Soliciting for some feedback on un-dividing Fischer Blvd in Toms River. From my analysis it meets all the requirements to not be divided. I understand we normally do not arbitrarily make these changes, but I feel the current divided setup is not conducive to proper routing and may proliferate additional workarounds like this.
I spotted this issue from analyzing UR 5026274. For the record (once this UR is closed someday) the reporter is complaining about accessing the Siam Spice restaurant when approaching from the south. They received a route by making a left before the restaurant (probably Shore Blvd), proceed north behind the restaurant via back roads, and then approach the restaurant from the correct side of the street by traveling south.
There are a number of street segments between the divided segments, but these will not be utilized since they meet the requirements of u-turn prevention. Consequently, Waze has no concept of the middle left turn lane. This could be better facilitated by a single two-way Fischer Blvd with proper turn restrictions. I’m not really happy with proliferating some of the existing cross segments along this road that are a workaround for simulating the center turn lane, and this will only become more complicated as more destinations are identified and mapped. While u-turn prevention could be disabled at these cross streets, there’s a possibility of many more being added in a short distance, once again as more destinations are mapped.
Funny, I started looking at this today and thinking the same thing. I haven’t dug in to as much detail yet as you probably have, but from what I could see, the majority, if not all, destinations can be controlled properly using turn restrictions. Most destinations on the opposite side can be reached via left turns, and to your point, I would agree that it being divided is causing more errors than preventing. I planned on looking in to this more this week but definitely support your argument right now. If you get the green light and need any assistance, please let me know.
Hi, just jumping in here… I’m not sure how UTP would apply in this situation. Since you are coming up/down Fischer Blvd, and making a left turn only into a Parking Lot, there is no reason for Waze to even consider blocking a potential U-Turn. These U-turn’s could be further prevented from even showing up by adding red arrows on the median segments.
:idea: The wazer’s destination in this UR Siam Spice also does not have a mapped parking lot, this could be a quick fix solution for the Wazer here.
:arrow: I would agree that this road doesn’t meet the requirements to be divided, and it specifically violates the rule about center turn lanes, but I only see the one UR here. Perhaps it would not be worth the lost data to attempt to undivide the roadway.
Regarding UTP, I’m merely saying that a NB Wazer is not able to access a SB segment of Fischer Blvd without being given an odd route (the complaint of this UR). Given the lack of PLR presently, a SB Fischer Blvd segment would presumably be the stop point for this, and other destinations. A PLR into this location would only be effective if it was accessible from both SB and NB, and would wind up resembling examples like this. I agree creating something like this again would solve the problem for this parking lot, but there’s other various destinations along this stretch potentially resulting in a lot of these unnecessary cross-over segments. Preventing u-turns at this point is not my concern since they should easily fit the requirements for UTP or be controlled via turn restrictions as you point out.
The real question to your point, is how the probability of this unwanted complication growing stacks up against temporarily losing some data per the un-divide process. This is the first roadway I’ve seriously considered un-dividing, since most of the other countless examples I’ve seen aren’t really harming anything other than not strictly meeting the requirements to be divided, and the guidance does say to leave situations like this alone. In this case, there may only be one UR, but I see the potential for additional issues and complexity looking at the big picture here. For proper routing, you’d need a crossover segment for every parking lot along this strip to essentially allow Waze to use the center turn lane, and the trend has already started. Given that this is one of the explicit reasons not to divide a roadway, I figured this warranted some attention.
Thanks for the clarification, I do understand more of what you’re attempting to do here.
When I re-read this entire thread, I realized we have a similar situation in Staten Island, New York. The roadway is Hylan Boulevard. This specific PL has three segments which have center turn lanes in them, and Hylan is setup bow-tiey. For these three segments where there are center turn lanes, the roadway was just a two-way single segment. For the rest of the roadway, it was split.
Perhaps something to this effect could be utilized. Further south on Fischer, I don’t notice any more center turn lanes. So for this little shopping center, if that’s what it is? I did not look too closely, you could combine the segments here. This would go a long way in attempting to address the problems you are looking out for, as well as maintaining the data with the other segments that are not too urgent.
Edit: I was unsure from your original posts how much of Fischer you intended to merge, so the bottom part of my post may be a bit redundant/obsolete.
Do we have any HNs or places along this section of Fischer Blvd not set back far enough for a PLR but would still be awkward to get a “you have arrived at your destination on the left” despite a physical median blocking the way? I’ve actually been splitting roads just because of this reason and nothing else due to numerous URs complaining about what I just mentioned.
Are U turns allowed at those other openings that don’t seem to work now? Are the 2 directions of Fischer Blvd set back far enough from each other to not prevent U turns? I don’t think road type is a problem; seems like they’re already being routed to a street, right?
Any areas I saw that have a physical median could be handled by PLRs, and I don’t see many of them to begin with. I could see how HNs for residences could be particulary troublesome in an example like this, but there’s always a center turn lane through these areas. I’m also not opposed to easing into this, and perhaps only addressing the northern sections with a focus on eliminating the existing hacks and solving the UR in question.
I don’t think u-turns are an issue here. The current configuration is favorable to trigger u-turn prevention penalties, which do indeed seem justified. While this is a correct setup, my only point it bringing UTP up is that it exacerbates the current issue with being routed to destinations on the other side of the road. Waze takes you out of the way because there’s no concept of the center turn lane without creating lots of cross segments. I’m personally not keen on continuing this workaround, and would rather correct the root problem here as I see it.
Alright, so back to my initial question - are there any places on the left side, going in either direction that are too close to the road for PLRs and cannot be turned into directly because of a physical median? I’m mobile right now so can’t tell.
Anything resembling the situation you’re describing could be addressed by a PLR. Most of the roadway does not even have a physical or painted barrier, which is the primary problem.
Given there’s no further concerns, it’s my intention to proceed on this action, at least on the northern sections. Please let me know if there’s anymore issues to consider.
What happens when people complain about being told their destination is on the left despite a physical median blocking access to it? I still don’t think unsplitting this road is a good idea. Elsewhere around the state were having to re-split divided roads that have been unsplit because of this issue, and PLRs with turn restrictions didn’t adequately help.
I understand the concept, and I thought we’ve covered already that there’s very low risk in this happening given the Waze places and HN setup along this road. Only issue I could see is if people are presented random pin coordinates from third-party searches along this road. Even if that is a concern, there are large stretches of this roadway that are not physically divided and have a left-turn lane. This can cause just as many problems since the divided nature of this roadway, as currently mapped, doesn’t account for them. Case in point is UR 5026274. Since U-turns are not allowed (and mapped correctly with UTP), approaching routes for destinations on the other side of the road start to become a bit ridiculous, especially when there’s a left-turn lane.
I understand if this is your guidance. If that’s the final word, do you have any suggestions to address UR 5026274, without using hacks like this? I’m at a loss on what to tell the reporter.
How about a compromise? :twisted: What if only the stretches lacking a physical median are un-divided? If alternating back and forth is undesirable, there’s two main stretches that could at least be addressed which include the location of this UR and some existing hacks.
I’m not trying to be a wise-guy here, but do you have examples of this? I’m having trouble visualizing how a single two-way with proper turn restrictions wouldn’t address the problem. Perhaps an example would help me see additional risk in doing this and better understand.
If there are ANY house numbers along the barriered two-way segments, or ANY destinations that “pin” to those segments, we run into this issue. You would have to remove all HN, and make sure there is a PLR for EVERY address and every place, plus we’d have to make sure that nobody adds a place without a pLR. The latter is impossible to absolutely control, given that we have trusted drivers reporting places from the app. On top of that, as you mention, there are Google results that may pin to teh road instead of a PLR.
Well from analysis I don’t think there’s anything currently there that couldn’t be solved with a PLR as I said previously. To your point, however, there’s always the possibility of someone adding something or getting a third party search result. It would seem a hack would be needed either way, unless: what about un-dividing the sections that do not have a median?
To answer this UR, I see adding another hack by allowing a PLR to extend over the divided segment or un-dividing the road entirely where there’s a left turn lane and no physical median. I understand divided roadways help where there’s a physical barrier and destinations on the other side, but left turn lanes would seem just as problematic when mapped as a divided roadway. The guidance specifically calls left-turn lanes out as a reason to un-divide. Why not just map this in such a way that doesn’t need hacks for either scenario?
The 2 main roads I have on the backlog for resplitting are US-46 in Lodi and Garfield, and White Horse Pike (US-30) in Absecon and Galloway Twp. Both have many URs, some of which are still open, with complaints regarding poor correlation between divided and split roads. You’ll have to search for those cities and streets though, since I’m mobile for the day.
I have no problem with a hybrid approach in which we would leave only the parts with a physical median blocking turns or an HN point split, right?
From what I can see, there is one house north of bay ave that is on the opposite side of a physical barrier median and then a few businesses on the corner of Bay Ave that may pose a problem such as described by phantomsoul. We can easily add the two or three plrs there to address. The majority of the barrier on this roadway exists only where there are no destinations on either side, or where destinations exist, turn lanes are provided. There are a number of houses along this stretch that can not be accessed from Fischer, but are close enough to Fischer to possibly be routed there if the HN are not set properly. Regardless though, these would require editor attention as routing would be wrong if the road was divided or not. In my opinion, there is very little risk here and the proposed changes would improve overall experience in this area.
Ok. But I’m just saying that if there end up being complaints about a physical median blocking the way to/from nearby destinations, we will have to re-split the road.
Appreciate all the feedback. I will concentrate on the portions that are not physically separated for now. With that completed, there should be plenty of time for a very thorough inspection and evaluation for the remainder. May make sense to phase this in slowly anyway and monitor for issues.
FYI, much of the 46 has long been on my list, too. I asked to do it when I was ran 2 or 3, and at the time, Orbit wasn’t ready to let me do it. I’ve since divided I think a few miles of it.