Gravel portions with profile road bike

openrouteservice directions
1

Hello,

has anything changed recently (about a month) in the route calculation algorithm (road bike)?
Or is it coming from the OSM data not being up to date for the road surface?

Because recently I’ve edited several routes with my editor based on the Openroute API and the road uses gravel portions that are difficult to ride with a road bike (route in France).
I have updated the OSM data for some of the sections used.

Apart from that, it works very well and Openroute is the most efficient routing service I’ve tested and use.

Thank you for that!

2

Hi @gegeweb,

The best source of information for changes regarding the backend and the API is the CHANGELOG.md.

It might have to do with #665

Best regards

3

Hi, thanks for the reply.

Before these changes, I recorded a track with the points (recommended, cycle route) for which the calculated routes between the points were indeed the desired route:

But now I have to move some points.
In particular n°7 and n°17. For the 17, the right and main road is “Ave de Nozay”.
The shotest way “Chemin du Collège” is good for mountain bike, but not cycling road wich will always take the main road.

Capture d’écran 2020-08-31 à 07.50.37
Capture d’écran 2020-08-31 à 07.50.37842×530 165 KB

Capture d’écran 2020-08-31 à 07.51.55
Capture d’écran 2020-08-31 à 07.51.55654×504 292 KB

And bettween 7 and 8, I need to move the 7 for taking “Route de Rambouillet”

Capture d’écran 2020-08-31 à 07.56.11
Capture d’écran 2020-08-31 à 07.56.111136×776 285 KB

Capture d’écran 2020-08-31 à 07.57.20
Capture d’écran 2020-08-31 à 07.57.201036×872 622 KB

The good tracks need to be now:

It’s like keeping shortest and not recommended.

4

For the portion of road at Marcoussis (91,FR), I updated the OSM data. It is explicitly forbidden to cycles.

But the “recommended” algorithm was much better before merging with “faster” and it’s removal.