# Uni-directional U-Turn Prevention - H Intersection
**URL:** https://www.waze.com/discuss/t/uni-directional-u-turn-prevention-h-intersection/155130
**Category:** United States
**Created:** [October 13, 2017, 9:28pm UTC](https://www.waze.com/discuss/t/uni-directional-u-turn-prevention-h-intersection/155130 "2017-10-13T21:28:39Z")
**Posts on this page:** 20
**Page:** 1
### Author:  [@nzahn1](https://www.waze.com/discuss/u/nzahn1)
#### Post date: [October 13, 2017, 9:28pm UTC](https://www.waze.com/discuss/t/uni-directional-u-turn-prevention-h-intersection/155130/1 "2017-10-13T21:28:39Z")
There are some instances when a divided highway intersects with a cross street and u-turns are disabled in one direction only.
Until now, [it was prescribed](https://wazeopedia.waze.com/wiki/USA/Junction_Style_Guide/Intersections#Allowing_median_U-turns) to break the u-turn prevention mechanism in a single direction by adding a micro-dogleg using a geometry node.
With the advent of reliable Junction Boxes, Waze staff is now recommending discontinuing the practice of configuring intersections using precise geometry to ‘hack’ the U-turn Prevention Mechanism. Instead, please contact a rank 5+ editor to install a junction box where a unidirectional u-turn prohibition exists, disabling the prohibited route. See the image above for an example. If a R5+ editor is not immediately available in real-time chat, submit an unlock request detailing the needed changes.
NOTE: There is no need to hunt these situations down for updates, only for future instances or when the intersections are updated in the course of other work.
---
### Author:  [@ldriveskier](https://www.waze.com/discuss/u/ldriveskier)
#### Post date: [October 13, 2017, 10:54pm UTC](https://www.waze.com/discuss/t/uni-directional-u-turn-prevention-h-intersection/155130/2 "2017-10-13T22:54:04Z")
Understood, but I have a couple of concerns:
1. JBs are currently limited to R5+. Besides the fact that we will need an R5+ editor to add each one, this will also effectively lock this type of intersection to R5. A divided road with a unidirectional U-turn could currently be locked as low as L2 depending on the road type and local lock rules.
2. Every time any work needs to be done at an intersection with a JB like this on any segment touched by the JB, an R5+ will need to remove the JB for other editors to do work, then recreate the JB.
---
### Author:  [@nzahn1](https://www.waze.com/discuss/u/nzahn1)
#### Post date: [October 13, 2017, 11:00pm UTC](https://www.waze.com/discuss/t/uni-directional-u-turn-prevention-h-intersection/155130/3 "2017-10-13T23:00:05Z")
> [@ldriveskier](#):
>
> Understood, but I have a couple of concerns:
>
> 1. JBs are currently limited to R5+. Besides the fact that we will need an R5+ editor to add each one, this will also effectively lock this type of intersection to R5. A divided road with a unidirectional U-turn could currently be locked as low as L2 depending on the road type and local lock rules.
>
> 2. Every time any work needs to be done at an intersection with a JB like this on any segment touched by the JB, an R5+ will need to remove the JB for other editors to do work, then recreate the JB.
Agreed this is a definite limitation of the current JB restrictions. If we gather enough feedback, and can justify the change, it may be possible to lower the restriction to lower ranks. The L5 restriction is an artificial of when JB funcarionality and limitations were less understood; now that JB are becoming more mature, it may be possible to adjust that limit.
---
### Author:  [@juliansean](https://www.waze.com/discuss/u/juliansean)
#### Post date: [October 13, 2017, 11:03pm UTC](https://www.waze.com/discuss/t/uni-directional-u-turn-prevention-h-intersection/155130/4 "2017-10-13T23:03:31Z")
To clarify if the u-turn needs to be prevented in both directions we stick with our current practice but if allowed in one direction but not the other then use the JB? Thanks.
Sent from my iPhone using Tapatalk
---
### Author:  [@dBsooner](https://www.waze.com/discuss/u/dBsooner)
#### Post date: [October 13, 2017, 11:18pm UTC](https://www.waze.com/discuss/t/uni-directional-u-turn-prevention-h-intersection/155130/5 "2017-10-13T23:18:15Z")
> [@nzahn1](#):
>
> > [@ldriveskier](#):
> >
> > Understood, but I have a couple of concerns:
> >
> > 1. JBs are currently limited to R5+. Besides the fact that we will need an R5+ editor to add each one, this will also effectively lock this type of intersection to R5. A divided road with a unidirectional U-turn could currently be locked as low as L2 depending on the road type and local lock rules.
> >
> > 2. Every time any work needs to be done at an intersection with a JB like this on any segment touched by the JB, an R5+ will need to remove the JB for other editors to do work, then recreate the JB.
>
> Agreed this is a definite limitation of the current JB restrictions. If we gather enough feedback, and can justify the change, it may be possible to lower the restriction to lower ranks. The L5 restriction is an artificial of when JB funcarionality and limitations were less understood; now that JB are becoming more mature, it may be possible to adjust that limit.
To be honest, in the situation of having mDL’s to prevent the u-turn on segments locked as low as L2 is dangerous enough. As an R2 Editor, they may not be paying attention to, or not even know what the mDL is for and inadvertently remove it. I know a lot of people put MC’s by them, but still, it’s dangerous.
As for JB being lowered, I’d be for that. But before they are lowered, I would like to see the WoP documentation spruced up for them. Right now, the page does exist, but can be confusing to some. As was the situation when Route Preference was added. We still have very little, if not any, documentation on route preferences. I’d hate to see JB released to the lower ranks (me being one too) without a more clear explanation of what they do, how to use them, when to use them, etc. Yes a page does exist, but few lower ranks know exactly how to use them or what the pitfalls are. Especially if they set them up incorrectly. I understand JB and love them. But we should make sure all editors that have the access understand them as well.
I also know JB 2.0 is coming out… so that could change the game yet again.
Sent from my iPhone using Tapatalk
---
### Author:  [@anon33958190](https://www.waze.com/discuss/u/anon33958190)
#### Post date: [October 14, 2017, 7:43am UTC](https://www.waze.com/discuss/t/uni-directional-u-turn-prevention-h-intersection/155130/6 "2017-10-14T07:43:38Z")
There are a fair number of these types of intersections in my area, none of which are locked higher than 3. I would support lowering JBs to 4, for all the reasons that Lisa and DB said. Isn’t the JB rank lock based on country? I saw a post where it was 2 in parts of Europe, but not sure if that’s intentional
Sent from my LG-H811 using Tapatalk
---
### Author:  [@dBsooner](https://www.waze.com/discuss/u/dBsooner)
#### Post date: [October 14, 2017, 11:49pm UTC](https://www.waze.com/discuss/t/uni-directional-u-turn-prevention-h-intersection/155130/7 "2017-10-14T23:49:51Z")
> [@juliansean](#):
>
> To clarify if the u-turn needs to be prevented in both directions we stick with our current practice but if allowed in one direction but not the other then use the JB? Thanks.
>
> Sent from my iPhone using Tapatalk
This brings up a good point… if they want us to steer clear of mDL’s, why is it not guidance to use a JB when both u-turns are allowed?
Sent from my iPhone using Tapatalk
---
### Author:  [@nzahn1](https://www.waze.com/discuss/u/nzahn1)
#### Post date: [October 14, 2017, 11:55pm UTC](https://www.waze.com/discuss/t/uni-directional-u-turn-prevention-h-intersection/155130/8 "2017-10-14T23:55:12Z")
> [@dBsooner](#):
>
> > [@juliansean](#):
> >
> > To clarify if the u-turn needs to be prevented in both directions we stick with our current practice but if allowed in one direction but not the other then use the JB? Thanks.
> >
> > Sent from my iPhone using Tapatalk
>
> This brings up a good point… if they want us to steer clear of mDL’s, why is it not guidance to use a JB when both u-turns are restricted?
>
> Sent from my iPhone using Tapatalk
Maybe it’s my personal experience blinders on. When I have two restricted u-turns, I shorten the cross segment to prevent both turns, rather than mDLing the angles.
---
### Author:  [@anon33958190](https://www.waze.com/discuss/u/anon33958190)
#### Post date: [October 14, 2017, 11:56pm UTC](https://www.waze.com/discuss/t/uni-directional-u-turn-prevention-h-intersection/155130/9 "2017-10-14T23:56:40Z")
Not sure how мDLs could be used to prevent a U-turn anyway. They just break the penalty
Sent from my LG-H811 using Tapatalk
Edit: that said, the guidance should probably say to just not use мDLs for any U-turns and just use a \>15 meter median if allowing both U-turns
---
### Author:  [@dBsooner](https://www.waze.com/discuss/u/dBsooner)
#### Post date: [October 15, 2017, 12:09am UTC](https://www.waze.com/discuss/t/uni-directional-u-turn-prevention-h-intersection/155130/10 "2017-10-15T00:09:43Z")
Yeah… that was a noob question. I completely spaced on that forgetting the 3 things that have to be present for the double left penalty to be in place. So as long as the median is =\<14m, the penalty exists so neither u turn (double left) is allowed. Making the segment \>14m breaks the penalty, thus allows the turn in both. mDL’s would be wasteful if you’re wanting to break it both ways. mDL’s were used to allow it in only one direction, by breaking the penalty. Sorry for the confusion. Carry on.
Sent from my iPhone using Tapatalk
---
### Author:  [@bz2012](https://www.waze.com/discuss/u/bz2012)
#### Post date: [October 15, 2017, 4:03pm UTC](https://www.waze.com/discuss/t/uni-directional-u-turn-prevention-h-intersection/155130/11 "2017-10-15T16:03:49Z")
recent discussion on discord Waze USA general:
> [@](#):
>
> [9:14 AM] bz2012 (5) SM La: Comment on the Waze change “encouraging” using more JBs. JBs have a MAJOR flaw: The presence of a JB anywhere on a particular segment “LOCKS” that segment, preventing many actions. That ‘LOCK’ can only be removed by removing the JB.
> For example : I was attempting to connect a PLR to a segment
> [9:16 AM] bz2012 (5) SM La: There happens to be a JB at the other end of the segment I attempted connect to
> [9:16 AM] bz2012 (5) SM La: This is a major pain! Especially as the more widespread use of JBs is being encouraged. That ‘LOCKING’ needs to be removed or a tool provided that will allow us to ‘suspend’ the JB while we do things and then re-impose it without losing the data it has been collecting and the configuration it has.(edited)
> [9:20 AM] johnsninja58 (5) VT SM: will that also destroy any of the JB data it had saved?
> [9:21 AM] enembee (5): Seems like this is more a comment on JBs as a whole, rather than recent guidance. Since, in reality, there likely aren’t a ton of intersections that require this uni-directional u-turn prevention technique.
> [9:24 AM] dBsooner (4) AM OK: It’s the JB as a whole, yes. The lock is there because if you modify any of the segments of the JB, you modify the endpoints of the JB.
> [9:24 AM] dBsooner (4) AM OK: I’m hoping JB 2.0 finds a way around this.
> [9:25 AM] enembee (5): And that was a concern I had, @johnsninja58 (5) VT SM.
> [9:25 AM] dBsooner (4) AM OK: @johnsninja58 (5) VT SM yep. It will destroy it if you delete it.
> [9:26 AM] dBsooner (4) AM OK: I thought you could modify parts of the JB (like name of a segment, etc) from beta WME, without removing the JB. I wonder if you can add a segment? I’m no longer a beta WME Editor or R5, so can’t test.(edited)
> [9:27 AM] bz2012 (5) SM La: Let me try working around in WME beta and see what happens. No joy so far. As far as I can tell, installing a JB anywhere for any reason essentially “LOCKS” all segments passing into that JB, preventing many normal operations on those segments and putting a ‘high price (data loss and time lost to replicate the configuration)’ on temporary removal of the JB in order to make desirable/necessary or even vital changes to those segments.(edited)
> [9:45 AM] johnsninja58 (5) VT SM: We may also be over thinking this.
> [9:45 AM] JDelosa (5): Maybe segments have to be cut at the JB allowing the parts out side the JB access
> [9:45 AM] johnsninja58 (5) VT SM: This doesn’t necessarily mean they will be used everywhere and proper configuration in advance prevents much of this
> [9:46 AM] johnsninja58 (5) VT SM: Should this be in the #routing or #wazeopedia\_usa section btw?
> [9:51 AM] bz2012 (5) SM La: JDelosa’s idea sounds like an ideal ‘cheap solution’ ! "The boarders of every JB could automatically insert a ‘cutting node’ in every segment. " That would allow normal manipulation of the external segments.
> [9:59 AM] ldriveskier (4) Lisa: Discussion probably should be in #routing , but wouldn’t cutting the segments defeat the traffic collection data that is done by the JB for the segments?
> [10:02 AM] bz2012 (5) SM La: I wouldn’t think so, I think when a segment is cut, the (speed) data replicates to both segments . That is one of the reasons behind the procedure we follow to divide or undivide a roadway. (data preservation!) I think that when segments are combined, the speed is averaged (hopefull on a basis of the relative lengths).(edited)
> [10:02 AM] dfortney (5): I’d super hate cutting segments entering a junction box when the only purpose of that junction box is to eliminate a micro-dogleg for a U-turn. I’d rather stick with the one simple hack vs. multiple road cuts
> [10:03 AM] bz2012 (5) SM La: This would be an automatic cut. Not one you would see or worry about normally.
> [10:05 AM] doctorblah (4): a la roundabout?
> [10:09 AM] bz2012 (5) SM La: except those are physically visible and apparent as soon as the roundabout is drawn.
> [10:12 AM] ldriveskier (4) Lisa: But after it is cut, the speed data will be collected separately on the two now different segments. Sometimes JBs are used to fix places where speed data gets messed up easily, so I would think this would be a bad thing.
> [10:13 AM] bz2012 (5) SM La: The segments would be in a series circuit. Identical current flows in a series leg.
> [10:18 AM] dfortney (5): I don’t know what that means. Maybe it means we’ll actually get a “U-turn” instruction on divided road U-turns 😉
> [10:19 AM] dude463 (5): create one problem while fixing another
> [10:25 AM] bz2012 (5) SM La: It means that if you have a segment that is 1 mile long and cars take 1 min to travel it, and you divide it into 2 half mile segments, the data for each segment will still show the same number of cars and that they are traveling at 60 mph and take 1/2 minute to travel the 1/2 mile. The same no matter where the division takes place.(edited)
> [10:51 AM] JDelosa (5): I do believe we should move this discussion to #routing as it will get lost here accept by us.
> [10:55 AM] bz2012 (5) SM La: It isn’t really ‘routing’, more like general editing, however, I agree it can effect routing. Many things do.
> [10:56 AM] JDelosa (5): Well maybe the #wazeopedia\_usa Channel then but here it will probably get lost
> [10:58 AM] johnsninja58 (5) VT SM: We could always bring it to the forums?
---
### Author:  [@ldriveskier](https://www.waze.com/discuss/u/ldriveskier)
#### Post date: [October 16, 2017, 4:23pm UTC](https://www.waze.com/discuss/t/uni-directional-u-turn-prevention-h-intersection/155130/12 "2017-10-16T16:23:44Z")
> [@ldriveskier](#):
>
> Understood, but I have a couple of concerns:
>
> 1. JBs are currently limited to R5+. Besides the fact that we will need an R5+ editor to add each one, this will also effectively lock this type of intersection to R5. A divided road with a unidirectional U-turn could currently be locked as low as L2 depending on the road type and local lock rules.
>
> 2. Every time any work needs to be done at an intersection with a JB like this on any segment touched by the JB, an R5+ will need to remove the JB for other editors to do work, then recreate the JB.
As a follow-up, one of my unvoiced concerns was the inability to add an RTC to segments that are otherwise locked by the addition of a JB. After further testing, RTCs can be added to such segments as long as the editor meets the minimum rank required for the segment placing an RTC.
While I still have a small concern about JBs being locked to L5, after further conversations I have less concern that I did when I left my first comment. I would support making JBs available to R4 editors in the USA, especially since there are many SMs that are at R4. Alternately, if it can be set by role (this may be too complicated), require SM+ or any R5+ (since not all R5’s are SM+) be able to edit JBs. Despite my poor choice of example in my previous comment, I really wasn’t suggesting the lock be lowered to L2 and this may be a good way to prevent inadvertent damage by lower ranked editors at these more delicate intersections.
---
### Author:  [@TheChrisK](https://www.waze.com/discuss/u/TheChrisK)
#### Post date: [October 16, 2017, 7:17pm UTC](https://www.waze.com/discuss/t/uni-directional-u-turn-prevention-h-intersection/155130/13 "2017-10-16T19:17:30Z")
I agree that R4 would be an acceptable lock level for JBs now that they’ve had time to simmer in the map and more people understand what their purpose is. It is a bit of a task to ask an R5 to remove a JB (while keeping a note of it’s allowed routes), allowing an edit to be made and re-drawing. This can be somewhat complicated for complex intersections.
It would be nice if there was a “disable” option for JBs to temporarily unlock segments for editing and then re-enable it with the routes restored as they we’re (aside from major segment changes where a re-draw is needed anyways).
---
### Author:  [@juliansean](https://www.waze.com/discuss/u/juliansean)
#### Post date: [October 16, 2017, 10:19pm UTC](https://www.waze.com/discuss/t/uni-directional-u-turn-prevention-h-intersection/155130/14 "2017-10-16T22:19:41Z")
The RTC is a big issue. I guess our R5’s are going to have to get creative with closures.
Sent from my iPhone using Tapatalk
---
### Author:  [@subs5](https://www.waze.com/discuss/u/subs5)
#### Post date: [October 17, 2017, 10:16am UTC](https://www.waze.com/discuss/t/uni-directional-u-turn-prevention-h-intersection/155130/15 "2017-10-17T10:16:53Z")
You will still have it so that two mechanisms have to be looked for to evaluate a an intersection. The length of the cross segment (\<14 m even if you have true length script) and angles (180 +/- 5) to see if the u-turn is prevented. Then you have to look for JB which is rank locked to R5 as stated above. Not sure how this fits into the simplicity portion of editing motto. Since it doesn’t even take out the checking angle measurements.
If we are to remove u-turn prevention entirely and go with JB for either preventing both or only one then fine. Yes that would be a lot of work but less so than Speed Limits or parking lot areas, two other recent projects still ongoing.
---
### Author:  [@ldriveskier](https://www.waze.com/discuss/u/ldriveskier)
#### Post date: [October 17, 2017, 10:23am UTC](https://www.waze.com/discuss/t/uni-directional-u-turn-prevention-h-intersection/155130/16 "2017-10-17T10:23:20Z")
> [@juliansean](#):
>
> The RTC is a big issue. I guess our R5’s are going to have to get creative with closures.
Maybe my post wasn’t clear enough. RTCs are NOT an issue. I was able to set an RTC on a segment that was included in a JB without disturbing the JB or needing help - and I’m not R5.
---
### Author:  [@juliansean](https://www.waze.com/discuss/u/juliansean)
#### Post date: [October 17, 2017, 11:58am UTC](https://www.waze.com/discuss/t/uni-directional-u-turn-prevention-h-intersection/155130/17 "2017-10-17T11:58:05Z")
> [@ldriveskier](#):
>
> > [@juliansean](#):
> >
> > The RTC is a big issue. I guess our R5’s are going to have to get creative with closures.
>
> Maybe my post wasn’t clear enough. RTCs are NOT an issue. I was able to set an RTC on a segment that was included in a JB without disturbing the JB or needing help - and I’m not R5.
That’s great to hear! I hadn’t tried yet and did think that’s what you meant! Thanks for clarifying.
Sent from my iPhone using Tapatalk
---
### Author:  [@anon33958190](https://www.waze.com/discuss/u/anon33958190)
#### Post date: [October 17, 2017, 12:33pm UTC](https://www.waze.com/discuss/t/uni-directional-u-turn-prevention-h-intersection/155130/18 "2017-10-17T12:33:29Z")
> [@subs5](#):
>
> You will still have it so that two mechanisms have to be looked for to evaluate a an intersection. The length of the cross segment (\<14 m even if you have true length script) and angles (180 +/- 5) to see if the u-turn is prevented. Then you have to look for JB which is rank locked to R5 as stated above. Not sure how this fits into the simplicity portion of editing motto. Since it doesn’t even take out the checking angle measurements.
Junction boxes remove the median u-turn penalty mechanisms, so you don’t have to look at or change any segment lengths or angles. Just map the intersection naturally, slap a JB on it and restrict what needs restricted. In cases where U-turns are illegal over a wide median or legal over a narrow median, it would also seem more sensible to use a JB than make the median segments some weird length.
---
### Author:  [@voludu21](https://www.waze.com/discuss/u/voludu21)
#### Post date: [October 17, 2017, 12:39pm UTC](https://www.waze.com/discuss/t/uni-directional-u-turn-prevention-h-intersection/155130/19 "2017-10-17T12:39:08Z")
Even if we can pesuade waze team to lower the minimum editor rank for JB creation, there is still a problem with the proliferation of JB:
We still have to remove an entire JB and all the turn restrictions set in that JB, in order to make any changes to the intersection.
This is a potential source of error, – recreating the JB after it has been deleted and some small adjustment has been made to the intersection.
It’s going to be a problem.
---
### Author:  [@nzahn1](https://www.waze.com/discuss/u/nzahn1)
#### Post date: [October 17, 2017, 1:48pm UTC](https://www.waze.com/discuss/t/uni-directional-u-turn-prevention-h-intersection/155130/20 "2017-10-17T13:48:51Z")
I want to share some of the discussion that occurred in discord.
Consensus facts:
- JBs are an unambiguous tool for restricting illegal routes across multiple segments, and staff has conveyed a preference that JBs used for restricting unidirectional U-turns on divided highways, over using mDL to ‘hack’ the U-turn Prevention Mechanism.
- JBs do not prevent the addition/editing/removal of real-time closures, as long as the user meets the minimum rank required for the road segment and RTC tool.
- Changes to JBs are saved to the tile; so, when adding/editing/removing a JB, changes can be compared to the existing tile, as long as the user meets the minimum rank required for the ‘Snapshot View’ tool.
- Removing and readding a JB does not delete the accumulated data.
- (Though, there is some question whether if the data is associated with specific segIDs, and if these segIDs change what happens to the data. Therefore, see #1 below).
Consensus opinions/findings
- Whenever a JB is required (for U-turn prevention or otherwise), make sure that everything else at the intersection is perfect first, including AGCs and PLR connections. A well wme-built intersection shouldn’t need to be revisited for years, unless a major construction project is undertaken, in which case you’d want to remove/rework a JB anyway.
- The counterpoint to the status quo: If you have a unidirectional U-turn setup in the old mDL style, you’d want to lock the intersection at/above 3 anyway, due to the possibility that a well-meaning editor could break the U-turn prevention mechanism installed.
- Though the process of adding and editing JBs is complex, the USA community feels that it may be appropriate to lower the feature lock on JBs to rank 4 editors. [i]
- (Note: I believe that the JB feature lock is currently global, so moving it may require staff to make a more substantial change.)
- (Note also: [JBs are currently undergoing an overhaul (beta editors only)](https://www.waze.com/forum/viewtopic.php?f=211&t=225660), and so a wiki update and editor training may be prudent before lowering the JB feature lock once JB2.0 is implemented.)[/i]
I’m proposing changes to the ‘Allowing median U-turns’ section of the [Controling U-turn penalties](https://wazeopedia.waze.com/wiki/USA/Routing_penalties/Controlling_U-turn_penalties) wiki. The only change is to remove the instructions for adding mDLs for unidirectional U-turns, and replaced with an instruction to install a JB in this specific situation. [See the new draft section on my wiki userspace](https://wazeopedia.waze.com/wiki/USA/index.php?title=User:Nzahn1/unidirectional_U-turn); the remainder of the page is unchanged.
[Next page](https://www.waze.com/discuss/t/uni-directional-u-turn-prevention-h-intersection/155130.md?page=2)