Do they need to have the same endpoints? Couldn’t one ramp segment just go directly from Charles St to 695, and the other to the 83 ramp (or the junction node of 695 and the 83 ramp)? This would avoid the “same-endpoints-multiple-segments” problem. I drew railroad segments at the interchange to illustrate my idea. A final layout would have the two ramps much closer together. There are two drawbacks I can think of: drivers from Charles St N wouldn’t get an exit instruction at the 83 ramp, and drivers from Charles St S and Bellona Ave wouldn’t get a “keep left at 695” instruction.
Actually, scratch that idea. Those two missing instructions are pretty important. On a side note, does anybody have documentation on the same endpoint-double segment bug? I’ve inadvertently laid roads out like this (usually single segment parking lot roads that have two entrances to the public road) and I’d like to know more about the issue.
I don’t. As is apparent in the WME Toolbox and WME Validator threads, there is some question regarding the prevalence and severity of this bug.
My experience is that waze sometimes (that is not always) fails to distinguish properly between two segments that connect the same two junction nodes. Sometimes waze uses one of the segments for one purpose and the other for another purpose. So issues may occur like:
- the UR purple line may show differently from the actual instructions given.
- if routing to or from a destination on one of the segments, waze may use the other in the routing instructions, and
- if you deviate from the route shown to the other segment, the deviation may not trigger a route recalculation.
It happens rarely but often enough and in serious enough circumstances that it’s worth editing around, in many editors’ opinions.
I agree that it should be locked very high. I understand that there’s been resistance to doubling up roads like this, but proper directions should always win out over editing simplicity and “keeping the database size down”, imo. The lock takes care of the first and the second isn’t really a compelling reason to leave directions worse than they could be.
The left turn is not prohibited because that’s the only way from N Charles to 83. If you’re going for 695, there’s a whole 'nother ramp there for you; it’s safer and easier, and honestly probably usually faster anyway since you don’t need to wait for a left turn light. Further, I don’t think Waze should disregard the DOT’s intentions just to save maybe a couple seconds once in a while.
So, plan of action:
- Unlock Segments (or CM makes changes)
- Double ramp from Charles to split at 695/83 ramp.
- Name Ramp A: “to I-83 N / York PA”
- Name Ramp B: "to I-695 W / to I-83 N / York PA
- Set turn restrictions so traffic from Charles St NB turns left onto Ramp A; All other traffic enters Ramp B.
- Existing C/D-type ramp at I-695/83 keeps current “to I-695 W” and “unnamed” designations.
- All road types “Ramp;” all locks rank 5 or 6.
Anything I am missing? Any changes to prevent the “endpoint double segment bug”?
Sounds correct to me.
To prevent the “endpoint double segment bug,” I’d include this junction node.
I’ve lowered the segment locks.
In thinking about this, I don’t think the drivers from Charles St N will get an exit instruction at the 83 ramp with either approach.
I would remove “York PA” from this ramp. It doesn’t appear to be present on any of the relevant signs.
What’s the deal with that roundabout in the aerial/some street view images? No longer there, or new?
Okay, I’ll remove York, PA. But I’m driving through tonight, so I’ll check the signs again.
Hit refresh and check it out.
The circle was removed during construction, its now a double traffic light. You get a good view during late stages of construction in streetview, and the finished product in trooster10’s videos from earlier in this thread.
EDIT: The streetview on Clark is the finished product: Google Link
I didn’t directly overlap the ramp so editors could see the intention, but its within a few meters, so I don’t think the users’ client will display two roads.
Done? Lock it?
Aren’t the turn restriction at the end of “Ramp A” reversed?
Yup. :oops: Thanks.
Looks good. I’ve locked back to 5. If you need to make adjustments after your drive through, let me know.
I’m also linking to the other thread that discussed this interchange for reference purposes.
Didn’t notice any changes needed after my drive thru.
Also, the Live Map and Clients have updated and seem to be giving good directions: Live Map Link
Still getting UR’s.
While what we did was for the best (from a safety standpoint) since it prevents routing NB Charles St drivers from using the northern ramp (to I-83N) to access WB I-695, Waze will still send those heading up I-83 on the first ramp (WB I-695).
That’s not as much a concern. Using the loop ramp to get to 83 doesn’t have the same safety implications as using the left turn ramp to get to 695 — the loop ramp puts you in the rightmost lane anyway, so it’s no different than 695 traffic going to 83. At this point, if you’re going to 83, Waze’ll just pick which one is faster.
I concur.
I still find this ramp unreliable in the app. I have been routed the wrong way consistently, and feel that the map should reflect reality, not just what the sign prescribes.
Locals use the straight ramp for both 83 and 695 access, every time. It is faster, obviously, and there is no lane switching to go either direction.
Waze has always been about using the most efficient route, avoiding traffic and so on. We have police and traffic camera reporting, so why not simplify this intersection? Proposal:
- Ramp 25.8 should just intersect 695 like any other ramp
- Ramp 14.2 should just divide right around the light rail overpass
- Let the app routing decide the best ramp to use
I haven’t driven ramp “25.8” since the completion of the construction. My understanding was that the merge to the beltway was difficult.
I’m not sure how there is no lane switching to either way. But if its straight-forward to go to I-695 W and faster, then that’s the way I would want waze to route me. Seems the reason that this ramp is unnamed is that the signage is really that you do take exit 24. So I don’t see a disadvantage to your proposal, except that some users no doubt will report that waze isn’t routing them through the loop 25.1 built for this movement.
Simply not true, there is a two lane merge to travel west on 695. I’m a local, and I frequently take the “right-hand” ramp from Charles, because its easier to move left on 695 and avoid merging traffic.
Correct. From the 25.8 on-ramp, you have to merge two lanes to the left to proceed on I-695 W. When you have traffic from 695 moving right to exit onto I-83, this can be annoying at least, dangerous at most, especially if you are driving there for the first time.
The traffic pattern looks as follows, with the bottom being 695 W coming from Towson:
|^|^|^| (3 Left Lanes Continue)
|^|^|^|
|^|^|^|<| (Right Lane of 695 Ends)
|^|^|^|<|
|^|^|^|<|---------------|>||>|
|^|^|^|<|----------|>||>|
|^|^|^|<|----|>||>|
|^|^|^|<||>||>|
|^|^|^|↔|>| (Two Right Lanes Split to I-83)
|^|^|^|>|>|
|^|^|^|>|>|
|^|^|^|>|>|
|^|^|^|>|<| (“25.8” Adds on)
|^|^|^|>||<|
|^|^|^|>|—|<|
|^|^|^|>|------|<|
|^|^|^|>|-----------|<|
==================== (Charles St)
|^|^|^|^|
|^|^|^|^|
|^|^|^|<| (“25.1” Adds on)
|^|^|^||<|
|^|^|^|-------|<|
|^|^|^|--------------|<|
|^|^|^|
|^|^|^|
OK I drove through and let me retract the above. However, it is still only one(ish see below) lane change if you take the 25.8.
I made a color coded drawing to illustrate this all. Blue + Yellow = Green. 
There’s also some GSV links in there.
The thru beltway is red. At the bottom of the ramp you’re destined to get on 83 North, unless you change one lane over to the left. That’s like a pretty typical onramp. Once you’re in that lane it diverges and the driver can either bear right to get on 83 or left to 695. The diverging lane is the same one that you would have been in if you took the loop ramp “25.1”. After the split that lane ends (after 1/4 mi), and so the driver must then merge again left. However that’s after the 83/695 craziness, and so that’s just like another onramp merge. You can either look at it as two total lane shifts, or one + one. Further, there are the drivers that are coming from Lutherville on Bellona. If they want to use the beltway they have to use the 25.8 ramp.
I attended the SHA meetings for this interchange and that diverging lane was specifically added for this exact issue. They knew drivers would use 25.8 for beltway entrance, and there was no inexpensive way to prevent that. (Early plans had a flyover from 695 to 83 before the Charles St. overpass, under Charles, over 25.8, eliminating the merge, hence the super wide underpass, but that was canned.)
Nick, sorry to put words in your mouth. I guess not everyone sees things the same.
Either way, these double ramps and stuff are really just workarounds to “trick” the routing engine right? If something changes on the backend, those could break something though, right?
This poses a tough question I think. Do we map for safety, reality, or the routing engine? I feel like this can be one of those triangles, with each of those things at the points, and we have to put the dot somewhere in the middle.
