Thanks to all of you for the informative discussions. I'm new to this thread (I had to look up "LPI").
Safe Systems Approach to Urban Planning (thank's
@reaperexpress) reminded me of Safety Management Systems (SMS) from my aviation background . A cornerstone of SMS is the non punitive reporting of all near miss incidents. Accordingly, and thanks to a prompt from
@Northern Light, I did send an email with the dash cam video to management at Toronto Transportation Services. I haven't heard back, not that I expect to, but I'll post on this thread if I do.
I went back to that intersection today with the drone to see if I could capture the traffic signal sequence.
I wasn't sure the signals would be visible from the drone so I also shot a clip with my phone from where the pedestrian was waiting to cross.
The drone is facing west.
Thanks for the additional context. It seems the issue was indeed that the westbound LPI was not combined with a No Right Turn on Red restriction. It's possible the LPI even made the situation worse by making it less obvious to the truck driver (who would have been looking left waiting for a gap) that the east-west phases had begun.
Some thoughts about this specific intersection:
- There seem to be quite a lot of articulated trucks making the southbound left and westbound right turns.
- There is already a permissive-protected green arrow for the westbound right turn so adding a No Right Turn on Red [ Except on Green Arrow ] restriction wouldn't be that much of an impact on drivers. If necessary, they could install a setback detector in the westbound right turn lane to make sure the green arrow stays on long enough to clear the queue.
It doesn't seem like there are many location-specific factors other than the high truck turning volumes so any mitigation measures should be applied broadly based on the general risk factors, rather than just a knee-jerk reaction to a particular incident at a particular signal. We would also need to consider whether the mitigation measures could have unintended consequences that increase other types of collisions.
Assuming that No RTOR is the correct mitigation measure, I could see a few strategies with varying aggressiveness:
- Introduce No RTOR at intersections with a relatively large number of right-turning trucks, or
- Introduce No RTOR at all intersections with LPI, or
- Introduce No RTOR City-wide, unless there is a sign specifically permitting RTOR.
The main potential risk I could see related to adding a No RTOR restriction is a reduction in signal credibility. Basically if people feel like they're waiting for no reason, they feel like their time is not being respected by the traffic signals. As a result, they're less likely to respect the traffic signals, which has obvious safety consequences. We've already seen this a lot over the past decade where the proliferation of signals that count down way earlier than necessary (typically due to LPI in the subsequent phase) has resulted in widespread disegard for Flashing Don't Walk and even Don't Walk indications. I personally ignore Flashing Don't Walk signals at intersections where I know there are 11+ seconds between the end of FDW and the start of the next green (4s Amber, 2s All-Red, 5s LPI).
So to limit the reduction in signal credibility, I'd suggest the following measures as part of any No RTOR implementation:
- Install [ Bicycles Excepted ] tabs on all No RTOR signs, unless there's a specific reason not to
- Follow
City of Toronto signals operations policy 5.4.3, installing a right turn arrow wherever practical (this has already been done at Progress & Kennedy):
- Avoid unnecessary use of coordinated operations, such as during low-traffic periods or on signals where coordination can't work well anyway due to the spacing of signals or the variety of different travel speeds (e.g. different types of vehicles - buses, bikes etc).
Uncoordinated signals can respond immediately when someone is detected on the side street, unlike coordinated signals who make people wait ages for a particular moment in the cycle before getting a green.
- Consider implementing much more flexible signal timing, including permissive periods, main street detection and phase reservice. This would enable the signal to provide an extra right turn green arrow if there happens to be a gap in main street traffic. However, to be most effective, that would require the main street pedestrian phase to rest in Don't Walk in the absence of pedestrians (though it would usually jump straight to Walk as soon as someone presses the button). This is much more expensive than the current fixed-duration main street phases, so it should be combined with other policy changes to prevent the installation of unwarranted signals and eliminate existing unwarranted signals (to be replaced with geometric design improvemets to improve safety for crossing pedestrians).