`lcn` tagging in Australia

[Migwell]

I think we should formalise the tagging rules for lcn and above.

Motivation

My concern is that the city suburbs of Melbourne and Sydney at least (probably other cities too) are littered with lcn=yes which show up on CyclOSM and imply the existence of a network that either has signage or cycle infrastructure or local awareness, which may not be the case. Consequently, the cycle infrastructure can appear safer and more extensive on OSM than it is in real life. For that reason I’m inclined to remove many of these tags. Here is an example of this phenomenon in Melbourne using CyclOSM. The pale blue streets are lcn=true despite often having no physical signage:

There is some discussion about the difference between lcn, rcn etc here, but I’m less concerned by the difference between them. I simply want to find a rule to answer the question “is this street part of a cycling network?”

Each region can have its own tagging guidelines for lcn, and Australia’s are here. I’d like to focus on this line:

These routes are often maintained and signposted by local councils and can include fragmented infrastructure and bike lanes. Local routes may be identified only by destination signage rather than having a gazetted name.

This descriptor is vague enough to allow almost anything to count as lcn, which I don’t like.

Proposal

Here is a new definition that I’d propose:

A cycle network in Australia does not need to have a name, but must be identifiable by physical signage indicating that it is recognised by some level of government. A street is not part of a network just because it has cycle infrastructure, or connects to a prominent cycle route, or has signs towards other cycle infrastructure.

Examples

Here are some examples all from the Boroondara LGA, for consistency.

This is located about here. Although this is a dangerous route along a main road (Power St), I guess we have to tag this as lcn?

Location. This is an easy lcn=yes tag, because the local government recognises it and states as such.

Location. I don’t think this counts as an LCN, as this is just pointing to another notable bike trail, and doesn’t form a network itself.

Location. This has no cycle signage, so I think it’s not LCN. However it does show up in the Travelsmart map (see next section).

Council Maps

To add some confusion, many Melbourne councils have a so-called “Travelsmart” map, which is annotated with “Informal Bike Routes”. Here’s an example in Boroondara corresponding to the OSM screenshot I attached (or you can view the original here:
.

Almost none of these dotted blue lines have physical signage, but should they count as lcn=yes? If so, we set a precedent for virtual information being used in the absence of physical signage.

---

Please share your opinions on my proposed definition and these examples, so we can work on a consensus!

[wolfy1339]

You probably should have posted this in Oceania which is the category for the Australian community and a couple other countries.

This discussion seems more relevant to your local community than worldwide

[Migwell]

Okay, I’ve now moved it.

[cs09736 (Sam Cantwell)]

I think nowadays many routes aren’t marked by physical signage but are instead listed only on council websites. For example, a council might create a new cycle route, paint sharrows on the roads belonging to the route, publish the route on their website, and just rely on the sharrows to show the route in person without adding any other signs indicating the route. Would such a case count as an lcn under the proposed definition? In my opinion these cases should be included.

[Migwell]

No, in my proposed definition this would not count as lcn. In your example, the sharrows themselves would be mapped, but sharrows alone don’t define lcn.

My main concern was that online data fails OSM verifiability. However, on consideration, a lot of OSM data like locality boundaries is derived from government data anyway, and it would be a shame not to capture this in OSM. Maybe we can just define a lcn:source that allows for physical or online references but requires something. In a future post I’ll try to summarise this in a proposal.

Also as an aside, lcn=yes only makes sense when there isn’t a specific start and end point, because if there is, then it should be mapped as a relation.

[aharvey (Andrew Harvey)]

@cs09736

cil websites. For example, a council might create a new cycle route, paint sharrows on the roads belonging to the route,

To me the sharrows are signage and sometimes their placement create a clear intent that they form a route, so these could be mapped as a route, just without the origin/destination or name tags you might add to a route with more signposts.

[Migwell]

But then we have redundancy where all cycleway:lane=pictogram are lcn=yes.

[aharvey (Andrew Harvey)]

I think it depends if the bicycle pictograms are just added to every local street to reinforce the space is shared, or only along selected roads that join together. More sore if you have turn arrows painted on the road too. I don’t feel too strongly either way though, just to me these feel like a route.

eg.

at OpenStreetMap

[Migwell]

Yes, I think that makes sense as an lcn, but again it would be marked as a relation and not tagged on the streets.

[Migwell]

Here’s my revised definition:

A cycle network in Australia does not need to have a name, but it must be recognised by some level of government. This could be demonstrated at the site of the route using wayfinding signage, pictograms to direct a cycler, or other infrastructure that makes the route stand out from the nearby streets. Alternatively, the route might be recognised by government resources such as TravelSmart council maps. In either case, these sources should be indicated using source:lcn or source:network to ensure the route is verifiable.
A street is not part of a network just because you use it, because it’s popular, or because it’s safe. Unfortunately none of these criteria are verifiable.

[Peter_Elderson]

@Migwell

Location. I don’t think this counts as an LCN, as this is just pointing to another notable bike trail, and doesn’t form a network itself.

This can be incuded in a route relation as a member with the approach role.

[cs09736 (Sam Cantwell)]

How should we handle cases where a route exists on a website but has no physical infrastructure?
I agree that roads with sharrows should be included if they are also marked as a route on the council maps, but if a route exists on those maps and has nothing to back it up in the real world I’m less inclined to say it’s part of the lcn.
Often these cases exist when the council has declared a future or planned cycle network but haven’t physically implemented it yet

[Peter_Elderson]

@cs09736

Often these cases exist when the council has declared a future or planned cycle network but haven’t physically implemented it yet

In my experience: better wait until physical implementation has at least started, and even then be cautious because plans often change during implementation.

[Migwell]

Right, so then we’re back at my original definition, which excluded “virtual” routes.

The only reason I considered allowing them is because they sometimes contain useful information about which streets the council deem to be safe, low traffic, traffic calmed etc. However I agree that they lack verifiability.

How can we proceed if not everyone agrees?

[cs09736 (Sam Cantwell)]

I’m happy to exclude virtual routes that have no physical implementation.

So I’m clear about everything above, are we currently saying that the following is an lcn:

• Clear physical signage showing the directions of a route.
• bike routes marked on virtual government maps that also have physical infrastructure backing them up (such as lanes, sharrows, or way finding signage)
• Bike routes marked only by paint on the ground but which clearly defines a route by directing the rider in a particular direction (such as in the aerial image shown earlier)

I don’t believe there is meant to be a difference in meaning between tagging lcn on the way and tagging it on a full route. You would use a route relation when there is a defined start and end point of the route, while you would use an lcn tag when the layout of the route is unclear but the road still meets the lcn criteria above.

I’m happy with these criteria which I think is what has been said above.

[aharvey (Andrew Harvey)]

I’m hesitant to suggest mapping from government maps, most council don’t release their content under and open license and even less have completed the OSMF CC BY waiver. Therefore we should not be copying routes from council maps, at best you may use them as hints of places to check on the ground or other sources we can use.

[stevea]

I have a fair bit to say as an OSM mapper who maps local, regional and national bicycle routes, but in the USA, not Australia. So I want to be very careful about not stepping on any Down Under toes being a faraway Yank. About 15 years of (nationwide) bicycle tagging consensus can be found on our USA Bicycle Networks wiki, which starts out with a critical section we have only recently (in the last several years) added:

What we map / do NOT map (for cyclists, as either infrastructure or route relations).

This is crucial, please read it and please document it for Australia if you agree.

Then,

There is INFRASTRUCTURE tagging (actual cycleways, bike lanes, sharrows…) and
There is ROUTE tagging (network=lcn, rcn, ncn, icn) with route relations.
AND, in the USA, we find benefit in additionally tagging with cycle_network=* to describe WHICH network of routes is being denoted.

Examples I see here 100% benefit from “sharrows” tagging (cycleway=shared_lane). The question becomes whether they are also part of “an” lcn (local cycleway network as a local cycleway route relation). If you read our wiki, we allow “seems like the local jurisdiction wants this treated as a bike route…” to be tagged lcn=yes, either directly with that tag on the elements of the street, or collected as members of a relation also tagged lcn=yes. However, while we usually associate these with a green “Bike Route (bicycle glyph)” sign (white text and glyph), some mappers don’t always follow this. It seems like @cs09736 (Sam) would say “no sign, no route,” but others want to make these into a route (relation). A (fairly stated) “virtual” route. That’s up to you (Aussies), but really, if I may offer helpful suggestions, please keep infrastructure tagging separate from route tagging. The former (infrastructure) DOES render (on CycleMap layer and CyclOSM), even if what you think you prefer to see (rendered) is the latter (a route relation at a particular network level, maybe because you think it ought to have a “badge” or a “number” on it).

Regarding whether what a local council “publishes” as a “route network” (for bicycles), well, (copyright) laws differ in Oz regarding applicability to whether these are ODbL-compliant, so I’ve nothing to say about that. But if these routes ARE signed (whether simply an lcn=yes route or an actual network of numbered routes where you put the number into a ref=* tag), OSM can map those. And should.

So: infrastructure, yes, map it! “Assembled into a route?” Well, there are steps along the way that you’ll need to come to some consensus about that (regarding signage, and how local councils do or don’t publish such things). The good news is that you can certainly do this (we have in the USA, and it’s working). It (simply?!) takes talking about it, just like you are here.

Wishing you the best from California!

[Migwell]

@cs09736 Generally yes, but I would cut point #2 because it’s already covered by the other points and we don’t want to rely on virtual information.

I think, as @stevea has alluded to, we need to distinguish networks from infrastructure: they must form an actual route. This can be indicated by signs or arrows but they need to be directing you somewhere in a way that other infrastructure does not.

[cs09736 (Sam Cantwell)]

After reading through the US’s guidelines and considering Andrew’s comment about copyright I think I agree, point 2 can be cut.

Im happy to go with your definition. I think we could work on the wording to eliminate any ambiguity, but that can always be done later.

It seems like the US guide says similar things to what we’ve discussed in this post. The thing that needs to be figured out and made clear is which signs designate a route, and which are merely used as general wayfinding. I haven’t seen any numbered local bike routes in Melbourne where I map. I mostly see the blue wayfinding signs, however, as shown in some of the pics above, these don’t always specify a route and are sometimes only used to show where significant things are.

We could also use this as an opportunity to clarify physical infrastructure tags in Australia on the wiki. There is a detailed page on the tags used in recent mapping in Melbourne which could be copied into the main Australian wiki page after discussion.
https://wiki.openstreetmap.org/wiki/Melbourne_Bike_Lane_Project

[Migwell]

Okay great. Here’s a revised definition. I don’t think the Wiki itself has a review process (?) hence why I’m drafting definitions here.

A cycle network in Australia is defined by directional signage. Typically this involves signposts or occasionally street markings that indicate a route designed for bicycles.

Specifically, a network needs:

• Explicit recognition that this route is for bicycles, either in words or using iconography
• One or more unambiguous directions, involving roads, paths or cycleways that the rider is directed down using physical signage

A cycle network may have, but cannot be defined exclusively by:

• A name
• Bicycle-specific infrastructure such as bike lanes, sharrows or shared paths
• Appearance in council maps or other online resources

A good rule of thumb is that, if you see a sign that has both a bike and an arrow on it, then it probably indicates a cycle network.