Evaluating the Mapping of UK National Cycle Network (NCN) Routes

Communities United Kingdom
1

I’d like to open a discussion regarding how we currently map the routes designated as part of the UK National Cycle Network (NCN).

We all presumably agree on the importance of accurately mapping dedicated cycle paths and infrastructure (e.g., using highway=cycleway). These elements are crucial for all users, particularly less-confident or family cyclists.

Mixed-Traffic NCN Segments And Their Issues

The primary concern I want to raise is with NCN segments that abruptly transition from dedicated paths to stretches along or across busy, mixed-traffic roads, often with:

  • Minimal or Zero Dedicated Crossing Infrastructure: Cyclists are expected to negotiate high-volume traffic without signals, dedicated lanes, or safe refuge points. Often they are expected to dismount on these segments of pavement as well making them wholly unsuitable for continuous cycling as you risk running afoul of the law if you fail to dismount.
  • Inadequate Cycling Provision: These are often simply busy roads with an NCN sign, offering no tangible benefit or safety improvement for a cyclist over any other road. Worse yet they can go from calm forest cycle-ways with ample safety to throwing you into mixed traffic roads without even the benefit of a painted on cycle lane.

Currently, if an NCN sign exists, we map the route using appropriate network=ncn and route=cycle tags. However, just because a segment is signed as part of a national network, does that mean it should be mapped as a contiguous, high-utility cycle_network route in the same way we map true, protected infrastructure?

As many cyclists can attest experienced riders will often ignore these poorly provisioned segments, preferring faster, more direct road routes. Timid or less experienced riders are effectively deterred by the sudden and significant increase in risk when a designated “cycle route” throws them into heavy, fast-moving traffic. The presence of the route tag may misleadingly suggest a consistent level of safety.

So I really struggle to understand who these lower quality NCN routes practically benefit. I fully understand the benefit to younger riders, those less experienced and less confident if the cycle route is a continuous route or has proper physical barriers and dedicated lanes when it joins roadways.

Proposal for Discussion

Should we consider a more nuanced approach to mapping these mixed-traffic NCN segments to better reflect their real-world utility and safety level?

For example, could we explore:

  • Utilizing tags that denote the quality or separation, such as adding route:quality=poor or similar, to flag segments that lack dedicated infrastructure? (This is a place-holder idea, and I welcome suggestions for established or new tagging practices or using tags that already exist I was unable to locate). (EDIT: Thank to @LordGarySugar on the discord for identifying this tag Key:class:bicycle - OpenStreetMap Wiki which would work very well in this case for tagging route segments.)
  • A discussion on whether the route=cycle tag itself should be universally applied to every signed NCN segment, especially when that segment is a high-traffic, non-separated highway.

The goal is not to remove data, but to ensure our mapping standards provide cyclists with a realistic and safe view of the quality of the route, not just its official designation.

What are the community’s thoughts on this, especially those who regularly cycle these routes?

1 Like
2

Just map the physical reality. If there’s an unsignalled crossing of a trunk road, map it as that. If there’s a section on a 1m-wide footway, map that. Any halfway competent router can and will take these factors into account

The relation merely means “this is part of a signposted cycle route with this number and this network”, nothing more.

Subjective assessments like route:quality=poor have no place in OSM. There’s a section of NCN near me which I think is glorious because of the scenery and the traffic-free nature, but others would rate poorly because of a rough surface and a hostile gate. Fortunately, we can already tag all these things.

6 Likes
3

(just for info in case anyone stumbles upon this thread and is unfamiliar with some of the previous discussions)

A while ago Sustrans (as it then was) started a process called “Paths for Everyone” which was designed to remove some of the scarier on-road sections from the national cycle network. That was one of the factors that led to this thread (and see also here in that).

That led to situations like this - the NCN here doesn’t run along the A170, but cyclists are still signed that way via generic blue signage (see changeset comment.

There are also plenty of other examples (and this bit of NCN67 was one) where there’s no on-the-ground signage, making it even more difficult to decide what should be in OSM.

I suspect that something might have got lost in translation there - taginfo suggests almost no route=cycle tags. network=ncn on ways is pretty rare too.

I think what people mostly do is “just add the way that is signed as part of a cycle route to that cycle route relation” and “make sure that all the relevant tags on the way (like surface etc.) are tagged on the way”.

What I’ve also been trying to do is to map the guideposts and route markers, with a role that indicates whether their “on the route” or “on the way towards the route”. That enables people to use an overpass query like this one to analyse where the signage is and how it corresponds to the route in OSM (often not well, and often it’s the signage that is “at best misleading”!).

1 Like