Hi,
I’m getting routed through a river sometimes when cycling in this area. The map does indicate a warning triangle there. I checked if it is something I could fix in OpenStreetMap, but it is accurately tagged there as a ford (with recent comments that it is deep/impassable). After trying it and having to turn back I looked for a setting to avoid this sort of thing but I don’t see anything like that.
Is the best solution to check a route for these warning icons, and manually add a via point? Or is there a global option I’m missing? I wonder if there is value in a negative via point option (i.e. ‘do not route via this point’).
here’s a route with the feature in question https://cycle.travel/map/journey/978960
Thanks :)
Comments
If the ford is genuinely, permanently impassable, it's probably best that OSM be updated to reflect that. Either by severing the path on each side, or putting appropriate access tags on the ford node if it's still passable by some vehicle types. With explanatory comments in either case.
Sometimes, in cases like this where a public right of way exists on paper but not in reality, people will add a stretch of "highway=no" over the missing bit, with a note explaining it.
Hi,
Currently, there is no method to introduce a 'negative' via point. Probably the optimum solution is to add a marker with an appropriate icon to at least remind you.
A similar issue was discussed previously.
Reading more into this, the OSM wiki page for the "ford" tag suggests using "smoothness=impassable" on the path segment crossing the river, or specifying a suitably high "depth" value on the ford node. I wonder if C.T checks for either of these?
Yep, the best way is simply to tag the ford with its depth. Use the depth or est_depth (i.e. estimated depth) tag. c.t will refuse to route through any ford of 0.3m depth or more.
Also in the Oxford area, you’ll note that it won't route through Islip Ford as that's tagged with depth=0.7 . I have carried my bike through that ford but I wouldn't want c.t to send me that way unwittingly!
Isn't the problem that river levels are rather variable? Even 0.3m in normal conditions is pretty challenging, but after rain could well be into dangerous adventure territory.
However, in the UK, fords with a rideable road surface will usually (almost always?) have a nearby footbridge that gives a feasible "bail out" alternative. That combination means most cyclists will not be stuck if the ford is impassable due to high water (#), so it would seem ideal if CT's algorithm were able to detect that combination and factor it into the calculation of the routing.
# (unless flood conditions also make the footbridge inaccessible ...)
I don't think that requires any special handling - if there's a footbridge then c.t will consider routing across it (as a dismounted cyclist!).
Apols, I have unintentionally managed to create a sub-thread pursuing fords and footbridges. I was looking at how 2 fords on the River Wear that do have adjacent footbridges are handled - see the test route in the subthread item created just now. Streetview shows the fords and footbridges well.
Appears to me that the routes are using the road through the ford (paved but rather rough and can be deep) rather than the adjacent footbridge - se especially the ford near St John's Chapel. I'm not saying this matters in these cases, as the bail-out to use the footbridge is obvious on the ground - but it would matter more if the algorithm would route the same way even if there is no footbridge. (I can't think of an example to test this, as most material fords have footbridges).