot now).

Another baffling communications issue from Metrolinx.

A train today was held at Clarkson for 7 minutes. At first, the CSA claimed that we were waiting for a signal (traumatic flashbacks to Monday morning...), but after we left he said that they'd had to fix a minor brake problem.

It was a minor delay, thankfully, but for what possible reason did they change the story?

I suspect either both (signal & brake) or what seems to be more likely, MX defaults to "signal issues" as the problem until they have more info. Wonder what a "minor brake" issue would be.

 

[WB62]

Another baffling communications issue from Metrolinx.

A train today was held at Clarkson for 7 minutes. At first, the CSA claimed that we were waiting for a signal (traumatic flashbacks to Monday morning...), but after we left he said that they'd had to fix a minor brake problem.

It was a minor delay, thankfully, but for what possible reason did they change the story?

Red signal is the first thing the customer service ambassador assumes when the train stops.

People want a reason why they’re being delayed as soon as it happens, the CSA isn’t at the front of the train, and doesn’t have immediate context to why the train may have stopped.

 

[Metroscapes]

^Begs the question, how diligent was the track inspection regime before the incident and why added diligence is now seen as necessary.

View attachment 713211

I think this photo shows why detailed track inspections are not being undertaken and defects would be hard to detect. The snow is caked over the corridor. Normally Toronto doesn't have so much snow, and we've had a couple undesirable melts due to temperature and/or rain to better expose the tracks. The dark ballast would also help accelerate melting.

It's going to be a lot of work to uncover all those tracks and inspect ties / plates. Especially in the picture, adjacent to the accident scene, where the snow got packed down by dozens of staff trampling around.

 

[smallspy]

I too always thought that there were onboard systems designed to determine the integrity of the consist. Something to tell the operators when a situation outside the norm arises.

There is. The MU, HEP, COMM and brake systems all do that - if the train comes apart, the failures will be immediately evident within the operating cab.

But none of those coming apart were applicable in the derailment on Monday.

Pardon the pun, but it's not my wheelhouse so find your salt and administer the appropriate dosage. My thought was that if track circuits are being used, the voltage being used to detect train presence would be measured to detect the locomotive is still on the tracks. Otherwise, the locomotive itself could provide a signal from one wheel set and detect it on another. Automagic detection of which situation a locomotive is operating in would probably be desired and would have to be certain not to trigger or disrupt track circuits as they are safety critical.

Admittedly it would add cost, complexity, and is yet another system that could fail leading to trains not moving when they should which are all possible reasons why such a thing hasn't been developed.

It sounds like what you're suggesting is a means of the train talking back to the signalling system. Which is in a round-about-way sort of how CBTC signal systems work. (Although they specifically can't detect derailments, either.)

But as you note, that would add a lot of cost and complexity.

And this is just it - this is one of those ideas that sounds like it should be a pretty easy thing to solve, but it really isn't - otherwise it would have been solved already and in use everywhere.

Dan

 

[mburrrrr]

I think this photo shows why detailed track inspections are not being undertaken and defects would be hard to detect. The snow is caked over the corridor. Normally Toronto doesn't have so much snow, and we've had a couple undesirable melts due to temperature and/or rain to better expose the tracks. The dark ballast would also help accelerate melting.

It's going to be a lot of work to uncover all those tracks and inspect ties / plates. Especially in the picture, adjacent to the accident scene, where the snow got packed down by dozens of staff trampling around.

I guess so. They will need lots of brooms. Picture from mid January.

UP conductor cleaning the white bugs off the train today.

GO conductors changing trains last week.

4 trains east of Jarvis where there are 6 tracks.

Can we squeeze a few Altos in there and maybe a soprano too.

 

[stevensontransit]

Metrolinx board meeting documents for next week is out:

Meeting 2026-02-12

Learn more about the Board of Directors' agenda for Thursday, February 12, 2026

www.metrolinx.com

Also in this report (https://assets.metrolinx.com/image/...rate/Item_11.4_-_Marketing_-_FINAL_ENG_Mx.pdf), amid those growing 4 and 5 days a week RTO mandates, events, and one fare program, why does go transit and up express ridership continues to set new heights while the TTC and other 905 transit systems ridership are either struggling or falling?

 

[Dan416]

Metrolinx board meeting documents for next week is out:

Meeting 2026-02-12

Learn more about the Board of Directors' agenda for Thursday, February 12, 2026

www.metrolinx.com

Also in this report (https://assets.metrolinx.com/image/...rate/Item_11.4_-_Marketing_-_FINAL_ENG_Mx.pdf), amid those growing 4 and 5 days a week RTO mandates, events, and one fare program, why does go transit and up express ridership continues to set new heights while the TTC and other 905 transit systems ridership are either struggling or falling?

I feel like the people taking TTC and other local agencies were working jobs that already required being at a physical location (think warehouses in Brampton and Mississauga). They never didn't have to go in.

On the other hand, GO handles commuters who typically work office jobs which were made hybrid during the pandemic. Forcing those workers back in affects GO more than the other transit systems, since they drive to their GO station, and then they often can just walk to their ultimate destination in the downtown core. So them being back in the office doesn't necessarily add to their local transit system usage or TTC's.

 

[Sauga89]

^Begs the question, how diligent was the track inspection regime before the incident and why added diligence is now seen as necessary.

That transparent accident investigation may provide some interesting detail about the rigour (or lack thereof) of track inspections at USRC. The incident happened at a time when the track had been buried in snow for some time and not easily inspected.

A curved track right where the shoving engine is powering up is a pretty obvious and likely spot for wheels to misbehave, especially if the gauge is out of spec or rail fasteners are weakened..

No surprise that the rearmost coach and the leading end of the locomotive were the ones to derail - that's where the buff forces would be greatest. I imagine the investigators will have downloaded the brake and throttle data. The train moved a fair distance before coming to a stop, I would suspect that the engineer (at the other end of the train, in the cab car) didn't even have any indication that something had happened.

- Paul

I think this photo shows why detailed track inspections are not being undertaken and defects would be hard to detect. The snow is caked over the corridor. Normally Toronto doesn't have so much snow, and we've had a couple undesirable melts due to temperature and/or rain to better expose the tracks. The dark ballast would also help accelerate melting.

It's going to be a lot of work to uncover all those tracks and inspect ties / plates. Especially in the picture, adjacent to the accident scene, where the snow got packed down by dozens of staff trampling around.

One other question: What did the data from the hundreds of trains that passed over this track before the derailment show?

I've read a bit about AI being used to spot issues for rail maintenance, and I'm wondering how far Metrolinx has gotten with it. I did find this pilot project from 2021, but I'm struggling to recall anything more recent. It would be interesting to see what you would find if you took the speed and location data for every train in the Union corridor for an entire week, and compared it to previous weeks to look for abnormalities.

I also get the value of looking at a rail with human eyes every week, but could Metrolinx, with Lidar and other sensors on trains, be able to know if a track is out of gauge, even with the snow? I don't mean to suggest it would be easy, but I wonder if it would be possible.

 

[mburrrrr]

This looks like a fun job. Not.
East of Union Station today.

 

[crs1026]

I also get the value of looking at a rail with human eyes every week, but could Metrolinx, with Lidar and other sensors on trains, be able to know if a track is out of gauge, even with the snow? I don't mean to suggest it would be easy, but I wonder if it would be possible.

There are a number of such technologies under development, I can recall seeing writeups of other similar products in the industry press.
There are lots of tried and true technologies already out there to detect defects or deviations in rail and track conditions.
Certainly vendors do claim that their products are superior than traditional inspection methods, but in general the jury may still be out on that point, and the industry is still collecting data.
I would caution that one successful test application does not necessarily imply a product that is ready for mass use..
I don’t know much about what ML has or hasn’t deployed, recently - they adhere to the prevailing regulatory requirements and are no doubt interested in staying current as new methods are put on the market.
When and how often and how the track in the USRC is monitored, I can’t say.

- Paul

 

[nfitz]

... over the next several cycles of maps and signage, the GO train lines will all adopt letters, so this may be confusing in practice.

Again? This was the official map 15 years ago. Personally I thought the current two-letter codes (LE, LW, KI, etc.) are clearer:

 

[lastcommodore]

Personally, I am not in favour of abandoning the GO naming scheme where the ref and name reflect where the train goes... I think it would be very boring to do so too
The only reforms I think I would bring onto GO is to change naming from terminal -> corridor (So that if it ever gets extended it doesn't need renaming) and to put different services as part of the naming scheme, as an example, LW-HA for trains that terminate to Hamilton GO, or LW-NI for Niagara, or LW-EX for express. Granted, LW has too many service patterns which really should be coalesced.

The current scheme prevents any potential naming conflicts with local transit, which is good.

 

[WB62]

Again? This was the official map 15 years ago. Personally I thought the current two-letter codes (LE, LW, KI, etc.) are clearer:

View attachment 714041

These had letters to make the reader know which line was which on a black and white printed version. The letters were never used in times tables, on departure boards, or by word.

My assumption for this was the map was too busy to put the full names, and 2 letter codes weren’t thought up yet.

You can see this present in even older maps that had multiple lines sharing colours:

I’m not going to participate in the debate of transitioning to lettered lines from the inconstant Geographic based (LE LW), End City based, KI MI RH BR ST)/and Destination based (UP) line names, because I’ve already done it a dozen times elsewhere and I can’t go through arguing about this again.

But I will add that no reasonable person today is against the concept of the subway lines being numbered, why be against the concept of regional rail lines being lettered?