The Worst of Road Signs

[codyg1985]

There's a sign on some side street in Somerville or Chelsea or Lynn or some other nearby city north of Boston that I always see out the window of commuter trains that's a Speed Limit 11 I believe. Looks fully standard.

And back in Virginia, the entrance and exit roads to Busch Gardens are signed at 18 mph.

And in Nashville, the speed limit on the road that goes around the Opryland Hotel/Convention Center is 24 mph.  I can't remember whether or not that also applies to the road around Opry Mills as well.

It does. I remember seeing those signs around Opry Mills as well. Weird.

[J N Winkler]

There are other examples of the Australian-style diagrammatics on a recently improved length of SR 89A in Cottonwood, just northwest of the SR 260 junction.  Unfortunately StreetView in this area is currently useless since it shows the roundabouts half-built, with no signing work done.  The intersections converted to roundabouts include Avenida Centerville, Valley View Rd., Cement Plant Rd., Clarkdale Pkwy, Lisa St./Lincoln Dr. and the work was done under TRACS H412901C.

You mean these?

(Sign pictures snipped for brevity)

Yes--those exactly.  Many thanks for digging these up.

As an aside, I checked and it seems I was wrong.  The Queensland MUTCD does not, in fact, mention the multilane diagrammatics Arizona DOT has been trying.  Instead, they are mentioned briefly in a separate Main Roads Queensland publication, the Traffic and Road Use Management Manual (section 1.27):

http://www.tmr.qld.gov.au/~/media/92ae4c3e-f3c4-493c-be14-b86cee533fdb/trumvolume1127.pdf

It seems that Queensland uses the same 90° chamfer for stub arms that the British do and which I personally recommend.  I don't know where Arizona DOT is getting their 60° (or whatever) chamfer.

For roundabouts in general the Queensland MUTCD recommends plain-vanilla Chapter 7 diagrammatics similar to what Arizona DOT tried on Loop 202.  The TRUM Manual describes a number of traffic engineering techniques used on the Main Roads network which aren't referenced directly in the MUTCD.

The TRUM Manual section referenced above refers the reader to section 14.3.4 of the Road Planning and Design Manual, which sheds some light on the motivation for these roundabout diagrammatics.  Apparently they are used in Queensland because the authorities want there to be a clear lane assignment for every possible trajectory through the roundabout, so that if an accident occurs and one of the drivers involved is found to be in the wrong lane for his or her intended path, fault can be assigned without ambiguity.

I think this is wrongheaded because the diagrammatics contain too much information (especially for complex intersections) to be processed effectively at speed.  Instead drivers should be taught simple rules for lane selection and the general ethical principle of defensive driving, and otherwise be left alone, as in Britain.

[vtk]

Apparently they are used in Queensland because the authorities want there to be a clear lane assignment for every possible trajectory through the roundabout, so that if an accident occurs and one of the drivers involved is found to be in the wrong lane for his or her intended path, fault can be assigned without ambiguity.

I think this is wrongheaded because the diagrammatics contain too much information (especially for complex intersections) to be processed effectively at speed.  Instead drivers should be taught simple rules for lane selection and the general ethical principle of defensive driving, and otherwise be left alone, as in Britain.

Roundabouts, like regular intersections, can vary considerably from example to example, including lane assignment.  There are no "simple rules for lane selection" that can be universally applied to all roundabouts without requiring the knowledge of specific roundabouts' lane configurations.  MUCTC chapter 3C offers several examples of different roundabout configurations; some of them have multiple and/or mandatory turn lanes, which must be considered in choosing a lane.

Of course, this information can be conveyed without diagrammatic signage.  The lane control signs described in chapter 2B are sufficient to choose a lane (and assign fault) for just about any roundabout with 3 or 4 entries/exits.  On the other hand, those don't convey route or destination guidance, which is probably ADOT's reason for going with the diagrammatics in the first place.

[J N Winkler]

Roundabouts, like regular intersections, can vary considerably from example to example, including lane assignment.  There are no "simple rules for lane selection" that can be universally applied to all roundabouts without requiring the knowledge of specific roundabouts' lane configurations.

Oh, there are simple rules--they are included in the Highway Code in Britain.  Left-hand lane if you are turning left, either lane if you are going straight ahead, right-hand lane if you are turning right, etc.  It is true that these rules are difficult to apply at some types of multilane roundabout, particularly ones which are not sited at a right-angle intersection of two roads or have mandatory lane assignments, but this is why positive guidance is typically provided with lane-assignment direction signs and markings on the pavement (which, in Britain, can include road numbers and destinations).

The underlying point nevertheless remains:  if you know the basic rules for navigating roundabouts, you know what to expect from the lane assignment signing when and where that is provided.  When this information is spread out in space and time, as is basically the case with the British and plain-vanilla MUTCD approaches, this adds to driver confidence since it eliminates the need (real or perceived) to process a complex diagrammatic (as shown in the Queensland manual) within a short period of time.

Of course, this information can be conveyed without diagrammatic signage.  The lane control signs described in chapter 2B are sufficient to choose a lane (and assign fault) for just about any roundabout with 3 or 4 entries/exits.  On the other hand, those don't convey route or destination guidance, which is probably ADOT's reason for going with the diagrammatics in the first place.

I disagree.  First, it is myopic to choose a particular design of sign for purposes of assigning fault when you can resort to a better, more robust, and more thoroughly tested approach to avoid the need to assign fault in the first place.  Second, a basic principle of traffic design is not to force drivers to process large amounts of information within a small window of space and time.  The Australian approach presents drivers with routes, destinations, and lane assignment on one sign, while the British and vanilla MUTCD approaches break up the lane assignment and route/destination information across multiple panels which are processed separately.

In context the Australian approach "works" in Arizona largely because redundant lane assignment signing is provided, which allows drivers to ignore lane assignment on the diagrammatic, observe the general tendency of the stub arms, and then look to downstream signing and pavement markings for lane selection.  So why even attempt to do lane assignment on the diagrammatic itself?

Britain has for decades had thousands of roundabouts signed using diagrammatic advance direction signs.  Many of these have five or more entries and exits, including some arms which are entry-only, as well as slip lanes allowing certain streams of traffic to avoid the circulatory carriageway.  There is a sophisticated visual grammar for presenting the road layout on a diagrammatic advance direction sign which allows drivers to interpret it quickly and accurately, without needing to be fed lane assignment information on the same sign.  I don't think the Australian approach is the proverbial better mousetrap.

Edit:  Looking at Corco's pictures of the SR 89A diagrammatics closely, it seems we are confronted with one instance of a single-lane roundabout with single-lane approach and two instances of double-lane roundabouts with double-lane approaches, all roundabouts being situated at right-angle intersections.  In the case of the two multilane roundabouts, the lane positioning rules taught in Britain tell the driver everything he or she needs to know in order to choose the correct lane.  In Britain these roundabouts would normally be signed with a diagrammatic only, without any lane assignment signs and also, I believe, without any arrows on the pavement.

[vtk]

Looking at Corco's pictures of the SR 89A diagrammatics closely, it seems we are confronted with one instance of a single-lane roundabout with single-lane approach and two instances of double-lane roundabouts with double-lane approaches, all roundabouts being situated at right-angle intersections.  In the case of the two multilane roundabouts, the lane positioning rules taught in Britain tell the driver everything he or she needs to know in order to choose the correct lane.  In Britain these roundabouts would normally be signed with a diagrammatic only, without any lane assignment signs and also, I believe, without any arrows on the pavement.

None of those Arizona roundabouts, except for the first one whose diagram looks like a chicken, have mandatory turn lanes.  The "lane positioning rules taught in Britain" work fine, as long as there aren't mandatory turn lanes or other deviations from typical roundabout design.  My point was that, for roundabouts with such deviations, additional info is necessary besides "lane positioning rules taught in Britain".  I didn't attempt to say that one method of conveying that information is better than another. 

[J N Winkler]

The "lane positioning rules taught in Britain" work fine, as long as there aren't mandatory turn lanes or other deviations from typical roundabout design.

Actually, they do work quite well, even when these deviations exist and the basic rules are overriden by lane assignment signing.  The underlying issue is one of socialization and driver expectancy.  If you know the rules, you can anticipate with reasonable certainty which lane you will be assigned and you can be prepared to make any necessary lane changes well in advance.

My point was that, for roundabouts with such deviations, additional info is necessary besides "lane positioning rules taught in Britain".

I don't think there is any disagreement that such information should be, and is, provided, or that it overrides the basic rules.  I also take your point that the SR 89A roundabouts don't have mandatory turn lanes or other unusual elements (and that, in fact, is why I argue the lane assignment signing is redundant).

In relation to the Wickenburg roundabout, clearly explicit lane assignment signing has to be provided because there are three lanes on the approach, the left of which goes to US 60 and Los Angeles, the right of which goes to Las Vegas and US 93, and the center lane of which is an option lane for either destination.  Note that if no lane assignment signing were provided, drivers following the British rules would know for sure that the left lane would go to US 60/Los Angeles while the right lane would go to US 93/Las Vegas.  The main benefits of lane assignment signing at this junction are to ensure adequate usage of the option lane and to prevent unnecessary lane changes on approach (to get into the lane "guaranteed" to go in one direction or the other) and on exit (to get into correct lane position for an upcoming turn).

At comparable junctions in Britain, provision of lane assignment signing can be very basic or very elaborate.  Basic provision would be a roundabout diagrammatic with pavement markings (no vertical signs) to assign lanes.  Intermediate provision might include an upright sign with lane assignment information, positioned downstream of the roundabout diagrammatic.  Finally, the lane assignment information can be put on overhead signs, also downstream of the diagrammatic.  In Britain this approach tends to be confined to motorway interchanges, though it looks like it is used at the Wickenburg roundabout and it is being used extensively for interchange roundabouts being built as part of the US 41 upgrade in Wisconsin.

Traffic volumes (including turning volumes) are important considerations in choosing among these approaches.

My concern about the Wickenburg diagrammatic is that the only bit of information it adds that would not be communicated by a vanilla Chapter 7 diagrammatic is the lane assignment, including the fact that the center lane is an option lane for US 60 and US 93.  This bit of information is lost at speed and its inclusion on the sign makes the whole difficult to read at speed.

The lane choice rules for roundabouts are not unique to Britain and I am sure they are taught in Australia (where state driving tests tend to be difficult and licenses are exchangeable with British licenses, unlike ours).  I think it is telling that the example lane-assignment diagrammatic in the TRUM Manual shows a situation which is not addressed well by lane choice rules--two lanes on approach, two lanes on first exit, and then one lane each on the following three exits, with the right-hand lane on approach (remember, traffic circulates on the left in Australia) being an option lane for all of the exits.

[bigboi00069]

Other than the white state road colored west and arrow signs, why is this black and not blue?

[Ian]

When Warren County, NY started signing their county routes last year, they put up county route shields in a similar color:


I think the one here and the Broward CR 822 shield above are more of a really dark navy blue.

[formulanone]

Other than the white state road colored west and arrow signs, why is this black and not blue?

I'm noticing Broward is finally recognizing county routes (or at least, that they aren't completely state-maintained); those pentagon shields used to be nearly impossible to find until around 2008 or so.

There seems to be a trend of dark blue pentagon shields, though. Not the best shots, but you can see the contrast between old and new:


March 2011, along CR 78


October 2011, along CR 833

[Scott5114]

You would think they would notice the colors on banner and the pentagon don't match...

[brownpelican]

Trenton, Tennessee has a speed limit of 31 all over their town.

And here's the speed limit at Ole Miss:

[bigboi00069]

I'm noticing Broward is finally recognizing county routes (or at least, that they aren't completely state-maintained); those pentagon shields used to be nearly impossible to find until around 2008 or so.

Yeah it is nice to see the CR signs in Broward. I do recall not seeing any at all a few years ago. And at least every other one i have seen in Broward is the correct colors, even other CR 822 shields.

[Ian]

You would think they would notice the colors on banner and the pentagon don't match...

You can only get so much from a county! 

[apeman33]

Trenton, Tennessee has a speed limit of 31 all over their town.

And here's the speed limit at Ole Miss:

Supposedly, it honors Archie Manning. He wore 18 as QB at Ole Miss.

[1995hoo]

Not far from my house one can find the very nice I-95 shield shown on the left in the picture below. Less than half a block later, behind the tree as indicated by the arrow, one can find the very ugly I-95 shield shown on the right. It may not be apparent from the image, but the "9" is not on a straight line with the "5" and is instead raised slightly.

[empirestate]

Not far from my house one can find the very nice I-95 shield shown on the left in the picture below. Less than half a block later, behind the tree as indicated by the arrow, one can find the very ugly I-95 shield shown on the right. It may not be apparent from the image, but the "9" is not on a straight line with the "5" and is instead raised slightly.

Not to mention that SR 613 crosses your path twice, according to the photos: once at the intersection in the first photo (hope that side street continues on the other side!), and again at the signal in the second.

[Quillz]

http://en.wikipedia.org/wiki/Virginia_State_Route_613_(Fairfax_County)

Assuming I have the state right, sounds like this particular route has multiple starting and stopping points.

[1995hoo]

Not far from my house one can find the very nice I-95 shield shown on the left in the picture below. Less than half a block later, behind the tree as indicated by the arrow, one can find the very ugly I-95 shield shown on the right. It may not be apparent from the image, but the "9" is not on a straight line with the "5" and is instead raised slightly.

Not to mention that SR 613 crosses your path twice, according to the photos: once at the intersection in the first photo (hope that side street continues on the other side!), and again at the signal in the second.

http://en.wikipedia.org/wiki/Virginia_State_Route_613_(Fairfax_County)

Assuming I have the state right, sounds like this particular route has multiple starting and stopping points.

It does. The road on which I was driving when I took the video from which those pictures are captured is Franconia Road, which is primarily numbered as 644. Van Dorn Street, which crosses Franconia at the light in front of me there, is numbered 613. So is Beulah Street, which heads south from Franconia Road perhaps half a mile behind me as I took those pictures. The part of Van Dorn to the south of Franconia Road (to the right in the pictures) is more recent construction than Beulah and so the duplicated numbering is probably an historical artifact. My Acura's sat-nav stubbornly lists the portion of Franconia Road between Beulah and Van Dorn as Route 613, whereas it lists the rest of it as Franconia Road. I'm not sure why it does that. Either way, though, I don't know anyone who ever refers to any road around here as "613." The radio traffic reporters often refer to "644" because the road to which that route number applies changes names as it crosses I-95 in Springfield (Franconia Road to the east, Old Keene Mill Road to the west) and both are important local arterials. I suppose on the radio it's easier just to say the number in that situation.

The street to the right in the first photo, Villa Street, isn't part of 613 and so the sign may be seen as misleading. Villa does not continue on the other side, although there is a church car park entrance.

[empirestate]

The street to the right in the first photo, Villa Street, isn't part of 613 and so the sign may be seen as misleading. Villa does not continue on the other side, although there is a church car park entrance.

Yeah, that's what I was getting at. According to the first photo, SR 613 is left and right from this spot right here, while I-95 is to the left at a spot up ahead. In the second photo we're at the spot up ahead, and I-95 is signed to the left as predicted, but here's 613 again, even though we've already passed the spot where the first sign said it turned.

I'm sure this isn't confusing when you're seeing it in real life, and I'm also not sure there's an official arrow for a route that goes both ways from a spot up ahead. Just struck me as an extra quirk adding to the signage's worst-ness.

[hbelkins]

You really need to understand Virginia's signing practices for secondary roads to get that one. Virginia usually uses an advance arrow prior to the turn. In cases where the route crosses, sometimes they will use the double arrow twice, and sometimes they will use duplicate route markers with the same number with an advance left and an advance right on one assembly. Too lazy to go through my photos to find one, but I have seen that done many, many times.

[SSOWorld]

I've seen some signage for Ohio SR directions that have up-arrows that have a right arrow sticking out the middle of it - like it has a boner.

[Android]

I've seen some signage for Ohio SR directions that have up-arrows that have a right arrow sticking out the middle of it - like it has a boner.

You mean like the ones in Nebraska?  I've often found those to be somewhat awkward. 




To change the focus to a different sign I find a bit awkward...  I've seen these in rural Weld County, Colorado over the years and always thought they were poorly worded.  Yesterday I bothered to take a photo of one.  These are on county roads in areas of poor drainage and the roadbed is not raised up any, so it can develop standing water on the rare times there is signifcant precipitation.  But still, shouldn't it read something like: Standing Water Possible, or Road Subject To Flooding?

[Brian556]

Argyle, TX

[SSOWorld]

where is it?

[1995hoo]

The "Caution Rising Water" sign makes more sense than what VDOT used to have posted on Woodburn Road in Annandale, Virginia. When I was a kid there was a one-lane bridge that used to flood out almost every time it rained. Eventually they put up a warning sign that was supposed to be turned sideways and then rotated 90 degrees whenever the road flooded, as it said "Road Closed High Water." But they must have gotten tired of turning it because eventually that sign was facing traffic at all times, which kind of defeated the purpose of having the warning sign at all.

I do not have a picture. Street View doesn't show the sign. I was on that road about two weeks ago but wasn't thinking of this issue and probably wouldn't have noticed the sign if I had, as I've been on that road literally thousands of times since access to the neighborhood where we lived when I was a kid required you to use Woodburn Road.