Route GuidesRoutes City GuidesCities Map Log in

Can users suggest local improvements to the algorithm?

5 Sep 2020
Find a better bike route. Try our map & route-planner »

Become a supporter

I love this site, and it offers the best routes of any I have experimented with. I just used it to map and then follow a 120 mile cross-country route from Cambridge to Evesham. But there were a couple of points where the suggested route deviated from the NCN route it had otherwise been following. One of them was actually quite nasty and dangerous: in Bedford, the Cycle.Travel app routed me across a really nasty junction onto the main road out of town whereas the NCN route continued for maybe half a mile along the river before crossing back to the same road. 

I imagine this is because there is a misleadingly labelled section of bike path along the main road in the underlying OSM/OpenCycle map. It's actually a lumpy shared use pavement and to reach it you must cross four lanes of traffic (at lights, but still).

Is there some way to feed into this site/algorithm a note to avoid specific junctions like that one, or should I try to edit OSM in some way to warn others? 


Sun 6 Sep 2020, 20:52

On the wider question – not currently. It’s something I’ve looked at but the challenge is that if a confident roadie comes along and marks (say) the A49 as suitable for cycling, it would break the routing for everyone else. So I’m a little wary of adding a subjective element to the routing like that.

But on the specific issue of Bedford, the reason it’s diverting is that OSM currently has NCN 51 marked as going up a set of steps from the river to Prebend Street, and tries to avoid steps. My recollection is that the route doesn’t go down the steps and OSM is wrong, so I’ll tweak the mapping and that’ll get fixed next time takes an update.

Mon 7 Sep 2020, 11:10

Thanks for Bedford!  

I do see the problem with differing ideas of what constitutes "safe" roads. Round Saffron Walden, where I used to live, there are some B roads I would avoid at almost any cost, yet there were regular time trials run up and down the A11 dual carriageway for the really hard core. 

The only algorithmic/non-subjective method I can think of is to look at traffic densities as recorded by google maps/Waze and avoid at all times the stretches which are jammed at rush hour but of course you don't have access to that data.