Roundabout safety with 2 phase traffic light

[kphoger]

It shouldn't be hard to cross as a pedestrian, as you're effectively crossing a series of one-way roads due to the medians. Just cross when it's clear. It may take longer, but it's not more dangerous.

I think the major difference that's been overlooked so far in this discussion is that, with roundabouts, there's no "clumping" of traffic–so "when it's clear" isn't really a thing.  There's a more-or-less constant flow of traffic coming out of the roundabout.  This is unlike at a stoplight intersection, where–even if a pedestrian doesn't even press the button at all–there will eventually be a red phase to work with.

[jakeroot]

It shouldn't be hard to cross as a pedestrian, as you're effectively crossing a series of one-way roads due to the medians. Just cross when it's clear. It may take longer, but it's not more dangerous.

I think the major difference that's been overlooked so far in this discussion is that, with roundabouts, there's no "clumping" of traffic–so "when it's clear" isn't really a thing.  There's a more-or-less constant flow of traffic coming out of the roundabout.  This is unlike at a stoplight intersection, where–even if a pedestrian doesn't even press the button at all–there will eventually be a red phase to work with.

Well said. Some arterials in the Carmel and Fishers area, with lots of roundabouts, can start to feel like freeways with the constant flow of decently-fast traffic.

There is also a degree of "assertiveness" with roundabouts, since your presence is what stops traffic. At that point, you may have to activate an RRFB, grab a flag, stare at drivers, check multiple lanes, scope out a refuge island ... it's not quite as simple as hitting the ped button at a signal or just waiting for the walk sign in a timed zone (although even then, yes, you have to be cautious when crossing).

As an examples to accompany my above post, even if it is an extreme example, let's look at this roundabout in Carmel, IN (E 116 St @ Hazel Dell Pkwy).

No signals. Cool for cars, sure. But for pedestrians, this is a nightmare. At a minimum, at least four lanes of traffic have to stop to let you go, and who knows how long that will take.

Crossing from one corner to the far corner? That may involve crossing as many as 11 lanes. Holy cow.


(image source: https://www.dronegenuity.com/projects/carmel-indiana-roundabout/)

For almost all of the crossings you would be only doing a 2 lane cross at one time (with the exception of the one southbound leg with a third lane.  There are pedestrian refuge islands between the groups of lanes.   That said this area has very little pedestrian traffic.   If there were more pedestrian traffic Carmel would likely add RRFB flashing beacon lights like they've done at many other roundabouts in the area.

Understood. I actually have family near the Fishers/Carmel border and am quite familiar with the intersection and surrounding neighborhood. I am aware that this area is very car-centric, but you have to wonder how much of that is the result of the infrastructure. Nothing about that roundabout screams "cross me!" Which is a shame, since Carmel and Fishers have quite a lot of good pedestrian paths.

[cjw2001]

Understood. I actually have family near the Fishers/Carmel border and am quite familiar with the intersection and surrounding neighborhood. I am aware that this area is very car-centric, but you have to wonder how much of that is the result of the infrastructure. Nothing about that roundabout screams "cross me!" Which is a shame, since Carmel and Fishers have quite a lot of good pedestrian paths.

I'm in Westfield just barely north of Carmel - I do 6 mile walks several times a week that includes crossing at a roundabout, 2 lanes at a time.   While nowhere near as busy as the Hazel Dell one referenced, I don't have any trouble finding an opportunity to cross.  Much of the time the cars will come to a stop when they see me waiting at the side (sometimes I'll delay approaching until I see a gap just to avoid that).

I also cross the same street at a traditional intersection on my way back from the walk.  I feel much more exposed at the normal intersection than at the roundabout since I have to worry about traffic flow in four directions to cross (main road plus traffic entering and exiting two subdivisions).

[Mr Kite]

I find it interesting that the whole point of roundabouts is seen by many as being that you don't have to stop.

The main benefit of roundabouts is capacity and in appropriate locations* they perform better than traffic signals in this regard.

You will only get to go through without stopping if there's nothing already in the circle. At busy locations during busy periods, you can find yourself waiting a while. Waiting for pedestrians to cross isn't much different in principle. A lot of Europe happily sticks marked crosswalks right before the circle, where pedestrians have priority over traffic. Also, at most roundabouts, you have to slow down before entering as the geometry insists upon it (quite deliberately too), so it's not quite like putting an uncontrolled crosswalk midblock on a high speed road.

*roundabouts are not the magic solution to every scenario. The UK has many roundabouts but many more traffic signals.

[kalvado]

I find it interesting that the whole point of roundabouts is seen by many as being that you don't have to stop.

The main benefit of roundabouts is capacity and in appropriate locations* they perform better than traffic signals in this regard.

You will only get to go through without stopping if there's nothing already in the circle. At busy locations during busy periods, you can find yourself waiting a while. Waiting for pedestrians to cross isn't much different in principle. A lot of Europe happily sticks marked crosswalks right before the circle, where pedestrians have priority over traffic. Also, at most roundabouts, you have to slow down before entering as the geometry insists upon it (quite deliberately too), so it's not quite like putting an uncontrolled crosswalk midblock on a high speed road.

*roundabouts are not the magic solution to every scenario. The UK has many roundabouts but many more traffic signals.

Nope. Capacity-wise roundabouts cover the upper end of stop sign controlled intersection and lower end of traffic light intersections.
Most optimistic recommendations talk about roundabouts at 30k AADT, while traffic lights can handle 40-50k.

[Mr Kite]

Not necessarily. It depends on the geometry of the junction, plus how heavy the flows are in any particular direction i.e. if there's less of a N/S E/W relationship between the main flows. If every situation suggested signals were more efficient, then roundabouts wouldn't be a thing. And like I said above, the UK has plenty of roundabouts but plenty more traffic signals. Nevertheless, it's a worthwhile option to have in the locker. The other advantage I forgot to mention is that it forgoes the need for the cost of signal installation and maintenance.

[kalvado]

No necessarily. It depends on the geometry of the junction, plus how heavy the flows are in any particular direction i.e. if there's less of a N/S E/W relationship between the main flows. If every situation suggested signals were more efficient, then roundabouts wouldn't be a thing. And like I said above, the UK has plenty of roundabouts but plenty more traffic signals. Nevertheless, it's a worthwhile option to have in the locker. The other advantage I forgot to mention is that it forgoes the need for the cost of signal installation and maintenance.

True, roundabouts can be more efficient - primarily for time spent at the intersection - at low traffic flows, or in primarily unidirectional traffic. Traffic lights will be better at the high-flow end of things.
Roundabout is an option - but not a scalable option. There has to be a very good understanding that once traffic pattern changes significantly, a total rebuilt will be needed.

[jakeroot]

Understood. I actually have family near the Fishers/Carmel border and am quite familiar with the intersection and surrounding neighborhood. I am aware that this area is very car-centric, but you have to wonder how much of that is the result of the infrastructure. Nothing about that roundabout screams "cross me!" Which is a shame, since Carmel and Fishers have quite a lot of good pedestrian paths.

I'm in Westfield just barely north of Carmel - I do 6 mile walks several times a week that includes crossing at a roundabout, 2 lanes at a time.   While nowhere near as busy as the Hazel Dell one referenced, I don't have any trouble finding an opportunity to cross.  Much of the time the cars will come to a stop when they see me waiting at the side (sometimes I'll delay approaching until I see a gap just to avoid that).

I also cross the same street at a traditional intersection on my way back from the walk.  I feel much more exposed at the normal intersection than at the roundabout since I have to worry about traffic flow in four directions to cross (main road plus traffic entering and exiting two subdivisions).

I don't mean to imply that crossing at a roundabout is always an inferior experience. I think there are many roundabouts where the experience is more than safe. Pretty much every single-lane roundabout is an exceptional experience.

Where I start to question the experience is when it comes to those who maybe don't move as fast, or those who cannot see as well. For the average person, crossing at a roundabout shouldn't be difficult. If you time it right, it could even be quite fast. But for those who are not as advantageous in their ability to move at a normal pace or see as-well, I do begin to wonder if large roundabouts (like that Hazel Dell example) become an obstacle to everyday tasks like leaving the house, since the protections of a signalized crossing are nowhere to be found.

The Hazel Dell example may be extreme, but then why was it built like that? The roundabout could certainly have been made smaller. The extra slip lanes and triple lane movements seem overkill, and removing them could have improved the pedestrian experience. Not every roundabout needs to be single-lane, but like, really, a double right turn slip lane? Those aren't even normally allowed at signalized intersections. Why here?

Carmel's original roundabouts to the north at E 126 St and E Main are great examples of pedestrian-friendly but high-capacity roundabouts. E 116 St is a busier street than either of those, but it still seems overkill.

[kalvado]

Understood. I actually have family near the Fishers/Carmel border and am quite familiar with the intersection and surrounding neighborhood. I am aware that this area is very car-centric, but you have to wonder how much of that is the result of the infrastructure. Nothing about that roundabout screams "cross me!" Which is a shame, since Carmel and Fishers have quite a lot of good pedestrian paths.

I'm in Westfield just barely north of Carmel - I do 6 mile walks several times a week that includes crossing at a roundabout, 2 lanes at a time.   While nowhere near as busy as the Hazel Dell one referenced, I don't have any trouble finding an opportunity to cross.  Much of the time the cars will come to a stop when they see me waiting at the side (sometimes I'll delay approaching until I see a gap just to avoid that).

I also cross the same street at a traditional intersection on my way back from the walk.  I feel much more exposed at the normal intersection than at the roundabout since I have to worry about traffic flow in four directions to cross (main road plus traffic entering and exiting two subdivisions).

I don't mean to imply that crossing at a roundabout is always an inferior experience. I think there are many roundabouts where the experience is more than safe. Pretty much every single-lane roundabout is an exceptional experience.

Where I start to question the experience is when it comes to those who maybe don't move as fast, or those who cannot see as well. For the average person, crossing at a roundabout shouldn't be difficult. If you time it right, it could even be quite fast. But for those who are not as advantageous in their ability to move at a normal pace or see as-well, I do begin to wonder if large roundabouts (like that Hazel Dell example) become an obstacle to everyday tasks like leaving the house, since the protections of a signalized crossing are nowhere to be found.

The Hazel Dell example may be extreme, but then why was it built like that? The roundabout could certainly have been made smaller. The extra slip lanes and triple lane movements seem overkill, and removing them could have improved the pedestrian experience. Not every roundabout needs to be single-lane, but like, really, a double right turn slip lane? Those aren't even normally allowed at signalized intersections. Why here?

Carmel's original roundabouts to the north at E 126 St and E Main are great examples of pedestrian-friendly but high-capacity roundabouts. E 116 St is a busier street than either of those, but it still seems overkill.

As long as there are safety islands between lane groups, multiple groups would be mostly psychological issue. That doesn't change overall trip length, nor increases number of vehicles to deal with. 
My main issue with roundabouts I crossed myself both walking and driving, is ability of drivers already in roundabout to see pedestrian in time, being able to stop in time, and doing so without blocking the flow - especially during higher traffic periods.

[jakeroot]

As long as there are safety islands between lane groups, multiple groups would be mostly psychological issue. That doesn't change overall trip length, nor increases number of vehicles to deal with. 
My main issue with roundabouts I crossed myself both walking and driving, is ability of drivers already in roundabout to see pedestrian in time, being able to stop in time, and doing so without blocking the flow - especially during higher traffic periods.

Issue is number of lanes combined with number of lane groups. All but one group at the Hazel Dell/116 roundabout is two lanes. Can be a problem for pedestrians due to multiple lanes needing to stop for pedestrians, and (a) pedestrians do not check that both lanes have stopped, and (b) the stopped car creates a blind spot for the other car, and the other car fails to see the pedestrian in time. Double-lane entries are fine but should only be used as necessary. Slip lanes don't need two lanes apart from very rare situations.

Drivers exiting a roundabout are usually not checking for anyone, so yes it is also a problem, but drivers entering roundabout seem focused (rightly so) on cars, failing to consider the potential for other traffic like peds or bikes.

[cjw2001]

I thought some more about the Hazel Dell and 116th roundabout from a pedestrian standpoint, contemplating crossing at the old stoplight vs the new roundabout setup. 

Even with the stoplight the previous setup would be quite dangerous with having to monitor the high volume of turning traffic from multiple directions while having to cross 5 lanes with no refuge areas.   https://www.google.com/maps/@39.9567248,-86.0736785,3a,75y,254.6h,84.29t/data=!3m6!1e1!3m4!1sSQxalH2EwOb0sDrml8CnmQ!2e0!7i3328!8i1664

With the new setup you only have to monitor traffic from a single direction and each crossing is much shorter distance.   As a pedestrian I'd much rather deal with the new setup.  Not saying it couldn't be further improved, just that it is far better than what was there before.

[vdeane]

Studies have, in fact, shown that pedestrians with impaired do have issues with roundabouts due to the sound of circulating traffic, and those problems are much worse at multi-lane roundabouts (in addition, traffic tends to move faster in those, due to straighter entry/exit lanes).  As such, the PROWAG ADA standards (which FHWA has been looking to adopt) require some form of signalization, such as a HAWK or regular traffic light at multi-lane roundabouts.

[kphoger]

Even with the stoplight the previous setup would be quite dangerous with having to monitor the high volume of turning traffic from multiple directions while having to cross 5 lanes with no refuge areas.

But, back then, peds had a stoplight-controlled crossing phase.

[jakeroot]

Even with the stoplight the previous setup would be quite dangerous with having to monitor the high volume of turning traffic from multiple directions while having to cross 5 lanes with no refuge areas.

But, back then, peds had a stoplight-controlled crossing phase.

Indeed. And it should be noted that, no matter how large an intersection, the same three crossing movements are apparent: both right turns (RTOR and right on green), plus left turns. Theoretically, all can be modified to eliminate movements while the walk sign is on, as is often done in the Netherlands.

Studies have, in fact, shown that pedestrians with impaired do have issues with roundabouts due to the sound of circulating traffic, and those problems are much worse at multi-lane roundabouts (in addition, traffic tends to move faster in those, due to straighter entry/exit lanes).  As such, the PROWAG ADA standards (which FHWA has been looking to adopt) require some form of signalization, such as a HAWK or regular traffic light at multi-lane roundabouts.

I was not familiar with those standards until now. I will be reading more about it. Thank you.

[kphoger]

Studies have, in fact, shown that pedestrians with impaired do have issues with roundabouts due to the sound of circulating traffic, and those problems are much worse at multi-lane roundabouts (in addition, traffic tends to move faster in those, due to straighter entry/exit lanes).  As such, the PROWAG ADA standards (which FHWA has been looking to adopt) require some form of signalization, such as a HAWK or regular traffic light at multi-lane roundabouts.

I was not familiar with those standards until now. I will be reading more about it. Thank you.

Ooh, yes, interesting...

https://www.adainfo.org/sites/default/files/1.4-Public-Rights-of-Way-2-slides-per-page.pdf  (see page 46)

[cjw2001]

Even with the stoplight the previous setup would be quite dangerous with having to monitor the high volume of turning traffic from multiple directions while having to cross 5 lanes with no refuge areas.

But, back then, peds had a stoplight-controlled crossing phase.

Understood, but there was left and right turn traffic that conflicted with the walk light.   I'd still take the roundabout over the long crossing distance with a light.

[jakeroot]

Even with the stoplight the previous setup would be quite dangerous with having to monitor the high volume of turning traffic from multiple directions while having to cross 5 lanes with no refuge areas.

But, back then, peds had a stoplight-controlled crossing phase.

Understood, but there was left and right turn traffic that conflicted with the walk light.   I'd still take the roundabout over the long crossing distance with a light.

True, but those issues can be mitigated through other measures as necessary, including improved signalization, leading pedestrian intervals, pedestrian lockout (red arrows during walk phase); plus, there's still all of the additional ADA 'equipment' that you don't have at roundabouts, such as actual walk/don't walk signals and sounds ('walk signal is on to cross...').

I also ran some distance measurements on the old signal vs the new one at Hazel Dell/E 116. The old signal took 233 feet to cross from one corner to the opposite corner. The new roundabout takes about 330 to 400+ feet to cover the same distance. Because crosswalks at roundabouts are offset, there is some level of back-tracking required. Not all of that distance are you exposed to traffic, to be fair, but the same is true at signals where the approach lanes have red lights and do not conflict with the walk sign (apart from right-on-red).

[yand]

No necessarily. It depends on the geometry of the junction, plus how heavy the flows are in any particular direction i.e. if there's less of a N/S E/W relationship between the main flows. If every situation suggested signals were more efficient, then roundabouts wouldn't be a thing. And like I said above, the UK has plenty of roundabouts but plenty more traffic signals. Nevertheless, it's a worthwhile option to have in the locker. The other advantage I forgot to mention is that it forgoes the need for the cost of signal installation and maintenance.

True, roundabouts can be more efficient - primarily for time spent at the intersection - at low traffic flows, or in primarily unidirectional traffic. Traffic lights will be better at the high-flow end of things.
Roundabout is an option - but not a scalable option. There has to be a very good understanding that once traffic pattern changes significantly, a total rebuilt will be needed.

as an adjacent topic, I feel like the poor performance of busy roundabouts in simulations and real life are in large part due to improper technique rather than inherent to roundabouts. People and AI treat roundabouts as right turns rather than freeway merges. If drivers just leave a little bit of space while waiting, they will have room to accelerate and easily take small gaps in traffic, and it will result in much shorter wait times.

[cjw2001]

True, but those issues can be mitigated through other measures as necessary, including improved signalization, leading pedestrian intervals, pedestrian lockout (red arrows during walk phase); plus, there's still all of the additional ADA 'equipment' that you don't have at roundabouts, such as actual walk/don't walk signals and sounds ('walk signal is on to cross...').

I still think comparing apples to apples (what was/is actually there) is valid vs what could have been and might be.  And what is there now is better than what was there.   

I'm still haunted by the memory of seeing a trash truck driver cleaning up a spill in the road on 116th not far from that intersection and driving by thinking "that's not safe" only to see on the news later that night he was killed minutes later by a car that didn't see him.   I just don't trust drivers to see pedestrians and would rather have a shorter distance to cross where I only have to watch for a small subset of the total intersection traffic that is moving in only one direction.