Watch Streetcars Live on Next Bus Beta

Next Bus has the TTC streetcar system up in a beta version on their site.

That link will take you to a display of the Queen car, although as I write this, no service appears to be running west of Humber Loop.  You can select other routes from a menu provided on the page.  Right now there is no Carlton service west of Bathurst or east of Broadview, Dundas cars are all east of Ossington.  Data for St. Clair is reported only for the streetcar portion east of Bathurst.

This is a trial.  Don’t complain to me if it doesn’t work.  That’s what trials are for.  Apparently there are still cars with no GPS units, and this causes problems.  Of course it is possible that the service really is that screwed up, but one quick cross-check is to compare the number of cars shown on the display with the number of runs that should be in service.  For example, the Bathurst route shows only two cars even though four are supposed to be in service at this hour.

The site requires that you have Java installed to use the live map feature.

33 thoughts on “Watch Streetcars Live on Next Bus Beta

  1. It’s 2009 and they wrote this in Java.

    Steve: Blame NextBus, not the TTC. Also, they have been providing this service to other systems for a while, and the code may have been built in an earlier era.


  2. Andrew: you think the service is bad according to that map? pfft, go to Spadina station and look at the 510 Spadina monitors. the right monitor has the location of all 510s. many times there are 5 bunched up at Spadina loop, a few streets down you have 4 more bunched up, and so forth. At least they have that directional arrow so you know North or South bound.

    Steve: just out of curiosity…where is the GPS unit? somewhere under where the passengers are? maybe the roof?

    Signal from streetcar goes to a satelite, from there it goes to an antenna, which then goes to some server on the ttc system, which then goes to places like I guess “transit control”/the site you linked above/the 510 monitors.

    What is the delay from streetcar until it gets to some kind of display?

    Steve: I don’t know how much of a lag there is in the TTC’s handoff of data to NextBus.


  3. It’s not, don’t look at any of the lines other then Spadina, 509 and maybe Queen – the rest clearly do not show all the streetcars that are in operation. 510 & 509 should be fine as those lines must (should) have GPS already installed on them

    Service on Queen and Spadina look good to me at this exact moment – granted it’s 1AM 🙂


  4. Let’s not get too worked up yet people.

    The TTC doesn’t even have real-time GPS tracking yet. This is based on schedule timed stops NOT GPS.

    In the future however, it likely will.

    GPS (for now) is only used for on-bus stop announcements. Position reporting is still done with Dead Reckoning via CIS. The TTC doesn’t have sufficient data bandwidth on their CIS network to be reporting with GPS. Once everything is moved from UHF LMR to CDMA on Bell, it wouldn’t be a problem.


  5. You may want to look at some demos of TTC NextBus applications at this U of T course Web page (click on Demo links).

    The applications illustrate how NextBus XML data can be deployed for a variety of purposes.

    Steve: Note that the links from this page may not work if you are behind a firewall that blocks most TCP/IP ports. Also, there is a certificate error on the website.

    I have found that these apps are not well-behaved.


  6. If anyone wants to see how NextBus looks on a full installation, call up the San Francisco Municipal Railway and check that one out. If watching the N JUDAH line bear in mind it runs two-car trains on most assignments. For the F line I’m not certain if all of the historic cars sometimes used have GPS.

    Steve: Here is a link to the F Market line. You can select other lines from the Lines menu.


  7. “Let’s not get too worked up yet people.

    The TTC doesn’t even have real-time GPS tracking yet. This is based on schedule timed stops NOT GPS.”

    Are you sure about that Mike? I’m looking out my office window, down Carlton from Yonge to Parliament. The positions on the map match the streetcars’ actual positions with a lag of 30-45 seconds. The map seems to accurately show gaps too, which certainly aren’t according to schedule.


  8. This is just a test system based on existing data its not “Live” or “real time” just yet. This is a sample created for the commission however the data it is using is based on timetables not actual vehicle data. For example I watched an intersection on Queen Street during the AM rush, it was saying the next car was 20 minutes away … 4 cars came and went. And the System seemed to thing there was a grand total of 11 Streetcars (both ways) on the entire route (Neville to Longbranch) at 8am …

    is a link to a press release regarding a contract they have with the TTC, however if you go to the TTC’s website and do a search on “nextbus” no pages are found, and the only info about next vehcile that has been released by the ttc has to do with their PILOT PROJECT on the 510 SPADINA route (with the screens and the maps at Spadina and Union Stations).

    Steve: Yes, everyone here should know it’s a trial. The point is that info about this has been circulating for a few days in the fan community and I thought it worthwhile for a larger audience to watch the progress. It’s also good to look at other busy systems like MUNI to see how a fully rolled-out display behaves.


  9. Hmm, suddenly only 509 and 510 data is available. Perhaps this beta wasn’t meant for public consumption?

    Steve: Oh well, it was nice while it lasted.


  10. According to Spacing, “Next Bus, the company providing the GPS tracker, is going to disable the arrival time feeds for all routes except for the Spadina 510 and the Harbourfront 509 streetcars. The TTC said all other routes are not GPS-to-web enabled. Next Bus is only providing estimated times on the other routes since the info is being pulled from the schedule and not the location of the nearest streetcar. The TTC said Next Bus does not want to provide information to the public that is not yet reliable.’ I actually tested it on the 504 this morning and it seemed quite accurate for the 3 cars I saw but I guess it does not make sense to test it on routes where not all cars are GPS equipped.

    Steve: All I can say is that if they based their displays on the schedules, then the TTC has some mighty strange schedules. I suspect that the real problem is for NextBus to actually get meaningful, reliable data. For the TTC, the big issue is that the claim I have heard at some public meetings about a full conversion to GPS for all vehicles is somewhat oversold. It runs the stop announcements, but not the vehicle tracking.

    Also, if route 501 isn’t working yet, how can route supervisors be using this sort of display to manage the route?


  11. As usual, the Spadina streetcar is really bunched up. Note to riders: PLEASE wait for the last streetcar if there are two (or more) streetcars bunched up like that. It always surprises me that riders aren’t willing to wait one minute longer to ride in a much less crowded streetcar.


  12. People will not wait one minute for anything: mememememememe.

    Streetcar drivers should be instructed to leave people behind if other streetcars are right behind. Close the doors, clap!

    When the streetcar at front overloads:
    – It takes longer to load each passenger
    – It holds up traffic (although not on Spadina)
    – The ride is less comfortable
    – The front streetcar lengthens its delay
    – It makes bunching worse by allowing subsequent cars to catch up – and speed up because there is no one else to pick up.

    Why. no. one. at. the. TTC. does. anything. about. it. is. beyond. me.

    Steve: People get on the first car for one very good reason. If there is a pack of cars, some of them will be short-turned. If the first car suffers this fate, you can always drop back to the second or third one. If you wait for the third one, and it’s short turned, then you’re back standing in another gap waiting for service.


  13. I’ve been watching the estimated arrival times for the Queen’s Quay segment of the 510. When waiting at the Union Station terminus for the 510, I noticed that the overhead display consistently underestimated the time until the next arrival. By watching this map, I think I know why. The system does not distinguish between cars that short turn at Queen’s Quay and cars that continue on to Union Station. As a consequence, the system always reports that an estimated arrival time for the 510 on only a few minutes along Queen’s Quay, even though most of the cars end up short turning, and wait times are often up to 15 or 20 minutes.

    Bad information is worse than no information at all. They need to fix this. Imagine waiting at Harbourfront in the dead of winter at night, being repeatedly told that the next car is only 1 or 2 minutes away, when it fact it will be 20.


  14. It appears that this site has been de-activated. I can no longer get it and says that the agency does not exist. An article in this morning’s Star said that it was only an internal demo and was not supposed to be live. However the F line on San Francisco MUNI is running well on a 7.5 minute headway.


  15. Before it was de-activated, the Next Bus system showed the 509 Harbourfront line missing one car (as of 3:30 pm Wednesday, March 18th). Hence, there was about a 15 minute gap between cars (there is close bunching of the other four 509 cars; there are supposed to be five cars during the PM peak). Hopefully this is not the case and it’s simply that one missing car not having the GPS installed.


  16. The Stop announcement on buses/streetcars as far as I understand work on GPS…Every bus/streetcar that I have been on since end of summer 2008 has those stop announcement things. Something on each streetcar/bus is sending a signal to a satellite then the satellite sends the signal back to earth (as far as I understand how GPS works).

    Why can’t this be used to track buses/streetcars? I am sure there will be some seconds of delay for the signal to be sent from the vehicles to the server and eventually the website/my cell phone.

    Steve: The GPS info has to be integrated with the antique CIS system, and this has not been done for all cars. Also, any peculiarities in mapping of GPS info into the old CIS framework needs to be worked out. It’s not as straightforward as it seems.


  17. The Next Bus system is up and running again. Probably a glitch.

    Steve, I don’t think this system can’t tell if southbound 510 cars are truly heading to Union Station or just simply short-turning at Queens Quay/Spadina. Or will the system be able to track extra 509 cars that run during special events at the EX (eg. Toronto FC soccer games or the CNE)? It will be interesting to see how will Next Bus and the TTC will solve these and other issues. But as of today, I am very impressed with the system. It was very accurate last night and this morning with respect to the 509 data.

    Steve: As things now stand, the CIS system depends on the schedule to “know” where a car is going, and it gets really confused when a car short turns or wanders off route. GPS data allows this problem to be detected sooner, but until the TTC implements a scheme where the car’s destination is entered on the fly based on where it is actually going rather than what the schedule says, there will be problems.

    On Queen, life is much more complex because of the step forward and step back crewing schemes which cause cars to frequently change run numbers. I have found in my analysis of CIS data now in progress for Queen that run numbers are all but meaningless (and hence any link to a “schedule”), and vehicle numbers give far more reliable output. If the TTC is going to move away from hard schedules to “run as directed” service, this will defeat the entire basis of the NextBus “prediction” scheme.


  18. WWhen this thing is working, every driver should have a display. And a cell phone or walkie-talkie to communicate with the other drivers on a route. And make it the driver’s responsibility to avoid bunching.

    Fantasy scenario

    trailing streetcar – “How full are you?”
    Leading streetcar – “Packed. And 20 more on the platform.”
    Trailing streetcar – “Tell them to wait. I’ll be there in one minute, and I’m going right through to Union.”


  19. Just looked at the live map again (12:20 Friday) at saw 10 streetcars packed together southbound from around Queen to King. The blockage was just clearing as I watched. The labels for each car block the actual arrow image when they get close together or pass. I hope they are at least going to shorten the labels to the route number only. It would also be really helpful if each route had a unique colour so you can tell them apart immediately on shared trackage.


  20. I have been watching the 509 and 510 cars in the background while reading other material. It seems that the Spadina cars are herbivores because they seem to know that there is safety in numbers. It would appear that they like to operate in 4 or 5 car packs. This morning just after 9:00 there were 3 Spadina cars, one on Wellington at York, one on King at Parliament and one in Queen’s Quay Loop that said they were on a break. I assume that these were cars coming off another line to enter service on Spadina. It is interesting to watch how the gaps and bunches seem to continue all day with out ever getting fixed. From the display it seems that a curve at Spadina Circle to turn cars back south would help immensely.

    Steve: This switch was deliberately not included in the plans to avoid disturbing the ever so delicate equipment in the earth sciences building then under construction. How they manage to avoid upsets from other passing traffic I don’t know.

    I look forward to seeing route supervisors with handheld displays so that they can see where the bunching is. Of course, they already have this at Spadina Station, but it’s not practical to keep a big queue of cars there to space them out.


  21. Kristian Says:
    March 20th, 2009 at 12:27 pm

    “Just looked at the live map again (12:20 Friday) at saw 10 streetcars packed together southbound from around Queen to King. The blockage was just clearing as I watched. The labels for each car block the actual arrow image when they get close together or pass. I hope they are at least going to shorten the labels to the route number only. It would also be really helpful if each route had a unique colour so you can tell them apart immediately on shared trackage.”

    I have spent some time watching the San Francisco Muni and they have different colours for each line.

    Steve says: “This switch was deliberately not included in the plans to avoid disturbing the ever so delicate equipment in the earth sciences building then under construction. How they manage to avoid upsets from other passing traffic I don’t know”

    I remember doing a lab in the “Mining Engineering Building”, and that dates me, where we had a grown crystal that had a known mass hung on it. The object of the lab was to determine the tensile strength of the metal by determining a delta L (stretch) versus time graph for the experiment. The problem was that every minute there was a jump in the graph where a Carlton Car went across the special work at McCaul. When you eliminated the jump from the shaking of the street car you got a graph that usable, the only trouble was the Lab Dummy didn’t know about street cars.


  22. I observed something rather interesting on the King Car today. Just after a Route Supervisor walked out into the street to have the -18 car I was on turn at Roncesvalles, another operator got on at the next stop between Spadina and Bathurst. They explained they were part of a new Division Committee set up to help identify operational safety issues and would be attempting to observe and converse with every streetcar operator each month. They had a clipboard with a note-taking sheet headed with the words “No shame, no blame” and footed by “Don’t dread, We shread”. The main part of the exercise was supposed to be ensuring that individual operators didn’t have any dangerous habits, but the observer really just ended up taking down complaints and comments from the operator so that the committee could find ways to address them.
    I listened in to the conversation for the entire trip, although I didn’t have to try hard because they were talking really loudly. Pretty much every issue identified went back to problems with dangerously-badly maintained vehicles and lax CIS supervision. Detailing the entire story would take too long, but some highlights were as follows:

    – Carhouse staff knowingly fielding operationally unfit vehicles in hopes that the operators wouldn’t notice and reject the cars.
    – The speaking operator having had to drive a disabled car all the way back from St. Clair to Roncesvalles with only track brakes barely working and the line-truck driving behind instead of in front. They were told not to couple to another car, even though the carhouse attendant refused to take over at Roncesvalles until another car was coupled.
    – Being told it was okay to proceed with one working sander in bad winter weather, even though the length of the sand trail is one of the first things measured in an accident investigation.
    – Roncesvalles CIS Supervisors spending much of their time on smoke breaks, and with their feet up on their desk when at the terminal. They apparently spend little of their time actually managing the lines which leads to a number of problems which the operators end up in trouble for. It also results in dangerous driving often necessary to get cars back on schedule and cars following a long gap running out of transfers. Schedule adherance is still foremost in the line management theory no matter what the disastrous outcome to the public.
    – Numerous accidents have occured recently due to operator inexperience. The five-day Division familiarisation program was cited as being wholely inadequate.
    – Operator stress is running extremely high because they are the only ones held accountable for the problems resulting from any of the above.

    As we’ve already learned, service unrealiability has little to do with the oft-spouted “Traffic congestion”. What’s amazing is just how bad things have been allowed to get to the point of serious routine danger to all occupants of the road. A lot of this goes back to the politics of budgeting, but the operation is flawed at every level and the operators are feeling like they’re bearing the weight of it. This situation is an absolute disgrace!


  23. wow. I am still disappointed by Toronto’s latest public transit innovation. Is there really no way to plan a route online on the ttc’s website. Being a newcomer to toronto I was really surprised and disappointed to find that there is no online tool to help plan a route. Take for example Montreal’s online public transit website, all you do is enter a starting address and destination and you are given several routes to take by the system step-by-step, no need to look anything up, along with approximate times taken to go from point A to point B. Takes all of 2 minutes and I know exactly where I’m going, how to get there and how long it’ll take. and Montreal’s system has been in place for over a decade. Doesn’t toronto have anything like this??? I rather use a route planner like montreal’s site over nextbus any day.


  24. How long is it going to take for this to be both accurate and released to the public? That’s the real question.

    If anyone has looked at the next Streetcar arrival signs at Union Station or Spadina Station, you will note that they’re not only inaccurate with real life but don’t evem match up with the GPS maps at Spadina. That’s a problem.


  25. Wow, it’s not dealing with the Spadina to Union closure very well:

    It keeps loosing the remaining 509 cars, and can’t figure out they they never go east of Spadina.

    This kind of system is most useful when there’s some kind of detour going on; and one that is scheduled, such as this one, should be reflected in the data.

    Steve: This is an inherent problem of the TTC’s information system that predicts and tracks vehicles based on the schedules. I was on a Spadina car today going to Queen’s Quay, but the system thought it was a King short turn, and called “Oxley Street” as the next stop as we crossed Adelaide southbound.


  26. So it’s now fall 2009 and the Nextbus system is still in Beta, any idea as to how long this beta test will be on for?

    Steve: No idea at all. The online Trip Planner still has not made it out the door either.


  27. “The online Trip Planner still has not made it out the door either.”

    It’s been four years at least (and probably two since Enough is enough.

    I know the alternative is not zero cost in terms of data prep but god knows how much has been sunk into the notion of a TTC super-swishy-roll-our-own-planner at this point.


  28. Any idea how long the 9 million dollar contract with nextbus is for? so how much per year is the ttc paying?

    Steve: I wil have to look this up. Stay tuned.


Comments are closed.