Get a sneak peek at whats next for Permanent Hazards on our April 7th Office Hours!
This is the place to discuss issues that are relevant for locations in the US. For any other discussions, please use the main forums.
Post by qwaletee
Orbit,

The example they gave is 5th Ave between form 13th to 9th Streets.

The problem with the advice is it doesn't account for whether the intersections in the middle are actually closed, and it should be clearer about that. Sometimes, transverse traffic is allowed across a closed stretch of road, even though the road itself is closed. For example, say they were closing that stretch of 5th because they are laying new sewer pipe. They might bridge the three intersections shown so that 10th, 11th and 12th, even though hey cross right through the middle of the closed section of 5th, still allow traffic to flow from 6th Ave to University Pl (10th and 12th) or vice versa (11th). Because the more streets closed, the worse the traffic snarls up, and the more residents and businesses impacted. NYC DOT does this all the time.

In other situations, even 9th and 13th could be affected. If they had to close the entire length and could not bridge the intersections, they might also be fully or partially closing the two "end intersections.
qwaletee
EmeritusChamps
EmeritusChamps
Posts: 2939
Has thanked: 188 times
Been thanked: 958 times
Send a message
US Champ / Country Manager | State Manager NY, NJ, PA, CT, MA, RI, VT, ME, NH | Northeast ARC | Mentor | Responding to Map Issues

Post by qwaletee
For one way cross streets, there may be either extra work and/or a couple of holes in the system.

One-way street may be closed at their source, whether due to work on the street or because the intersection is closed due to work on the cross street. Thorugh traffic will generally not be permitted in such cases, but local traffic may be set up differently, in one of several ways:

1) the regular entrance may be open to local traffic, with only through traffic restricted

2) the intersection may be completely closed even to local traffic, but local traffic may be permitted to enter form the "wrong end" in the wrong direction

3) the street may simple be closed even to local traffic (intermittently or throughout the closure).

When it is truly-fruly closed (the latter case), it should get treated like a main street's closure - it is simply closed. Not complicated. But that's only if the closure is for the entire period, not intermittent. If intermittent, leave it open as per one of the the other two options, as appropriate.

For the middle option, when local traffic is permitted to enter in the normal fashion, but not through traffic, we get into the discussion seen above about possibly closing just the exit, which will prevent through routing. However, we must be careful in that case NOT to prevent traffic form entering that intersection. Otherwise, Waze will have no route into that local entrance. All entrances into the intersection may be allowed, or there may be a mix of open/closed. IN the NYC example earlier in the thread, since 5th Ave is closed, its entrances into the intersection are effectively closed, but the cross streets leading into 5th would need to be permitted at both ends, the cross streets leading away from 5th would need to open at the entrance, but closed at the exit. It is unclear to me how we enforce that. It seems to me the road closure spreadsheet would close the segment, and would be inappropriate, but we could use partial restrictions on the "exiting arrows" that lead onto or across 6th and University Pl.

Where reverse traffic flow is used, then it gets more complicated. It seems to me we would have to do the following:
1) well in advance of the closure, turn the affected segments into two-way, but with traffic restricted in the wrong direction EXCEPT during the closure, using two closre periods (e.g., from today's date through the last day of normal traffic, and from the day following the re-opening until say a month later)
2) use the spreadsheet to create three closures:
2a) the wrong direction until the closure period
2b) the wrong direction from the day after the re-opening until about a month later
2c) the normal direction for the period of the closure

After the reopening, delete the partial restrictions, and revert to one way. After the tile update, you can update the spreadsheet to indicate that the closure has ended, if you like. Just make sure to these steps before the month is up.

However, this only works if the closure is complete, 24x7, for the street closure period. If the street has normal night/weekend traffic (or vice versa), then I don't think the closure system works at all. We then only have the partial restrictions option available, which does not work as well. In a city like New York, that's probably no big deal, because at worst, Waze will give a wrong route for the area, but you can easily route around it. But if the streetscape is not such an obvious grid, and there's no local person direction detour traffic, it can make things difficult for drivers. This would be especially true if the roads are sparse with long stretches between public junctions.
qwaletee
EmeritusChamps
EmeritusChamps
Posts: 2939
Has thanked: 188 times
Been thanked: 958 times
Send a message
US Champ / Country Manager | State Manager NY, NJ, PA, CT, MA, RI, VT, ME, NH | Northeast ARC | Mentor | Responding to Map Issues

Post by qwaletee
StephenR1966 wrote:
PesachZ wrote: Two way segments can be closed in one direction only by noting the junction node IDs in order of travel direction you want closed.

For example this segment of King can be closed only in the direction leading away from Harbor Dr. by saying the closure should be in effect from node ID: 42525853 to node ID: 42525850



Sent using Tapatalk for Android 4.4.2
Thank you for that explanation. The node IDs field had me totally confused. Now it makes perfect sense. It brings other questions to mind, but I'll butt out now as to not hijack the discussion.
COlumn clearly needs a better title. Maybe "One way closure junction node ID order" if we're willing to have many words. Or slightly shorter "Junction IDs (1-way)," which is not as self-documenting, but at least conveys the gist and would be enough to jog the memory of someone who knew about it but couldn't remember quite how it works.
qwaletee
EmeritusChamps
EmeritusChamps
Posts: 2939
Has thanked: 188 times
Been thanked: 958 times
Send a message
US Champ / Country Manager | State Manager NY, NJ, PA, CT, MA, RI, VT, ME, NH | Northeast ARC | Mentor | Responding to Map Issues

Post by qwaletee
You forgot:

No warranty is expressed or implied. Past lack of performance may be indicative of future results. This illustration may contain certain forward looking statements that may not represent actual future commitments or performance.
dmcrandall wrote:
PesachZ wrote:Yes "soon"
"Soon" (c) 2014 by Waze.com. This disclaimer is protected by copyright and its use, copying, distribution and decompilation is restricted. All rights reserved. No part of this disclaimer or any attachments may be copied or reproduced, stored in a retrieval system, or transmitted, in any form, or by any means, optical, electronic, mechanical, photocopying, recording, telepathic, or otherwise, without the express witnessed and notarised prior written consent of the all holders of the relevant copyrights.
qwaletee
EmeritusChamps
EmeritusChamps
Posts: 2939
Has thanked: 188 times
Been thanked: 958 times
Send a message
US Champ / Country Manager | State Manager NY, NJ, PA, CT, MA, RI, VT, ME, NH | Northeast ARC | Mentor | Responding to Map Issues

Post by qwaletee
Correct, because Experimental is not in the link. Some might consider it a kind of a bug - we understand that the layers parameter should not record the temporary status of the experimental layer. But the PL generator should recognize that if you have standard roads off but experimental roads on, it should turn on the roads layer for you so you don't get ugly surprises like that.
qwaletee
EmeritusChamps
EmeritusChamps
Posts: 2939
Has thanked: 188 times
Been thanked: 958 times
Send a message
US Champ / Country Manager | State Manager NY, NJ, PA, CT, MA, RI, VT, ME, NH | Northeast ARC | Mentor | Responding to Map Issues

Post by russblau
Poncewattle wrote:Should the list of node IDs be separated/delimited by commas, spaces, or doesn't it matter and I'm just too obsessive about these sorts of details? :geek:
Yes. :lol:
russblau
State Manager
State Manager
Posts: 1797
Answers: 1
Has thanked: 356 times
Been thanked: 668 times
Send a message

Post by SkiDooGuy
taco909 wrote:Suggestion on the form,
The start/end time example is given as 00:00 and 23:59, but that results in an error that it does not follow format. Seconds are needed 00:00:00 or 23:59:59
It is correct in the example post in this thread, just needs to be added to the doc itself.
Actually it will all need to be adjusted to submit and show a time format of 12:00.

As that is what the new feature will be using.
SkiDooGuy
Coordinators
Coordinators
Posts: 2582
Answers: 1
Has thanked: 437 times
Been thanked: 1478 times
Send a message
Assistant Regional Coordinator: Great Lakes Region
Local Champ: Canada, USA
Country Manager: Canada, USA, Palau
Rank 6 editor

https://i.ibb.co/ZNbYyt5/RIPIain.png

Post by stephenr1966
PesachZ wrote:Your statement is true, we only have to worry about cross streets, if through traffic across those intersections of blocked. However the discussion is when we have to close side streets, which side should we close. Orbit is getting confirmation from Waze on that.
Another issue with the approach recommended (in the case that we really do need to close the crossing streets) is how do people get routed to the businesses and/or residences along those closed segements of 10th, 11th, and 12th streets? Assuming only the intersection is closed, it seems like it would be better to either disconnect the side streets leaving oneself a [NOTE] UR to remember to reconnect after the closure or add nodes on either side of each intersection creating 5 m segments and close those segments as well. It's more work for us, but what else is there for it until they make intersection blocking an optional part of the closure?
stephenr1966
EmeritusChamps
EmeritusChamps
Posts: 514
Has thanked: 888 times
Been thanked: 271 times
Send a message

Post by stephenr1966
PesachZ wrote: Don't disconnect any roads unless they are physically not roads anymore. That's the whole point of the closure system, so we no longer have to disconnect roads. Waze has asked us not to disconnect segments.

If the proposed change goes into effect, then people will be allowed to be navigate up until the closed intersection, but not across it, as the only segments actually closed would be one leading AWAY from the closure, which are inaccessible because you can't drive through the intersection.

Sent using Tapatalk for Android 4.4.2
Upon closer inspection of the example I see the crossing streets are all one-way roads so the proposed change to close the outgoing segments makes a lot of sense. I was thinking in terms of two-way crossing segments. How are we supposed to close the intersections without closing the whole side-street? Should we just restrict all the turns and the straight through arrow?

I offer this permalink from my area as an example. Harbor Drive is closed as are the intersections at County St and King St.
stephenr1966
EmeritusChamps
EmeritusChamps
Posts: 514
Has thanked: 888 times
Been thanked: 271 times
Send a message

Post by stephenr1966
PesachZ wrote: Two way segments can be closed in one direction only by noting the junction node IDs in order of travel direction you want closed.

For example this segment of King can be closed only in the direction leading away from Harbor Dr. by saying the closure should be in effect from node ID: 42525853 to node ID: 42525850



Sent using Tapatalk for Android 4.4.2
Thank you for that explanation. The node IDs field had me totally confused. Now it makes perfect sense. It brings other questions to mind, but I'll butt out now as to not hijack the discussion.
stephenr1966
EmeritusChamps
EmeritusChamps
Posts: 514
Has thanked: 888 times
Been thanked: 271 times
Send a message