Efficiency and Flexibility: Decoding Dutch Traffic Signal Design | Mobycon Academy
ฝัง
- เผยแพร่เมื่อ 3 มิ.ย. 2024
- What sets traffic signals in the Netherlands apart from their North American counterparts? In our latest Mobycon Academy, Integrated Mobility Consultant Narayan Donaldson explores the features of Dutch traffic signal design that reduce waiting times for all road users- pedestrians, cyclists, and drivers- while improving safety and efficiency. The main differences? The Netherlands uses more fully-protected signal phases, eliminating conflicts between traffic moving in different directions, and the signals respond better to real-time conditions.
See how these principles are applied in practice- and how they could be applied in North American cities seeking to innovate their transportation systems.
Make sure to subscribe to our TH-cam channel, and to Mobycon Academy to stay up to date on future videos: academy.mobycon.com/
Thank you for explaining what should have been obvious to me living in the Netherlands. Interesting story.
Great video! A lot of this is completely possible in the US with existing signal controllers and equipment. Choosing to operate signals this way is another issue. But I believe once you start seeing a fully actuated signal with logic driving coordinated ped crossings, and those ped crossings split in a reasonable way...its hard to turn back. Furthermore, a critical aspect to switching from permissive lefts/rights to protected only is detection...without it, or with poor detection configurations, you get a lot more non-compliance.
To implement this in the states, you also need to have dedicated right turn lanes to control right turn conflicts. There appear to be a lot more of those in the Netherlands. But we could certainly design our intersections to be operated like this without needing to change any laws that I'm aware of.
@@karlmacnair6089 yes, or where you don't have dedicated right turns, use the same geometric ideas of placing crosswalk users ahead and to the right a bit more than we typically do.
But not only do we need right turn lanes (that's not that uncommon), it's also that you have to control them by signals. We typically just assume right turning drivers can turn right whenever, so dont always signal that movement or use detection.
That was the case in the video below. We have a right turn lane, we have it controlled by signals...but didn't include detection. Lesson learned...add detection.
th-cam.com/video/N3DJHJ63hEo/w-d-xo.html
The biggest challenges we are facing implementing Dutch-style signal operations in North America are the clearance times and the control structure.
Dutch red clearance times are two dimensional, so the clearance duration depends on the following phase. For example at 3:38 you can see that the red clearance from the bicycle signal to the right turn signal is 0 seconds, while at 4:58 you can see that the red clearance time from that same bicycle signal to the side street phase is about 4 seconds. Although 2D clearances are legally possible in North America, we have not found any controllers which natively support them and are also compatible with American cabinet standards.
The other challenge is the control structure. Most North American controllers depend on the Ring & Barrier structure, which works great for very simple intersections, but becomes exponentially more complex and constrained as additional features such as right turn signals are introduced. The Block (a.k.a. Module) structures typically used in the Netherlands allow each phase to determine its own compatibility based on the conflict matrix, and automatically make use of unused time from other phases, without needing to manually program "overlaps" as you would with a Ring & Barrier structure.
@@narayandonaldsonQuite right. Both entry time and exit times are considered (oprijtijd en ontruimingstijd) and variable timing based on conflict areas. This data is entered as a matrix into the controller. North American controllers don't have this. This is why actuation on Dutch signals works well in the first place. For a technical explanation, see here: th-cam.com/video/7KPGVP85WpU/w-d-xo.html
@@karlmacnair6089depends. The Netherlands has very short minimum green times as well as allowing negative clearing time.
This was great. Thank you.
Now I know why the bike light at some intersections becomes green as soon as I surrender and put my foot down.
One thing I think would have been a beneficial add; how pedestrian crossings really do not have a long "yellow or green" phase or flashing red as in the states. Basically the only pedestrians expected to cross are the ones at the light as it turns green.
Much better than in NYC where the blinking phase leads to a lot of conflict.
One of the common issues at Dutch signals is that the bicycle detector is only about 25 m before the stop line, which is not quite enough time to call a green light if the signal is resting in green for a conflicting phase. This often causes the cyclist to stop briefly before the light changes to green. This situation can be reduced by setting all conflicting phases to rest in red and by disabling "meeverlengen" (extending with overlaps), like the signal at 3:13. Or alternatively the bicycle detectors can be moved further away (e.g. to 35 m).
Even at intersections where phases do default to red, there are often bicycle paths which lack an upstream bicycle detector, which means that although the signal can provide a green instantly when a cyclist is detected, it doesn't detect cyclists until they reach the stop line, by which time they have already lost their momentum.
Placing the detector further away has the downside of making the estimated arrival time at the signal less accurate, increasing the risk of extending the green too long, or not extending a green in a situation when the cyclist actually could have made it through with an extension. So detectors at 35 m should ideally measure the speed of the cyclist (using a camera or dual detectors) to improve the estimate.
@@narayandonaldsonor maybe cycle a bit slower within a built up area? You can't have it both ways: either regulate for all or choose one mode that has priority. Seems you forgot you react as a cyclist, move the detection point 10 meters and the other detection points HAVE to change too.
@@Paul_Cwhy slow down? Momentum for cyclists is a very important thing. People cycle in cities cause it’s more convenient and faster (most of the time) if signals aren’t improved than you’re stuck with non moving tech.
Another consideration is signal timing, which the actuation is actually bolted on top of (conflict management is done by the timing circuits). Ontario Traffic Man did an excellent video on it: th-cam.com/video/7KPGVP85WpU/w-d-xo.html
It's actually the same person in both videos.
☀️ "Promo SM"