Crash prone 'modern roundabouts'

[7/8]

I'm struggling to see how someone could defend this, it just seems wrong to me!

Pretty obvious, if you zoom out, you can clearly see which road goes through and which is turning.
And if average Joe is so stupid that he cannot see the big picture while driving through, that is not engineer's problem!.

I probably could have been more clear here, but I was referring to the solid white line in the roundabout. The sign on Cherry St says you can go straight or right, but the solid white line (next to the red car in the satellite image) suggests that you can only go right, since going straight would require you to cross a solid white line.

As to the problem of drivers thinking Cherry St to N Detroit Ave is "straight" instead of a right-turn, I can see both sides of the coin. And though I understand your point, the fact is (to a certain extent) we need to design roads for stupid drivers because many drivers are stupid. Especially when it comes to roundabouts (in North America anyway), as all the crash statistics in this thread prove.

[jeffandnicole]

I'm struggling to see how someone could defend this, it just seems wrong to me!

Pretty obvious, if you zoom out, you can clearly see which road goes through and which is turning.
And if average Joe is so stupid that he cannot see the big picture while driving through, that is not engineer's problem!.

I probably could have been more clear here, but I was referring to the solid white line in the roundabout. The sign on Cherry St says you can go straight or right, but the solid white line (next to the red car in the satellite image) suggests that you can only go right, since going straight would require you to cross a solid white line.

If anyone commenting missed that, then they need to turn in their roadgeek license!   

[kalvado]

I'm struggling to see how someone could defend this, it just seems wrong to me!

Pretty obvious, if you zoom out, you can clearly see which road goes through and which is turning.
And if average Joe is so stupid that he cannot see the big picture while driving through, that is not engineer's problem!.

I probably could have been more clear here, but I was referring to the solid white line in the roundabout. The sign on Cherry St says you can go straight or right, but the solid white line (next to the red car in the satellite image) suggests that you can only go right, since going straight would require you to cross a solid white line.

As to the problem of drivers thinking Cherry St to N Detroit Ave is "straight" instead of a right-turn, I can see both sides of the coin. And though I understand your point, the fact is (to a certain extent) we need to design roads for stupid drivers because many drivers are stupid. Especially when it comes to roundabouts (in North America anyway), as all the crash statistics in this thread prove.

I would interpret "straight" on that lane as "Straight to N Detroit", and "right" - as "right to Emmet st. off N Detroit".
If they had that in mind .. but probably not - sign at the roundabout entry shows that right lane goes to Detroit and Berdan, and left.... to Berdan and Detroit as well.

Talk about levels of confusion!

[7/8]

I'm struggling to see how someone could defend this, it just seems wrong to me!

Pretty obvious, if you zoom out, you can clearly see which road goes through and which is turning.
And if average Joe is so stupid that he cannot see the big picture while driving through, that is not engineer's problem!.

I probably could have been more clear here, but I was referring to the solid white line in the roundabout. The sign on Cherry St says you can go straight or right, but the solid white line (next to the red car in the satellite image) suggests that you can only go right, since going straight would require you to cross a solid white line.

As to the problem of drivers thinking Cherry St to N Detroit Ave is "straight" instead of a right-turn, I can see both sides of the coin. And though I understand your point, the fact is (to a certain extent) we need to design roads for stupid drivers because many drivers are stupid. Especially when it comes to roundabouts (in North America anyway), as all the crash statistics in this thread prove.

I would interpret "straight" on that lane as "Straight to N Detroit", and "right" - as "right to Emmet st. off N Detroit".
If they had that in mind .. but probably not - sign at the roundabout entry shows that right lane goes to Detroit and Berdan, and left.... to Berdan and Detroit as well.

Talk about levels of confusion!

Interesting, Google Maps doesn't show Emmet St connecting to N Detroit Ave, but this GSV from Sept 2015 suggests it does. The bolded sentence above is another good point of potential motorist confusion!




Also, Google Maps shows (a second Cherry St?) connecting to the roundabout at W Manhattan Blvd going the wrong way!



But GSV confirms this is an error and that this street doesn't connect to the roundabout.



The outdated satellite imagery doesn't help with all this confusion

[jeffandnicole]

Uh....................

[7/8]

Uh....................

I'm not sure why you're confused. I posted the GSV image to show that the street in the background with the houses on it doesn't connect to the roundabout (there's grass between the two), whereas Google Maps seems to show them overlapping.

[kphoger]

Uh....................

I'm not sure why you're confused. I posted the GSV image to show that the street in the background with the houses on it doesn't connect to the roundabout (there's grass between the two), whereas Google Maps seems to show them overlapping.

I'm confused by cars driving in opposite directions on the same roundabout.

[7/8]

Uh....................

I'm not sure why you're confused. I posted the GSV image to show that the street in the background with the houses on it doesn't connect to the roundabout (there's grass between the two), whereas Google Maps seems to show them overlapping.

I'm confused by cars driving in opposite directions on the same roundabout.

Holy crap, I guess that's what he was referring to! I should post that in the Google Street View thread

UPDATE: Here's my post with more views of the wrong way driver: https://www.aaroads.com/forum/index.php?topic=2648.msg2165328#msg2165328

[kalvado]

Uh....................

I'm not sure why you're confused. I posted the GSV image to show that the street in the background with the houses on it doesn't connect to the roundabout (there's grass between the two), whereas Google Maps seems to show them overlapping.

FOr me new alignment shows up when I untilt the image, old one with tilted image.

And Emmet street is clearly connected to Detroit just off roundabout, maybe 70 feet down the street.
https://www.google.com/maps/place/Toledo,+OH/@41.6863319,-83.5568115,19z/data=!3m1!1e3!4m2!3m1!1s0x883b872dfc1e4e79:0x7c3cc89f453ac345
https://www.google.com/maps/place/Toledo,+OH/@41.6862413,-83.5560435,3a,66.8y,90t/data=!3m4!1e1!3m2!1sohHAkXupMp-umJA_rgki5w!2e0!4m2!3m1!1s0x883b872dfc1e4e79:0x7c3cc89f453ac345

[jeffandnicole]

Uh....................

I'm not sure why you're confused. I posted the GSV image to show that the street in the background with the houses on it doesn't connect to the roundabout (there's grass between the two), whereas Google Maps seems to show them overlapping.

I'm confused by cars driving in opposite directions on the same roundabout.

Holy crap, I guess that's what he was referring to!

That's exactly what I was referring to!!!!

[jakeroot]

What if you just removed the confusing markings? Install overhead signage that clearly shows the straight movement as being a "slight right then left". Or even change the arrows so that they point slightly left instead of straight.

I don't think you need to remove any lanes to fix the problem here. Although apparently we need more right-facing chevrons, because there's evidently people out there that still don't know which way to go around a roundabout....

[kalvado]

What if you just removed the confusing markings? Install overhead signage that clearly shows the straight movement as being a "slight right then left". Or even change the arrows so that they point slightly left instead of straight.

I don't think you need to remove any lanes to fix the problem here. Although apparently we need more right-facing chevrons, because there's evidently people out there that still don't know which way to go around a roundabout....

Actually I was staring at the map trying to visualize traffic patterns.. Looks like there is a massive amount of rebuild going on in the area - I-75 bridge, those two roundabouts, C/D road to I-75. And area looks mostly residential, with lots of homes. Commuter traffic struggling during roadwork in crazy traffic pattern?
Google also shows another roundabout in progress about 1/4 mile west.. Looks like a major realignment of everything..

[lordsutch]

What if you just removed the confusing markings? Install overhead signage that clearly shows the straight movement as being a "slight right then left". Or even change the arrows so that they point slightly left instead of straight.

Most non-signalized 2x2 or 2x1 roundabouts in the UK don't have lane markings, for what it's worth.

I don't think you need to remove any lanes to fix the problem here. Although apparently we need more right-facing chevrons, because there's evidently people out there that still don't know which way to go around a roundabout....

On the other hand, chevrons (along with a mandatory turn one-way arrow) are probably more, not less, emphasized in the UK (except mini-roundabouts, for obvious reasons) than most US installations. Using a W1-6 on top of the chevrons would probably get the point across better, but for some reason MUTCD doesn't specify that.

[tradephoric]

It would be interesting to see if this design would work or if the traffic volumes turning onto Berdan Ave are too high (requiring the dual circulating lanes).....

[johndoe]

I have a cheaper fix =)

What if you just removed the confusing markings?

Do you find the circulating roadway markings/arrows more confusing than the approach fishhooks?  Usually the normal arrows in the circulating roadway align nicely with the exit (unless the angle between legs is big like this), but it makes a lot of sense to me that "through"="exiting roundabout" and "left"="circulating roundabout" but there's a decent chance I've looked at these things too long =)  I guess I've never understood the desire to remove circulating markings. 

IMO if anything is confusing it's the fishhooks.  We've had lots of debate in our office about their effectiveness!  Some people think the angle between the arrows is most important, while others think it's more about demonstrating which lane can exit at which departure (for instance in the left lane you cannot use the FIRST exit but you can use the second and third).  Just be glad this is only four-legged; things get really weird with a fifth leg!

[cjw2001]

I'm struggling to see how someone could defend this, it just seems wrong to me!

Pretty obvious, if you zoom out, you can clearly see which road goes through and which is turning.
And if average Joe is so stupid that he cannot see the big picture while driving through, that is not engineer's problem!.

I probably could have been more clear here, but I was referring to the solid white line in the roundabout. The sign on Cherry St says you can go straight or right, but the solid white line (next to the red car in the satellite image) suggests that you can only go right, since going straight would require you to cross a solid white line.

As to the problem of drivers thinking Cherry St to N Detroit Ave is "straight" instead of a right-turn, I can see both sides of the coin. And though I understand your point, the fact is (to a certain extent) we need to design roads for stupid drivers because many drivers are stupid. Especially when it comes to roundabouts (in North America anyway), as all the crash statistics in this thread prove.

I would interpret "straight" on that lane as "Straight to N Detroit", and "right" - as "right to Emmet st. off N Detroit".
If they had that in mind .. but probably not - sign at the roundabout entry shows that right lane goes to Detroit and Berdan, and left.... to Berdan and Detroit as well.

Talk about levels of confusion!

Interesting, Google Maps doesn't show Emmet St connecting to N Detroit Ave, but this GSV from Sept 2015 suggests it does. The bolded sentence above is another good point of potential motorist confusion!




Also, Google Maps shows (a second Cherry St?) connecting to the roundabout at W Manhattan Blvd going the wrong way!



But GSV confirms this is an error and that this street doesn't connect to the roundabout.



The outdated satellite imagery doesn't help with all this confusion

I've corrected the geometry on the second cherry street to the left in mapmaker (it wasn't connected to the roundabout, just drawn too far to the right).   I also removed the building outline that was sitting on top of the roundabout and cleaned up some of the other geometry.   Also connected Emmet Street and split the lanes on North Detroit north of the roundabout.   Will require google moderation for the final touches to set the new lanes to the right highway priority.

Give it 24 hours for everything to show up on maps.

[jeffandnicole]

What if you just removed the confusing markings? Install overhead signage that clearly shows the straight movement as being a "slight right then left". Or even change the arrows so that they point slightly left instead of straight.

Most non-signalized 2x2 or 2x1 roundabouts in the UK don't have lane markings, for what it's worth.

Most NJ traffic circles don't have lane markings either.  People do have a tendency to creep into other lanes when they're not supposed to.  Plenty of horn honking.  And yet, they don't seem to have the issues many others report at these fully parked roundabouts.

[tradephoric]

Here is a good streetview compilation of the problem.  In the very first picture there is a sign that directs drivers what lane to be in depending if they want to go on Detroit Ave or Berdan Ave.  Track the white Lincoln Towncar through the roundabout.  The white Lincoln correctly follows the signage and continues onto Berdan Ave (although they have to cross a solid white line inside the roundabout to get there).  Now track the red Camero who was at one point directly beside the white Lincoln.  Both the red Camero and the vehicle behind them incorrectly continues onto Detroit Avenue.  This is your problem.. drivers in the left lane are exiting onto Detroit Ave and they shouldn't be. 
 

[tradephoric]

What if you just removed the confusing markings? Install overhead signage that clearly shows the straight movement as being a "slight right then left". Or even change the arrows so that they point slightly left instead of straight.

Most non-signalized 2x2 or 2x1 roundabouts in the UK don't have lane markings, for what it's worth.

Most NJ traffic circles don't have lane markings either.  People do have a tendency to creep into other lanes when they're not supposed to.  Plenty of horn honking.  And yet, they don't seem to have the issues many others report at these fully parked roundabouts.

I'm not convinced eliminating lane markings would make roundabouts any safer.  There was a notorious traffic circle in Augusta, Maine known as Cony Circle which averaged over 100 crashes per year.  The traffic circle had no pavement markings and it was basically a free-for-all for drivers.  It was the highest crash prone intersection in the entire state of Maine. 

They decided they had to do something about it and reconfigured the traffic circle with modern roundabout standards in 2008.  They striped the circle as a spiral roundabout and reconfigured the entry flares slightly, but the biggest change was the pavement markings themselves.  After the change, total accidents dropped by 50% and injury accidents dropped by 68%.  Even with these impressive numbers, the newly configured roundabout is still the 3rd highest crash location in Maine.   The crash problem didn't go away, but they definitely made it better with the striping.

http://www.centralmaine.com/2011/08/11/cony-circle-accidents-down_2011-08-10/
http://www.wgme.com/images/DANGEROUS%20ROADWAY%20INTERSECTIONS.pdf

Cony Circle:  Pre-2008


Cony Circle:  Post-2008


EDIT:  To be fair, i may be underestimating the effects the geometric changes had on Cony Circle in reducing the total crashes.  I'd be all for taking a modern roundabout that has averaged 100 crashes a year and eliminate the signage and pavement markings just to see what happens (don't do any other changes, just remove all the markings inside the roundabout).  At the very least it would be fun to see the results of that study.

[tradephoric]

According to PUBLICATION NO. FHWA-SA-15-072, there have been 46 fatal roundabout crashes between 2005 and 2013.  Here's a KMZ file that maps every fatal roundabout crash:

http://www.mediafire.com/download/adpyqy0yh6k3h1e/Fatal_Roundabout_crashes_%282005-2013%29.kmz

Two roundabouts had multiple fatal crashes during the 9 year analysis period:
Moore Chapel Rd./I-485 NB Ramps in Charlotte, NC: 2 fatal crashes
US 62/Big Tree Rd. in Hamburg, NY: 2 fatal crashes

Here are some statistics regarding the fatal roundabout crashes:
*83% involved single-vehicle crashes.
*52% involved impaired driving. 
*56% were classified as speed involved crashes (at least two fatal crashes involved vehicles exceeding 100 mph). 
*46% involved motorcycles.

[cjw2001]

Those stats sounds reasonable -- you most likely have to be driving impaired, speeding, or doing something else really stupid to get killed in a roundabout.  The high percentage of single vehicle crashes speaks volumes.   Good find on the stats.

[kalvado]

Those stats sounds reasonable -- you most likely have to be driving impaired, speeding, or doing something else really stupid to get killed in a roundabout.  The high percentage of single vehicle crashes speaks volumes.   Good find on the stats.

Well, lets compare to general statistics
US average fatal crashes: 56% single vehicle, 32% impaired drivers.
DC and RI (I would use those as very urban references) have higher single-vehicle rates - 70 and 77% respectively. Drunk - 47% and 33%.

I would say numbers are not way off national average, considering roundabouts are mostly on urban roads.
I don't know what excessive number of motorcycles mean (13% national, 9 and 14 DC/RI) likely  visibility problems on roundabouts?

[tradephoric]

^The study mentions the disproportionately high number of motorcycle fatalities at roundabouts.  All it takes is a motorcyclist to clip the curbing in the roundabout and the rider goes flying. 

[Tarkus]

^The study mentions the disproportionately high number of motorcycle fatalities at roundabouts.  All it takes is a motorcyclist to clip the curbing in the roundabout and the rider goes flying.

Indeed.  I've studied the accident statistics for a bunch of roundabouts in Oregon and have found that even low-volume roundabouts, involving roads with 25mph speed limits in largely residential areas, have had high-severity injury accidents.  In most cases, it's been a motorcyclist.  Having driven through a couple of roundabouts on my 49cc Honda Metropolitan scooter (including this one in Eugene), they're really quite awkward on that mode of transportation.  Lean too far or not far enough to make all those corners, and there's a ton of curbing all over the place with your name on it.

[jeffandnicole]

I can also see motorcyclists taking these roundabouts way too fast, as sort of their own obstacle course.  A motorcyclist dropping their bike at slow speeds generally isn't going to be fatal (helmets certainly do help though!).  And in terms of cornering, a roundabout and a general left or right turn has approximately the same radius.