Routing to Samsø Rederi in Hou (access=yes node)

[G4rdener]

🐞 routing report

Routing engine

• OsmAnd's in-app offline routing
• Any online routing provider (YOURS, openrouteservice, OSRM, etc.)

Routing Profile

car (default)

Start and end points

Start: Hamburg, Flensburg, Kolding, Skanderborg, Horsens, Aarhus or Haldrup

End: Samsø Rederi in Hou (https://www.openstreetmap.org/node/3041302832)

Actual and expected routes

Expected routing: Routing on E45 until Horsens V (56a) or Skanderborg V (52) then following the 451 or 445 to Hou

Is this a regression?

Didn't try before Yes, the previous version in which this bug was not present was: ....

🌍 Your Environment

OsmAnd Version:

4.2.7

Device and Android/iOS version:
Nokia 7.2
Android 11
Maps used (online or offline):

• Offline maps offered within the OsmAnd app for download.

Denmark (01.07.2022) (Phone is connected to Wifi)

• Online (tile / raster) maps

Anything else relevant?
Routing on OSM.org is working fine, and in OSMAnd it is working from e.g. Odder or Søvin

Graphhopper

OsmAnd

[scaidermern]

The reason could be this gate which has no access tags set.

[G4rdener]

Actually this was my first thought too. But why is the routing working from Odder?

[vshcherb]

The route from Hamburg is OK, so the problem is in the last segment (https://osmand.net/map?start=55.794333%2C9.661102&end=55.912501%2C10.255158&mode=car#10/55.6218/9.7462)

And I think @scaidermern points correctly to the gate (https://www.openstreetmap.org/node/2207108788)
https://osmand.net/map?start=55.794333%2C9.661102&end=55.912447%2C10.258752&mode=car#10/55.7642/9.7572

[G4rdener]

Yeah that's correct but when you Route from Odder it won't Route over the two ferry connections but Route like exspected (3rd picture in OSMAnd)

[xmd5a2]

Routing time (cost) is 18445:
https://osmand.net/map?start=55.890090%2C9.972732&end=55.912513%2C10.259116&mode=car#10/55.6385/10.4548
But if you move start point 30 m east then routing time is 1512:
https://osmand.net/map?start=55.890063%2C9.973542&end=55.912513%2C10.259116&mode=car#18/55.89006/9.97501
There is no something special at this place in OSM data.
Routing engine bug?

[alisa911]

osmandapp/OsmAnd-core-legacy#103

[alisa911]

test files:
osm - newTestv3.osm.zip
obf - Newtestv3.obf.zip
osm of result - res.osm.zip
inspector result for 14 zoom - res.txt.zip
pull request:
osmandapp/OsmAnd-tools#524
before:

after:

map on 14 zoom:

[alisa911]

https://osmand.net/map?start=55.959585%2C9.712569&end=55.577859%2C9.676552&mode=car#17/55.95961/9.71514
from 2603 to 2586

https://osmand.net/map?start=55.897141%2C9.810405&end=55.912513%2C10.259116&mode=car#11/55.9421/9.9890
from 2825 to 2554

[vshcherb]

#14924 - probably related

[vshcherb]

#15384

[vshcherb]

#11770

[RZR-UA]

I've re-tested Issue with new OsmAnd HH-routing.

Finish point set exactly to Node: Samsø - Hou (3041302832)

Results:

New routing (HH) - OK - 53 km, 0:45 time, 3508 cost

https://test.osmand.net/map/?start=55.794333,9.661102&finish=55.912459,10.259108&type=osmand&profile=car#11/55.8589/9.8335

Current routing (BRP Java) - OK - 48 km, 0:47 time, 3575 cost

https://test.osmand.net/map/?start=55.794333,9.661102&finish=55.912459,10.259108&type=osmand&profile=car&params=car,hhoff#11/55.8589/9.8335

Current routing (BRP C++ native) - OK - the same as Java

[RZR-UA]

Additional tests are successful:

Test from Hamburg:

https://test.osmand.net/map/?start=53.558003,9.998245&finish=55.912459,10.259108&type=osmand&profile=car#8/54.669/10.085

https://test.osmand.net/map/?start=53.557977,9.997902&finish=55.912459,10.259108&type=osmand&profile=car&params=car,hhoff#7/53.778/9.558

Test to Hamburg with swapped start/finish

https://test.osmand.net/map/?finish=53.558003,9.998245&start=55.912459,10.259108&type=osmand&profile=car#8/54.669/10.085

https://test.osmand.net/map/?finish=53.557977,9.997902&start=55.912459,10.259108&type=osmand&profile=car&params=car,hhoff#7/53.778/9.558

[vshcherb]

Issue is finally fixed with new Special Routing