Cycling infrastructure between towns

Help and support
41

I agree with @SimonPoole that a new highway=-tag for this purpose would be A Bad Idea™. For example @Richard’s prpoposal of a new route-type would be a much better way to do this.

However, there’s something else that has been bothering me in this proposal, namely:

How many cities are there in Finland that would connect to each other with uninterrupted cycleways?

Are any of those cities, for example, connected to each other by cycleways (outside of the Greater Helsinki Area, namely Helsinki, Vantaa, Kauniainen, Espoo and Kerava, which are all connected to each other by multiple cycleways)? I don’t think so, but maybe I’ve missed some cycleways that stretch that far uninterrupted!

There is no direct cycleway from Helsinki to Porvoo, for example. There is a Eurovelo route between them, however. Most of the way it is a =cycleway (though most of the way it is also foot=designated and segregated=no). Famously, there is (still) a six-kilometer stretch starting at Box (and the town really is called that) where the Eurovelo route goes through a highway=secondary with maxspeed=80.

One can bicycle from Helsinki to Porvoo on a slightly norther route through Hindsby, Hagalund and Hindhår (& it’s a much more beautiful route as well), but that goes mostly through highway=tertiarys with maxspeed=50. So neither (in fact none of the examples above) would fit @aktiivimallikansalainen’s proposal?

42

What works here for cars does not work for cyclists or pedestrians, the reason being practical maxspeed.

A cyclist will have roughly the same speed on a highway=trunk_cycleway as on a highway=path but a car will have a 3…4 times higher speed at highway=motorway versus highway=residential.

Same for pedestrians, a wide high quality pedestrian area is not faster then a highway=footway with surface=ground.

Just like roundabouts are a good idea for cars, do not use them for bicycle infrastructure.

43

Note that I’m talking about “villages” (or taajama for the actual Finnish concept).

Of the top of my head, there are uninterrupted cycleway connections between

  • Hyvinkää-Riihimäki
  • Riihimäki-Hausjärvi (Hikiä village, which is further connected to Hausjärven kirkonkylä.)
  • Hyvinkää-Nurmijärvi (Kirkonkylä, via villages of Noppo, Herunen and Rajamäki)
  • Nurmijärvi-Tuusula (uninterrupted to village of Nahkela, via villages of Palojoki and Murto. Between Nahkela and central Tuusula (Hyrylä) there’s a small section currently tagged as a normal road. This section used to be signposted as a normal cycleway, but now it should be tagged with highway=service, access=destination, cycle=yes, foot=yes, moped=yes.
  • Helsinki-Sipoo (village of Box, this cycleway qualifies even though it doesn’t go all the way to Porvoo. Box is just as valid end location as Porvoo would be)
  • Hämeenlinna-Janakkala (Turenki)

I have personally ridden these too on a bicycle, but can’t remember if there were interruptions

  • Lahti-Nastola
  • Oulu-Kempele-Oulunsalo-Tupos-Liminka-Lumijoki
  • Jyväskylä-Säynätsalo-Muurame

Other very probable locations:

  • Tampere and surrounding municipalities
  • Turku and surrounding municipalities

Other notable connections in the area Uusimaa area:

  • Hyvinkää-Tuusula (village of Jokela, is interrupted by a short section of residential road. In my concept, this would be a gap on the map by design, but I would qualify the rest of the cycleway, at least the part north part of this cycleway as it connects outskirts of Jokela to Hyvinkää)
    *Jokela-Järvenpää (this contains a short section that is tagged as cycleway, but in reality should follow the highway=service… tagset mentioned on the Nurmijärvi-Tuusula connection. I would argue the surrounding cycleway counts on other merits, but there would be a gap.

As this is meant to model cycling infrastructure, and not “fun and useful cycling routes” sometimes gaps are just the model working as intended. People who specifically want cycleways/lanes should be alerted that there’s something else going on, and decide what they’re comfortable with.

* I’m not going to get too deep into the local tagging practice discussion here, but situations like this have historically been signposted either as a cycleway (with limited access to motor vehicles) or by having no cycleway sign (but sometimes have a square routing sign for cycles), having a no motor vehicles sign with additional signs giving exceptions to what kind of motor vehicle access is ok.

The currently accepted tagging practice is problematic because it interruptes a de facto cycleway with a non-cycleway tag for a technicality. From a cyclist safety perspective, service vehicles can drive on any cycleway, if they’re servicing the cycleway infrastructure. The only thing that’s changed by these signs is that vehicles can use the cycleway to access a location (residence, other infrastructure that needs service).

44

Yeah, I’ve ridden many of the ones you mention in the south of Finland as well.

Earlier you said that routes aren’t what you’re after because they often contain gaps. Now you say that gaps are fine? Which is it?

Box (and mutatis mutandis for the other villages and taajama) is a rural community with eight hundred inhabitants. Porvoo is a city with sixty thusand inhabitants, a historic city-centre with a mediaeval cathedral. I wouln’t count them as equally valid endpoints (though more power to anyone if they happen to live in The Box :slight_smile: ).

I don’t see a lot of reason to tag these kinds of short connections in any particular way. Most routers will already find them fine anyways.

45

Citation, please. My objection to routes is that a route doesn’t tell the user anything about the infrastructure. Is it cycleway or highway=secondary? Who knows, a route will render regardless. An experienced road cyclist might not care, or even prefers the latter. But a parent riding with their kids? A commuter? Someone just starting riding for fitness? They’re looking for places to ride that are safe, and Eurovelo is very much not that.

Apart from Oulu, I’m not aware of Finland having any official on-the-ground route designations for normal bike commuting/routing, that also use roads. Every other bike route is sightseeing, bike touring etc. I really don’t understand the argument that highlighting core infrastructure should be done the same way international tourist routes are highlighted, and only that way. (I understand it somewhat in places like Belgium, where signposted as-safe-as-possible bike routes are commonplace.)

Let’s downgrade all the roads to Kouvola too while we’re at it. I mean who cares about Kouvola? :stuck_out_tongue:

It’s 35km to Box from downtown Helsinki, one way. I’m not aware of any OSM map that would currently render the cycleway at any zoom that fits the whole way on a normal-sized screen, even from the eastern edge of urban Helsinki.

Box has services, and it has people living in it. Some might even commute to Helsinki by bicycle. Of course Porvoo is more impressive from a touristic point of view, but that’s not even remotely the relevant to cycling infrastructure.

One thing I have said earlier, is that this is about maps, not routers. A router will find a route anywhere, some are better than others at communicating what that route is going to be on.

46

If this is the criterion, presumably all cycleways would fit it. I think @Woazboat nailed it above!

47

I’ve given a proposal of what the criteria might be above, but you chose not to comment on it. Instead you quickly jump on this one sentence, where I did not fully repeat myself.

This is a fairly frustrating way to have a serious conversation.

48

I’ve no idea. What I was trying to do was to help find a way of simplifying the problem, while still being at the scale (zoom 9) that the original question was posed at.

More people live in Porvoo than Box, so “infrastructure to Porvoo” is more relevant to more people than “infrastructure to Box”.

Indeed, and this is a very simple map (just an overpass query) that shows that there is a cycleway at least most of the way between Box and the outskirts of Helsinki. You absolutely could create a web map to show that a bit nicer.

49

I did read the criteria you listed. My serious question to you is that if you include every village and town, are there any cycleways left that wouldn’t fit the criteria. Are there cycleways that lead to nowhere?

A final suggestion from me: if you are only concerned about visualization and not about routing, couldn’t you solve the problem without any additional tags or changes. Namely: just make your own map where all highway=cycleways are always drawn with a thick blue line. You wouldn’t even need to filter out cycleways inside cities, since at very high zoom levels they will turn into a blue mush anyway. Ditto for the gaps: at high enough zoom levels, small gaps will become indistinguishable.

50

The vast majority of cycleways are fully contained in a single city/town/village. Are we having a serious miscommunication here?

That also is the reason it’s not possible to render all of them at low zoom, because what serious map wants to render a blue mush. How is this even a suggestion?

51

Fair enough. For the record, any argument from me for “some kind of cycleway hierarchy” can be assumed to include openness for more hierarchy levels than “residential” and “connecting villages”. I’m fine saying e.g. a cycleway from Helsinki to Box would be tertiary, and if it was ever extended to Porvoo it might become secondary or primary.

But I’m not at this moment prepared to argue the relative merits of villages and towns in Finland, and I kind of feel that would be jumping the gun at this stage of discussion. That kind of thing would presumably be up national level tagging guidelines, once a framework for cycleway hierarchy is established. Appreciate any guidance if that’s a mistaken assumption.

52

Firstly, you’ll never know what it looks like until you try it.

Secondly, Richard suggested a way above of showing things inside of and outside of urban areas differently. “is X inside polygon Y” is apparently supported natively by Tilemaker**, which I use for vector maps. I’ve not used it for that, but it sounds perfect here.

If you try and find reasons why something “might not work” I’m sure you’ll succeed, but it might be worth “just trying something simple” to see what does.

** which (in case you weren’t aware) Richard also wrote.

2 Likes
53

See my illustration of this problem here:

54

I concede Richard’s suggestion might solve the simple two level rendering issue. Of course I thought we just moved to having more levels than that :slight_smile:

My problem is basically that I’m not looking for clever hacks to create a personal map. As I have said multiple times now.

I use cycling map apps created by others, including three that I’m paying money for. These maps have existed for years, and none are doing what Richard suggested. I guess I could send them a message asking for it, but I don’t think I’m wrong to be less than optimistic. Still, writing those messages would be easier than arguing about tags here, so point taken.

But my thinking behind this topic is that rendering support for “simple hierarchy, just like roads, look at these nice tags with familiar names” would be simple to understand in a feature request from a customer. A known technique, predictable rendering, no potential dangers about what kind of polygons exist around the world etc. Should be an easy sell to any map under active development, but this is probably me being too optimistic for a change.

I also do genuinely think that cycleway hierarchy is to some degree “real/objective” data about real world objects, with inherent value, that is currently missing from OSM. I’m somewhat surprised that this aspect seems to be readily dismissed by the commenters (I can understand saying that proposed criteria are bad, maybe they are, but it feels like people aren’t even seriously entering that consideration.)

But simply from the perspective of helping me as an individual, I don’t feel like this second aspect is my fight.

55

You can just run a query against a bunch of built-up area polygons if you want to render stuff within towns differently - that’s what I do.

To clarify as I did not understand what I read at the first time, does polygon here refer to things like “residential area” or “park”? The local practice seems to be a bit erratic, but a lot of roads and cycleways inside towns are left outside any area like that, seemingly at random.

areas
areas415×364 17.2 KB

I think this would lead to undesirable amount of cycleways in town being rendered. And tons of partial cycleways would probably be more distracting than, say, a cafe or two appearing a bit earlier than expected.

Or if polygon means boundary (city borders), those at least locally compass all surrounding countryside as well, i.e. there’s no area that is not inside some city borders. Anything resembling “borders of built area” doesn’t seem to consistently exist. I can see some village or suburb boundaries, which would be perfect if they existed consistently, but unfortunately they don’t.

Drawing missing suburb/village boundaries might be a doable task, though I guess I’m a bit unsure what the expected quality of data here is (e.g. does an official definition exist, in a human digestable format…)

56

There are many other consequences of the road system/design.

Why would practical maxspeed, at this multiplier, be the sole criterium for determining if a road hierarchy “works”?

(also, to parrot the criticism of a hierarchy for cycleways: the practical max speed of the road network is best captured using other tags that describe the characteristics of the road, not some abstract hierarchy)

57

I know the number of lanes, can you give the many others?

Related question: What kind of indicators do you see to create a hierarchy for cycleways?

58

Continuity, priority, safety, accessibility, navigability, to mention some. To say that cyclists travel the same speed regardless of road design is perhaps true in a very limited context, but obviously not true when you consider the same elements that affect real speed for drivers – intersections, mixing with other transport modes, road type transitions, priority, etc.

These would have to be worked out locally, I don’t presume to make a list here that will fit wherever you are perfectly. Indicators, not written in stone. OK?

Some cycleways are directly numbered according to a hierarchy already. If I go to the online national roads database and click a cycleway, it might be labeled “cycleway fv44” or “cycleway ev39”. But many cycleways are simply not numbered that way.

Then there are signs. Signed routes are obviously one indicator where they exist and are properly mapped so you can know that it’s a route type you care about, etc.

Many places don’t have signed routes, or routes at all. They might have non-route signs, like this example where there’s no route, but the advised cycleway to get to Vikeså is clearly signed, and there’s the reciprocal sign at Vikeså pointing the opposite direction. This is clearly the main cycleway between these two villages, even if you hypothetically were able to find another way to get there.

image
image940×892 57 KB

image
image1657×1068 88.4 KB

There’s also ownership. Like with carriageways, ownership of cycleways is hierarchical. There is a correlation between who owns the road and what role it plays. A county-owned cycleway is likely to play a more signficant role than a municipal cycleway, just like a county-owned carriageway.

At this point we’re approaching highway=tertiary tagging territory. The guideline for NO to determine highway=tertiary says to estimate importance based on if driveways connect to it or it branches off to other roads for that, a design to cater to a greater amount of traffic, bus routes, etc. Likewise, a cycleway with no clear indicators from above could be judged based on connecting to properties or branching off to other cycleways, if it’s designed for a larger amount of traffic, there’s segregation between pedestrians and cyclists, etc.

And I’m sure smart people would contribute to this as well.

2 Likes
59

For the highway classification, correct me if I’m wrong but I understand OSM started out basically mirroring UK highway classification (A road, B road, C road, “unclassified”), and these were then more-or-less successfully adapted to realities in other regions. Most bear some relation to official classification (some variation of national, regional, local, etc), though there’s increasingly been pushback on depending solely on official classification. And the criteria differ by region, so while some regions will reserve a highway=primary to a high-throughput urban highway, others will end up using it for two-lane roads, sometimes unpaved. Creating analogies to cycle networks don’t seem too farfetched: base on official classification, probably a rating of importance based on destinations connected, then some rating of infrastructure-as-built as a proxy of importance to the building authorities.

I don’t think inventing new top-level tag values like highway=cycleway_trunk would be successful, but I could understand using highway=cycleway + cycleway=primary in regions that have enough cycleways to make a hierarchy of them. I can’t contribute with exact criteria because my region doesn’t have that :person_shrugging:

2 Likes
60

One of the photos below shows a way with two lanes, extensive guard rails, reflector posts and a sturdy bridge that doesn’t require lower speed limits. Not shown in the photo, but this way has signs saying where it goes and far it is. I don’t personally consider those features necessary for classification.

This is one of the cycleways I suggested as being higher in hierarchy than normal residential cycleways. It’s the only cycleway between towns of Lumijoki and Liminka, and it’s more than well-equipped to handle the normal levels of traffic on that route.

The other photo is an example of highway=trunk from the wiki. It may be an extreme example, but I know it’s not the only single lane trunk road in Northern Europe. Just as a reminder not to get too hung up on indicators.

https://upload.wikimedia.org/wikipedia/commons/0/06/Einbreid_bru%2C_Iceland.jpg(image larger than 4 MB)

IMG_5828
IMG_58282016×1512 1.49 MB

1 Like