47 ms·

I'm a little skeptical that it would be safer for pedestrians and bikes. Bikes need to merge with traffic already in the roundabout instead of simply going str

I'm a little skeptical that it would be safer for pedestrians and bikes. Bikes need to merge with traffic already in the roundabout instead of simply going straight through an intersection, and pedestrians would need to always check for oncoming cars rather than being able to mostly rely on lights telling them when they can cross. But then most of my experience with roundabouts is with large, multilane roundabouts so the traffic is still pretty fast - maybe smaller roundabouts with significantly slower traffic would be safer than I'm picturing.

ryukafalz 5y ago
If you have a bike path alongside the car lanes, they seem to work really well. See this one for example - the red paved path is for bikes: https://i.redd.it/1vfvbvntky411.jpg https://i.redd.it/1vfvbvntky411.jpg The bike path crosses before the point where cars merge in, so the driver's attention isn't split. Also, it may not be obvious from the picture, but the car lanes are raised to the level of the bike path where they intersect.
- raffraffraff 5y ago
  The problem with these roundabouts is that a car driver cannot simply "exit the roundabout". They have to watch front, side and rear for cyclists. IMHO a roundabout should only ever impede traffic on entry. Once you're in the roundabout your journey around and out should be entirely unimpeded. Following the basic principles, there should never be a situation where a vehicle is going around the outter circumference of the roundabout blocking a vehicle from exiting. (Even a bicycle)
  - Kim_Bruning 5y ago
    Only the front in theory: the crossing with the bicycle path is fairly close to 90 degrees. Even so -in practice- you should be checking side and rear anytime you make any turn on any kind of intersection (you fail the dutch driving exam if you forget even just once). I agree that even ideally-er the journey out should be 100% unimpeded, but in practice this setup seems to work ok. See also video I posted for you, above. :-)
    - whatevertrevor 5y ago
      What if they moved the bike/pedestrian crossings to further along the lanes away from the roundabout? This would give more room for the cars to exit the roundabout without encountering immediate cross traffic, and more "collection space" on the exit itself so if a driver does have to stop to let a pedestrian/cyclist through they don't impede the roundabout traffic.
      - Kim_Bruning 5y ago
        Posted video link of this type of roundabout in action just above ( https://news.ycombinator.com/item?id=27415631 https://news.ycombinator.com/item?id=27415631 ) There's just sufficient space for 1 car to tuck in each time. Note that the priority here is not necessarily to maximize car throughput at all costs; but to ensure safety for all modes of traffic; and to do so inside a given amount of space and cost. Given a different set of parameters, a different intersection type might be chosen. (+) That said, note in the video that there's a significant number of people using the bicycle lane and pedestrian path; I think they're carrying more people than the car lanes here. I'm not a traffic expert, but forcing the denser traffic to take a longer route in favor of the less dense traffic does not seem like the right choice here. ;-)
        raffraffraff 5y ago
        It seems to work during that 5 minute segment, but at 1.45 it looks like this system could get gridlocked. If vehicles are blocked from exiting a roundabout for long enough, there's a risk that the whole thing will fill up. When this happens, it can only get unblocked quickly if the vehicles blocking the exits wish to leave via those exits. Edit: nope, the system can continue to move, but when more exits are blocked, the system throughput drops sharply to zero. I'd love to see this roundabout during rush hour. For cyclists safety, this system works because the cycle lane is so far outside the vehicle lane, and it is right before pedestrian crossing. Most of the roundabouts with bicycle lanes that I have encountered, they just used a different coloured tarmac at the edge to show cyclists where they should risk their lives. They're not at 90 degrees to the cars, they're literally side-by-side. In that situation, cyclists are in greater danger of getting side-swiped by a car exiting the roundabout.
tvphan 5y ago
Uh I live next to a roundabout (single lane) and it’s the safest thing ever. I’m guessing you’re American?
orthoxerox 5y ago
Proper Dutch-style roundabouts have the path for cyclists that follows the sidewalk, not the car lane. https://commons.m.wikimedia.org/wiki/File:Alternative_dutch_roundabout.png https://commons.m.wikimedia.org/wiki/File:Alternative_dutch_... Multi-lane intersections on major roads should just have a left-turn lane with a separate signal phase.
- raffraffraff 5y ago
  Unless it's a motorway-style roundabout with many multiple lanes and high velocity traffic, I much prefer to "pretend to be a car" when I cycle through roundabouts. The busiest roundabout I regularly cycle through has 2 lanes and 6 roads intersect at it. I can't imagine wanting to use it if I had to operate outside the basic rules of roundabouts.
  - Kim_Bruning 5y ago
    Something like this? https://www.youtube.com/watch?v=FR5l48_h5Eo https://www.youtube.com/watch?v=FR5l48_h5Eo Mode separation for motor, bicycle, and foot traffic; all traffic crosses as close to 90 degrees as possible (for visibility reasons); priority markings on pavement; speed bump on one of the entrances to slow traffic; and a lot more features I couldn't tell you about. ( I sometimes watch https://www.youtube.com/c/NotJustBikes/ https://www.youtube.com/c/NotJustBikes/ , but I'm no traffic expert )
- Doxin 5y ago
  This is incorrect. To a rounding error all the roundabouts I have encountered (in the Netherlands, where I live) have the cycle lane follow the car lane, and give cyclists right-of-way over the cars. You'll occasionally see one where the cycle lane is separated from the roundabout, but I only know of one from the top of my head.

ryukafalz 5y ago

If you have a bike path alongside the car lanes, they seem to work really well. See this one for example - the red paved path is for bikes: https://i.redd.it/1vfvbvntky411.jpg https://i.redd.it/1vfvbvntky411.jpg The bike path crosses before the point where cars merge in, so the driver's attention isn't split. Also, it may not be obvious from the picture, but the car lanes are raised to the level of the bike path where they intersect.

raffraffraff 5y ago
The problem with these roundabouts is that a car driver cannot simply "exit the roundabout". They have to watch front, side and rear for cyclists. IMHO a roundabout should only ever impede traffic on entry. Once you're in the roundabout your journey around and out should be entirely unimpeded. Following the basic principles, there should never be a situation where a vehicle is going around the outter circumference of the roundabout blocking a vehicle from exiting. (Even a bicycle)
- Kim_Bruning 5y ago
  Only the front in theory: the crossing with the bicycle path is fairly close to 90 degrees. Even so -in practice- you should be checking side and rear anytime you make any turn on any kind of intersection (you fail the dutch driving exam if you forget even just once). I agree that even ideally-er the journey out should be 100% unimpeded, but in practice this setup seems to work ok. See also video I posted for you, above. :-)
  - whatevertrevor 5y ago
    What if they moved the bike/pedestrian crossings to further along the lanes away from the roundabout? This would give more room for the cars to exit the roundabout without encountering immediate cross traffic, and more "collection space" on the exit itself so if a driver does have to stop to let a pedestrian/cyclist through they don't impede the roundabout traffic.
    - Kim_Bruning 5y ago
      Posted video link of this type of roundabout in action just above ( https://news.ycombinator.com/item?id=27415631 https://news.ycombinator.com/item?id=27415631 ) There's just sufficient space for 1 car to tuck in each time. Note that the priority here is not necessarily to maximize car throughput at all costs; but to ensure safety for all modes of traffic; and to do so inside a given amount of space and cost. Given a different set of parameters, a different intersection type might be chosen. (+) That said, note in the video that there's a significant number of people using the bicycle lane and pedestrian path; I think they're carrying more people than the car lanes here. I'm not a traffic expert, but forcing the denser traffic to take a longer route in favor of the less dense traffic does not seem like the right choice here. ;-)
      - raffraffraff 5y ago
        It seems to work during that 5 minute segment, but at 1.45 it looks like this system could get gridlocked. If vehicles are blocked from exiting a roundabout for long enough, there's a risk that the whole thing will fill up. When this happens, it can only get unblocked quickly if the vehicles blocking the exits wish to leave via those exits. Edit: nope, the system can continue to move, but when more exits are blocked, the system throughput drops sharply to zero. I'd love to see this roundabout during rush hour. For cyclists safety, this system works because the cycle lane is so far outside the vehicle lane, and it is right before pedestrian crossing. Most of the roundabouts with bicycle lanes that I have encountered, they just used a different coloured tarmac at the edge to show cyclists where they should risk their lives. They're not at 90 degrees to the cars, they're literally side-by-side. In that situation, cyclists are in greater danger of getting side-swiped by a car exiting the roundabout.

tvphan 5y ago

Uh I live next to a roundabout (single lane) and it’s the safest thing ever. I’m guessing you’re American?

orthoxerox 5y ago

Proper Dutch-style roundabouts have the path for cyclists that follows the sidewalk, not the car lane. https://commons.m.wikimedia.org/wiki/File:Alternative_dutch_roundabout.png https://commons.m.wikimedia.org/wiki/File:Alternative_dutch_... Multi-lane intersections on major roads should just have a left-turn lane with a separate signal phase.

raffraffraff 5y ago
Unless it's a motorway-style roundabout with many multiple lanes and high velocity traffic, I much prefer to "pretend to be a car" when I cycle through roundabouts. The busiest roundabout I regularly cycle through has 2 lanes and 6 roads intersect at it. I can't imagine wanting to use it if I had to operate outside the basic rules of roundabouts.
- Kim_Bruning 5y ago
  Something like this? https://www.youtube.com/watch?v=FR5l48_h5Eo https://www.youtube.com/watch?v=FR5l48_h5Eo Mode separation for motor, bicycle, and foot traffic; all traffic crosses as close to 90 degrees as possible (for visibility reasons); priority markings on pavement; speed bump on one of the entrances to slow traffic; and a lot more features I couldn't tell you about. ( I sometimes watch https://www.youtube.com/c/NotJustBikes/ https://www.youtube.com/c/NotJustBikes/ , but I'm no traffic expert )
Doxin 5y ago
This is incorrect. To a rounding error all the roundabouts I have encountered (in the Netherlands, where I live) have the cycle lane follow the car lane, and give cyclists right-of-way over the cars. You'll occasionally see one where the cycle lane is separated from the roundabout, but I only know of one from the top of my head.