I've been using cycle.travel a lot (a big fan) and yesterday, for the first time in many km of touring, I came across a route suggestion (from the website) that was dangerous, on the way to Martigues. (I see that brouter.de also suggests doing the same thing). I have forced the route to follow the safer way we ended up taking at viapoints 1 and 2 in this route. If you remove those two points, the suggested route takes you up a busy off-ramp from P544 the wrong way with no verge . Instead we took the on ramp on the other side of the bridge, but then had to take our chances crossing the P544 to head south. (There really doesn't seem to be a legal, bike-friendly solution to getting onto this road, and there doesn't seem to be an alternative bike-friendly way to get to Martigue from this direction.)
cycle.travel doesn't class P544 as a busy road, but it was certainly busy on a Friday afternoon in January with lots of truck traffic. P544 is essentially a 1-lane autoroute, with a generally wide verge lane that was safe enough to ride along, but we took the opportunities to take quieter detours at via points 3 and 5. (Although the map looks like you might be able to take a right turn viapoint 4, there is a concrete barrier for 500-m either side of this point preventing it.) The only stretch that cycle.travel classes as busy is a short section where the verge disappears completely for ~100 m under the bridge at viapoint 6.
After turning off P544 just after viapoint 6, the roads were good.
Comments
Really interesting case – thanks for pointing that one out. I’ve encountered issues with Fos-sur-Mer before: as you say there basically isn’t any safe way to get there by bike.
Without any via points at all, c.t’s Arles–Martigues route is via Saint Martin de Crau and Istres, which is not brilliant but probably better (the D5 even has a cyclable shoulder of sorts).
Unfortunately there isn’t really enough source data in OpenStreetMap to differentiate between the various sliproads at the P544 junction – nothing about verges or otherwise – and the road is misleadingly mapped as a dual carriageway whereas in reality (as you found!) it’s a single carriageway which you can cross at your peril. I might attempt to fix that in OSM and see if that affects the decisions.
Thanks Richard