Crash prone 'modern roundabouts'

[jakeroot]

New turbo roundabout in McCalla, Alabama:

https://maps.app.goo.gl/Cghi3jxqWMmsyKN2A

[tradephoric]

Crashes increase but less severe at US 41 roundabouts vs. prior intersection
Despite an uptick in accidents in U.S. 41 roundabouts over prior intersection, the Sarasota-Manatee Metropolitan Planning Organization says they are functioning well.
https://www.yourobserver.com/news/2024/feb/08/crashes-increase-roundabouts-intersection/

The Gulfstream Avenue at US-41 roundabout in Sarasota has gone from 20 crashes per year in 2019 (before the roundabout and pre-pandemic) to 120 crashes in 2023 (post roundabout).  Compare that to the 10th Street at US-41 roundabout just down the street that went from 20 crashes in 2016 to 20 crashes in 2023.  One of these roundabouts is not like the other.

Gulfstream Avenue @ US-41:


10th Street @ US-41:

[kphoger]

The Gulfstream Avenue at US-41 roundabout in Sarasota has gone from 20 crashes per year in 2019 (before the roundabout and pre-pandemic) to 120 crashes in 2023 (post roundabout).  Compare that to the 10th Street at US-41 roundabout just down the street that went from 20 crashes in 2016 to 20 crashes in 2023.  One of these roundabouts is not like the other.

  I'd say not!

[tradephoric]

Bay Area's high-tech new turbo roundabout is huge disaster
https://www.dailymail.co.uk/news/article-14588847/bay-area-turbo-roundabout-crashes-driver-accidents.html

There have been a lot of crashes at the new State Route 25/Highway 156 turbo roundabout.  According to this article there were 77 crashes in 2024 upon completion of the roundabout.  That compares to 20 crashes in 2023 and 45 crashes in 2022 (but those were construction years?).

This article from August 2021 states that there were 126 collisions between 2009 and 2018 with two fatalities at the intersection. There have been no reported fatalities since the new roundabout came online but it's on pace to see more crashes in 2 years as a turbo roundabout than it did in a decade as a signalized intersection (looking at crash data prior to construction). 

New 'turbo' roundabout planned for 25/156 intersection
https://benitolink.com/new-turbo-roundabout-planned-for-25-156-intersection/

[tradephoric]

There are clear overhead signs indicating that drivers are approaching the State Route 25/Highway 156 turbo roundabout and there are no fixed objects in the raised central island.  Drivers could catch some air driving over the central island if they aren't paying attention but at least they aren't going to crash head first into a tree, retaining wall, or 'decorative boulders'.  Those types of fixed objects crashes have led to fatalities at other roundabouts nationwide.  But will the raised curbing separating the lanes of the turbo roundabout create issues for motorcyclists?  Has that been an issue in the Netherlands (specifically for motorcyclists) where these turbo roundabouts are so common?

I'm open to the idea that turbo roundabouts may perform better than a standard modern roundabout, but let's look at data before pushing the next "flavor of the month" roundabout design in America.  Even if turbo roundabouts work really well in Netherlands, no guarantee they will work well here.  I believe another turbo roundabout was built in Florida and Jake just mentioned one built in Alabama.  They are slowly catching on here but are they really better?

[kphoger]

By the way, AI image generators are awful at roundabouts.  Some of them make me laugh out loud.  Lots of fun.

[CoreySamson]

I would be interested to go back through the fates of the roundabouts listed early on in the thread to see what happened to them now that it has been a long time since this thread was instituted. Did the accident rates drop? Were the roundabouts redesigned or replaced? Would be a fascinating study.

[kalvado]

I would be interested to go back through the fates of the roundabouts listed early on in the thread to see what happened to them now that it has been a long time since this thread was instituted. Did the accident rates drop? Were the roundabouts redesigned or replaced? Would be a fascinating study.

I may not be up to the full task, but first one mentioned here in 2015, in Ann Arbor, was still shining in 2023
https://www.michiganautolaw.com/blog/2024/08/12/michigan-roundabouts-most-dangerous-2023/

[tradephoric]

I would be interested to go back through the fates of the roundabouts listed early on in the thread to see what happened to them now that it has been a long time since this thread was instituted. Did the accident rates drop? Were the roundabouts redesigned or replaced? Would be a fascinating study.

https://www.clickondetroit.com/news/local/2024/07/03/these-are-the-20-most-dangerous-michigan-intersections/

blake77777

Those 4 roundabouts have injury rates of 7%, 7%, 9%, and 4% compared to 24%, 30%, 25% and 33% from the four worst traditional intersections. The data is right before your eyes, roundabouts save lives and ER visits. edited

All the roundabouts listed here are over a decade old at this point.  Blake77777 makes an interesting comment.  Yes, you may only have a 7% chance of getting injured if you are involved in a crash at a roundabout (compared to ~25% at traditional intersections), but Blake77777 isn't taking into account the fact that you have a much higher chance of getting in a crash at the roundabout to begin with.  The local tow-company is getting a lot more work cleaning up all the fender benders but is it really that much safer?  A Minnesota study previously discussed on this thread found that there was actually an increase in injury crashes when looking at the safety of 2x2 and 3x2 roundabouts built in that state.

[kphoger]

Oh, wow!  Look at those injury totals!

Add up all five of the roundabouts on the list, and only 41 total injuries.  That's fewer than either of #1 or #3.  That's awesome.

[kalvado]

Oh, wow!  Look at those injury totals!

Add up all five of the roundabouts on the list, and only 41 total injuries.  That's fewer than either of #1 or #3.  That's awesome.

(image clipped)

Daily traffic and peak traffic numbers are still a missing part of the puzzle.

[kphoger]

Daily traffic and peak traffic numbers are still a missing part of the puzzle.

But saying that an intersection with so much fewer injuries is among the 'most dangerous' does seem to be misleading.  When the list only considers total crashes without weighing them by injuries, then it misses the whole reason roundabouts are advocated in the first place.

[paulthemapguy]

Lurking variable: the recent reconstruction of these intersections into roundabouts is likely because they were high-crash locations to begin with. What would really be useful is a set of data focused on each particular roundabout location, comparing the crash frequency and injury frequency with the same rates measured before the construction of the roundabout.

[kalvado]

Lurking variable: the recent reconstruction of these intersections into roundabouts is likely because they were high-crash locations to begin with. What would really be useful is a set of data focused on each particular roundabout location, comparing the crash frequency and injury frequency with the same rates measured before the construction of the roundabout.

Again, digging out old data isn't simple.. One datapoint for state& W ellisworth
https://www.mlive.com/news/ann-arbor/2015/03/roundabout_on_ann_arbors_south.html

The intersection had 16 crashes in 2012

Fine print: Costco opened there end of July, so annual number would be  1.5-2x higher (guessing). 1/3-1/4 injury crashes, so guessing 8-10 injury crashes out of 30 total per year with Costco full steam (guesstimate).
Actual data today is 5 injury out of 126 total.

Not as impressive of a reduction IMHO. 

[Rothman]

Looks like 11 Mile Road/I-696 @ Van Dyke needs a roundabout.

[kalvado]

Looks like 11 Mile Road/I-696 @ Van Dyke needs a roundabout.

Looks like DOTs should not be allowed to design roundabout until supervised by a responsible adult.

[tradephoric]

Take Telegraph Road.  There are a couple of high crash locations along that road (12 Mile and Schoolcraft).

Oh, wow!  Look at those injury totals!

Add up all five of the roundabouts on the list, and only 41 total injuries.  That's fewer than either of #1 or #3.  That's awesome.

That's true.  But if you picked at random 5 signalized intersections that carry comparative traffic to the 5 roundabouts on that list, you'd probably see around 40 total injuries too.  The difference is total crashes would be around 150-200 as opposed to 623.  The same number of injury crashes but a tripling to quadrupling in total crashes.  That's been the theme when agencies decide to replace a traditional traffic signal with a complex 2x2 or 3x2 roundabout. 

[Rothman]

Looks like 11 Mile Road/I-696 @ Van Dyke needs a roundabout.

Looks like DOTs should not be allowed to design roundabout until supervised by a responsible adult.

At least I make sense.

[kalvado]

Looks like 11 Mile Road/I-696 @ Van Dyke needs a roundabout.

Looks like DOTs should not be allowed to design roundabout until supervised by a responsible adult.

At least I make sense.

That's fine, make whatever you want - just don't design things.

[Rothman]

Looks like 11 Mile Road/I-696 @ Van Dyke needs a roundabout.

Looks like DOTs should not be allowed to design roundabout until supervised by a responsible adult.

At least I make sense.

That's fine, make whatever you want - just don't design things.

Ok.  We'll let engineers design the roundabout that will lower the injuries at the worst intersection in Michigan.

[jakeroot]

11 Mile / I-696 / Van Dyke is basically a giant signalized hamburger roundabout.

[kalvado]

Looks like 11 Mile Road/I-696 @ Van Dyke needs a roundabout.

Looks like DOTs should not be allowed to design roundabout until supervised by a responsible adult.

At least I make sense.

That's fine, make whatever you want - just don't design things.

Ok.  We'll let engineers design the roundabout that will lower the injuries at the worst intersection in Michigan.

Are they going to ask for more H-1Bs for that?

[Rothman]

Looks like 11 Mile Road/I-696 @ Van Dyke needs a roundabout.

Looks like DOTs should not be allowed to design roundabout until supervised by a responsible adult.

At least I make sense.

That's fine, make whatever you want - just don't design things.

Ok.  We'll let engineers design the roundabout that will lower the injuries at the worst intersection in Michigan.

Are they going to ask for more H-1Bs for that?

Not sure what you mean by this, but there's a very ugly worst case scenario behind your intent.

[kalvado]

Looks like 11 Mile Road/I-696 @ Van Dyke needs a roundabout.

Looks like DOTs should not be allowed to design roundabout until supervised by a responsible adult.

At least I make sense.

That's fine, make whatever you want - just don't design things.

Ok.  We'll let engineers design the roundabout that will lower the injuries at the worst intersection in Michigan.

Are they going to ask for more H-1Bs for that?

Not sure what you mean by this, but there's a very ugly worst case scenario behind your intent.

Just simple understanding that roundabout design and use isn't developed in US.
There is seemingly no progress in developing the understanding in the past 10 years, so learning from outside experts is the only option.
On a similar note, FAA - another A under the same secretary of transportation - turned to the world expertise after 737MAX fiasco. What followed was a demonstration of huge problems at FAA and Boeing.
I don't want the same realization at FHWA to cost hundreds of lives...

[tradephoric]

I had never seen this video before yesterday, but it is super interesting.  The roundabout uses stop lights for the pedestrian crossing and the video does a great job at touching on all the issues that come from that.  It's a very unique roundabout.

@meerkatmcr
So, at a HAWK signal, double alternating flashing red means "proceed if crossing clear", and at a railroad crossing it means "stop or you will be hit by a train". This is certainly an interesting design choice...

The video also talks about the use of HAWK signals and meerkatmcr made a very interesting comment. It's not just at railroad crossings either.  You come up to a school bus with alternating flashing red balls and it doesn't mean you can proceed with caution when you think it's safe to go.  How did the FHWA even approve the alternating flashing red at the HAWK?  I'm sure someone here can give a very valid reason why the HAWK signal operates the way it does, but it just seems so wrong.