Streetcar and light rail signals

[EnviroTO]

It is when the sensor is half way down the block and the light becomes green for the streetcar to the exclusion of all other traffic (which has reds in all directions) and the streetcar signal gets cancelled by the streetcar going through the intersection. Unless a streetcar going the opposite way has just gone through the signal, a streetcar triggering the light gets a signal to proceed exactly at the time it arrives at the intersection. The problem with the way it is currently set up is that the signal is normally there as they arrive which mean they have to slow down because they haven't got the proceed light yet. You can see it best when the streetcars are spaced out. It doesn't work that well with streetcars bunched up (because the streetcar ahead cancels the light before the one behind can get through) or when streetcars in opposite directions reach the light a less than a minute apart (because the first one through cancels the signal giving a green to one direction of cars which even when on an accelerated cycle will not be back in time for the second streetcar).

 

[nfitz]

I can see that it doesn't work well on Spadina (if activated at all) because of all the bunching ... but Queens Quay is seldom that busy ... and on the rare occasion I've run along it, it seems we hit most of the lights on red.

 

[W. K. Lis]

I put up a couple of ideas on the http://commonground.ontarioliberal.ca/ website.

The first is on to Expand to use other international transit traffic signals

and the other is on getting True transit priority signals.

 

[reaperexpress]

I put up a couple of ideas on the http://commonground.ontarioliberal.ca/ website.

The first is on to Expand to use other international transit traffic signals

and the other is on getting True transit priority signals.

The one about using different signal heads is a really good suggestion. Ontario is really restrictive when it comes to signal heads. Not only do we only get one transit signal, we can't have bicycle lenses, and arrows can only be used for green and amber, not for red. This results in us needing those "Transit Signal", "Bicycle Signal" or "Left Turn Signal" signs, which is far less obvious and intuitive than simply having different lenses for different signal types. And all it would take is to change some text in the Ontario Traffic Manual: super easy!

The second would be a lot harder to implement since it is differs case-by-case. There isn't any rule preventing proper signal priority, like there is against different signal heads. I guess we could change some wordings to suggest higher-priority measures such as postponing left-turn phases if they would delay transit vehicles, but I don't think we could do any more than suggest. I think the best way to accomplish that goal is to simply do it. Pick one median ROW and demonstrate that it can be done. My personal pick would be the Highway 7 Rapidway, since it seems to have advanced signal control systems, unlike the TTC's ROWs.

 

[TheTigerMaster]

Do you have any insight on signal priority for the LRT lines, reaperexpress?

 

[reaperexpress]

continued from Flexity Streetcars thread:

Interesting. Do you have any insight on how well signal priority will worth with ECLRT, FWLRT and ELCRT?

I don't know what the plans are as far as the TSP system that will be used. It's possible they would use our current outdated system. It's also possible they would go with an adaptive TSP system, which provides signals with plenty of advance notice of LRV arrivals by operating signals in groups controlled by a central system that is estimating LRVs' arrival time at each intersection. This also allows the signals to gradually adjust their offsets toward that target rather than our current system which adjusts them all at once at the last second. This results in much less traffic impact (including shorter pedestrian wait times) and a higher green rate for transit vehicles.

In any case, the fundamental differences on the suburban LRT lines are obviously the wider streets resulting in much slower signal reaction times, and the longer signal spacing providing earlier advance notice. To minimize the former issue, Metrolinx wanted to put two-stage pedestrian crossings, which vastly reduce minimum pedestrian countdown times. But the City has opposed it on the basis that it's bad for pedestrians as now they could get stuck waiting half a cycle in the middle of the street (i.e. on the LRT platform) when TSP is called.

I personally side with Metrolinx on this one, because many of the pedestrians are accessing the LRT and are only crossing half way anyway, and then will further benefit from faster journey times along the rest of their journey. My impression is that in suburban areas, a large proportion of pedestrians are travelling to and from transit stops.

One big factor that I think needs to be examined is phase rotation, which is re-ordering the phases to benefit transit. Suburban intersections tend to have more phases, so there will be an even bigger benefit on our new LRT lines than on our downtown routes.

For example in our typical median LRT phasing, we could skip the advanced left turn phase if there is an approaching LRV, and serve it at the end of the cycle ("lagging") instead. Phase rotation was highly impractical with our old "MTSS" signal control system, but the "TransSuite" system we now use actively advertises it as a feature. But we still don't use that feature because apparently people would be caught off-guard by the constantly varying signal cycle. I think that this is irrelevant, because all they need to know is provided to them on the signal display. If the light is red, then don't go.

 

[salsa]

One big factor that I think needs to be examined is phase rotation, which is re-ordering the phases to benefit transit. For example in our typical median LRT phasing, we could skip the advanced left turn phase if there is an approaching LRV, and serve it at the end of the cycle ("lagging") instead.

Yes please! I ride the Viva Hwy 7 East everyday to my new job in Markham, and more often than not the buses are held up at almost every intersection so that drivers can make a left turn. Jeez, why should they have to go first? The TSP system they installed there leaves much to be desired.

 

[Johnny Au]

TSP would have been much better with GPS technology.

 

[Rainforest]

Yes please! I ride the Viva Hwy 7 East everyday to my new job in Markham, and more often than not the buses are held up at almost every intersection so that drivers can make a left turn. Jeez, why should they have to go first? The TSP system they installed there leaves much to be desired.

Apparently, some urban experts are concerned that if the left-turn phase gets moved to the end of the cycle, pedestrians who did not finish crossing before the left-turn phase starts are at risk of being hit by turning vehicles.

 

[TheTigerMaster]

If a driver would run over a pedestrian that's right in front of them, then they shouldn't be on the road.

 

[TheTigerMaster]

continued from Flexity Streetcars thread:



I don't know what the plans are as far as the TSP system that will be used. It's possible they would use our current outdated system. It's also possible they would go with an adaptive TSP system, which provides signals with plenty of advance notice of LRV arrivals by operating signals in groups controlled by a central system that is estimating LRVs' arrival time at each intersection. This also allows the signals to gradually adjust their offsets toward that target rather than our current system which adjusts them all at once at the last second. This results in much less traffic impact (including shorter pedestrian wait times) and a higher green rate for transit vehicles.

In any case, the fundamental differences on the suburban LRT lines are obviously the wider streets resulting in much slower signal reaction times, and the longer signal spacing providing earlier advance notice. To minimize the former issue, Metrolinx wanted to put two-stage pedestrian crossings, which vastly reduce minimum pedestrian countdown times. But the City has opposed it on the basis that it's bad for pedestrians as now they could get stuck waiting half a cycle in the middle of the street (i.e. on the LRT platform) when TSP is called.

I personally side with Metrolinx on this one, because many of the pedestrians are accessing the LRT and are only crossing half way anyway, and then will further benefit from faster journey times along the rest of their journey. My impression is that in suburban areas, a large proportion of pedestrians are travelling to and from transit stops.

One big factor that I think needs to be examined is phase rotation, which is re-ordering the phases to benefit transit. Suburban intersections tend to have more phases, so there will be an even bigger benefit on our new LRT lines than on our downtown routes.

For example in our typical median LRT phasing, we could skip the advanced left turn phase if there is an approaching LRV, and serve it at the end of the cycle ("lagging") instead. Phase rotation was highly impractical with our old "MTSS" signal control system, but the "TransSuite" system we now use actively advertises it as a feature. But we still don't use that feature because apparently people would be caught off-guard by the constantly varying signal cycle. I think that this is irrelevant, because all they need to know is provided to them on the signal display. If the light is red, then don't go.

I'm concerned about the opposition to two stage crossing. Slowing down LRVs with hundreds of people on board, so that (often) a single pedestrian can cross is a bad idea.

Does the City's opposition to this mean that the two stage crossing proposal is dead?

 

[reaperexpress]

TSP would have been much better with GPS technology.

GPS-based TSP is indeed one of the options that I expect to do well in evaluation. It's mostly a question of various departments working together with the TTC to get buses' GPS units to be precise enough to work for TSP, which apparently their current units aren't.

Apparently, some urban experts are concerned that if the left-turn phase gets moved to the end of the cycle, pedestrians who did not finish crossing before the left-turn phase starts are at risk of being hit by turning vehicles.

Any pedestrian that is still in the crosswalk after the end of the pedestrian countdown will be always in conflict with some other movement. That's why we have pedestrian countdowns. In the current situation, pedestrians that somehow managed to not finish crossing before the end of the countdown are at risk of being hit by vehicles travelling along the cross street.

Keep in mind that we're talking entirely about fully protected turning movements here.

Yes please! I ride the Viva Hwy 7 East everyday to my new job in Markham, and more often than not the buses are held up at almost every intersection so that drivers can make a left turn. Jeez, why should they have to go first? The TSP system they installed there leaves much to be desired.

I've actually been impressed with how well the Hwy 7 signals are working. The pedestrian crossings are absolutely massive, so it's frankly amazing that they don't have immense delay at each intersection. Given how well it works, I'm guessing they use an adaptive TSP system.

Also keep in mind that York Region uses a TSP system that only prioritizes buses that are behind schedule. If your bus is not late, then you will not be experiencing TSP.

I'm guessing that Toronto will implement at selective TSP system as well, which will be greatly beneficial in reducing bunching as well as reducing pedestrian delays along TSP routes.

I personally am in favour of a graduated TSP system, which has various levels of priority depending on schedule adherence. For example, vehicles that are on-time or early get no priority, vehicles 1 to 3 min late get mild priority (green extensions but no reordering, or vice versa) and vehicles more than 3 min late get maximum priority, with phase rotation, green extensions and red truncations.

I'm concerned about the opposition to two stage crossing. Slowing down LRVs with hundreds of people on board, so that (often) a single pedestrian can cross is a bad idea.

Does the City's opposition to this mean that the two stage crossing proposal is dead?

I don't know what the current situation is, but all of the recent plans I've seen online have had single-stage pedestrian crossings.

 

[CDL.TO]

I personally am in favour of a graduated TSP system, which has various levels of priority depending on schedule adherence. For example, vehicles that are on-time or early get no priority, vehicles 1 to 3 min late get mild priority (green extensions but no reordering, or vice versa) and vehicles more than 3 min late get maximum priority, with phase rotation, green extensions and red truncations.

Thanks for your detailed posts on TSP. There's a lot of misinformation out there.

About your point above, you're dismissing any idea of transit priority serving a purpose of faster transit trips (under normal operating conditions), suggesting that it should only be used to increase reliability instead.

I certainly disagree. TSP should be designed to speed up transit. We have an obligation to do what we can to make transit a superior transportation choice. I do like the idea of the TSP having a more "aggressive" option for when vehicles are behind schedule, though.

 

[Cobra]

I believe in LRT and their capacity to provide rapid transit service, however, I have absolutely no faith in the City of Toronto to make it work which is one of the main reason for the scepticism towards LRT in this city

 

[smallspy]

GPS-based TSP is indeed one of the options that I expect to do well in evaluation. It's mostly a question of various departments working together with the TTC to get buses' GPS units to be precise enough to work for TSP, which apparently their current units aren't.

Commercial, non-military GPS is only accurate to about 7 meters or so, and that number is currently capped by the US administrators of the service. How do you propose that they make it more accurate than that?

Dan
Toronto, Ont.