What is needed for something to be classified as a 'cycle route' (London)

Previous Topic Next Topic
 
classic Classic list List threaded Threaded
8 messages Options
Reply | Threaded
Open this post in threaded view
|

What is needed for something to be classified as a 'cycle route' (London)

Simon Still
See discussion on 

There appear to be a large number of sections of road in some areas of London tagged as ‘cycle route’ that are no more than the occasional 1057 cycle symbol painted on the road.

They are not signed, and do not have any route numbering.

Based on the discussion it appears
- most were added by user MacLondon 
- they were the ‘lowest level’ of route designation by some councils at some time in the past. Pick some ‘useful routes’ on ‘quiet roads’ and just paint some symbols on them for people to follow 

Some of these appear on the last 2015 TfL cycle maps in yellow (routes were blue) keyed as ‘other roads recommended by cyclists’ 
 
My opinion is
- these are not followable on the ground 
- they do not meet TfL or borough quality criteria (and thus do not appear on any more recent maps) eg - they are not shown in any way on Lambeth councils 2017 cycle map https://www.lambeth.gov.uk/parking-transport-and-streets/cycling/lambeth-cycle-routes-map
- they decrease legibility of the map because they create a mass of dense blue lines from which it’s hard to pick out genuinely useful routes.
- they probably have a negative impact on routing engines as they are likely treated equally to actual signposted routes. 
- in many cases where they do show the most direct route through backstreets that is likely to be the busiest with rat running traffic as it’s where google and Waze will send drivers. 

Thoughts?



_______________________________________________
Talk-GB mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/talk-gb
Reply | Threaded
Open this post in threaded view
|

Re: What is needed for something to be classified as a 'cycle route' (London)

Jon Pennycook-2
Outside London, these days I tend to use direction signs or named route signs (eg named after planets/satellites/dwarf planets in Woking, or colours in Bracknell, or the "Cycling Discoveries" signs in north Hampshire) as an indication of an LCN/RCN. In the past, I had looser criteria, and I sometimes go back and remove LCNs that I added (eg there are some I added in Wokingham in the expectation that signs would appear, but they never did).



On Tue, 15 Dec 2020, 13:28 Simon Still, <[hidden email]> wrote:
See discussion on 

There appear to be a large number of sections of road in some areas of London tagged as ‘cycle route’ that are no more than the occasional 1057 cycle symbol painted on the road.

They are not signed, and do not have any route numbering.

Based on the discussion it appears
- most were added by user MacLondon 
- they were the ‘lowest level’ of route designation by some councils at some time in the past. Pick some ‘useful routes’ on ‘quiet roads’ and just paint some symbols on them for people to follow 

Some of these appear on the last 2015 TfL cycle maps in yellow (routes were blue) keyed as ‘other roads recommended by cyclists’ 
 
My opinion is
- these are not followable on the ground 
- they do not meet TfL or borough quality criteria (and thus do not appear on any more recent maps) eg - they are not shown in any way on Lambeth councils 2017 cycle map https://www.lambeth.gov.uk/parking-transport-and-streets/cycling/lambeth-cycle-routes-map
- they decrease legibility of the map because they create a mass of dense blue lines from which it’s hard to pick out genuinely useful routes.
- they probably have a negative impact on routing engines as they are likely treated equally to actual signposted routes. 
- in many cases where they do show the most direct route through backstreets that is likely to be the busiest with rat running traffic as it’s where google and Waze will send drivers. 

Thoughts?


_______________________________________________
Talk-GB mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/talk-gb

_______________________________________________
Talk-GB mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/talk-gb
Reply | Threaded
Open this post in threaded view
|

Re: What is needed for something to be classified as a 'cycle route' (London)

Robert Skedgell
In reply to this post by Simon Still
On 15/12/2020 13:26, Simon Still wrote:
> See discussion on 
> https://www.openstreetmap.org/changeset/95752985#map=18/51.46201/-0.12146&layers=C
> <https://www.openstreetmap.org/changeset/95752985#map=18/51.46201/-0.12146&layers=C>
>
> There appear to be a large number of sections of road in some areas of
> London tagged as ‘cycle route’ that are no more than the occasional 1057
> cycle symbol painted on the road.
>
> They are not signed, and do not have any route numbering.

Some of the LCN/LCN+ routes are signed with blue directional signs, but
often without route numbers.

You would need recent street level imagery or a survey to determine
whether a route really has degraded to only fading TSRGD diagram 1057
signs ("Cycle lane, track or route"). If 1057 is used on a carriageway
rather than on a lane or track, it presumably indicates a route,
although TSRGD 2016 does not elaborate upon this - is there an LTN which
does?

I am not very familiar with the area discussed in the changeset above,
but routes I have used in LB Hackney this summer certainly were.

>
> Based on the discussion it appears
> - most were added by user MacLondon 
> - they were the ‘lowest level’ of route designation by some councils at
> some time in the past. Pick some ‘useful routes’ on ‘quiet roads’ and
> just paint some symbols on them for people to follow 
>
> Some of these appear on the last 2015 TfL cycle maps in yellow (routes
> were blue) keyed as ‘other roads recommended by cyclists’ 
>  
> My opinion is
> - these are not followable on the ground 
> - they do not meet TfL or borough quality criteria (and thus do not
> appear on any more recent maps) eg - they are not shown in any way on
> Lambeth councils 2017 cycle
> map https://www.lambeth.gov.uk/parking-transport-and-streets/cycling/lambeth-cycle-routes-map
> <https://www.lambeth.gov.uk/parking-transport-and-streets/cycling/lambeth-cycle-routes-map>

I wouldn't trust most borough councils here, as the older LCN/LCN+
routes are likely to be the responsibility of TfL/GLA.

> - they decrease legibility of the map because they create a mass of
> dense blue lines from which it’s hard to pick out genuinely useful routes.
> - they probably have a negative impact on routing engines as they are
> likely treated equally to actual signposted routes. 
> - in many cases where they do show the most direct route through
> backstreets that is likely to be the busiest with rat running traffic as
> it’s where google and Waze will send drivers.

Unless there's a new modal filter as part of a Low Traffic
Neighbourhood, obviously.

>
> Thoughts?
>
>
>
> _______________________________________________
> Talk-GB mailing list
> [hidden email]
> https://lists.openstreetmap.org/listinfo/talk-gb
>


_______________________________________________
Talk-GB mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/talk-gb
Reply | Threaded
Open this post in threaded view
|

Re: What is needed for something to be classified as a 'cycle route' (London)

Simon Still


On 15 Dec 2020, at 14:35, Robert Skedgell <[hidden email]> wrote:

If 1057 is used on a carriageway
rather than on a lane or track, it presumably indicates a route,
although TSRGD 2016 does not elaborate upon this - is there an LTN which
does?

Not by any means.  1057’s are the ‘go-to’ way to DO SOMETHING for traffic engineers.  

- Cyclists getting hit by cars at a junction? Paint some 1057s across it ‘to alert drivers that there may be cyclists there” (though of course drivers should be conscious that there could be cyclists on any road) 

- can’t work out how to get cyclists around a bus stop or parked car? Paint a 1057 to indicate road position. 

OSM Wiki Cycle_routes
 
"Cycle routes or bicycle route are named or numbered or otherwise signed route” 

I would argue that a ‘route’ marked with nothing but 1057 symbols is not useful in any way and doesn’t meet that definition 

I have similar issues with London’s Q network - sections of un-numbered quietway.  However, these should indicate a certain level of service - ie that they meet TfL s quality criteria in terms of traffic volumes etc - but also have a point.  Q sections are supposed to be feeders for the strategic cycle network of QW and CS routes - ie follow a Q and you should soon get to a main, destination signposted, route.  (though again, naming and numbering being revised and all routes that meet *latest* quality standards will be C numbered)  

_______________________________________________
Talk-GB mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/talk-gb
Reply | Threaded
Open this post in threaded view
|

Re: What is needed for something to be classified as a 'cycle route' (London)

Andy Allan
On Tue, 15 Dec 2020 at 15:09, Simon Still <[hidden email]> wrote:

> Not by any means.  1057’s are the ‘go-to’ way to DO SOMETHING for traffic engineers.
>
> - Cyclists getting hit by cars at a junction? Paint some 1057s across it ‘to alert drivers that there may be cyclists there” (though of course drivers should be conscious that there could be cyclists on any road)
>
> - can’t work out how to get cyclists around a bus stop or parked car? Paint a 1057 to indicate road position.
>
> OSM Wiki Cycle_routes
>
> "Cycle routes or bicycle route are named or numbered or otherwise signed route”

I'm broadly in agreement with Simon's point of view on this one. I see
in many parts of the world the thought that if there is any form of
cycling infrastructure, it must be part of a route relation. This
isn't helpful. Some infrastructure is just there and not part of a
route. In fact, the "signed" bit of "signed cycle route" was not only
there to avoid enthusiastic mappers making up their own routes (from
whole cloth), but also to ensure that individual occurrences of
infrastructure aren't mistaken for routes.

* Not all bike paths are part of a larger signed cycling route.
* Not all bike lanes are part of a larger signed cycling route.
* Also, not all 1057 marked stretches of road are part of a larger
signed cycling route. (Same applies to sharrows, for our American
colleagues).

In saying all that, with the state of the art in the LCN era being so
low-quality, along with several years of neglect since then, it's hard
to tell just by looking at one stretch of road whether it is or is not
part of a longer route - often a lot of detective work is required! So
I think the best results are when there are some agreed broad outlines
of how we work, but we shouldn't be afraid to discuss and document in
detail specific edge cases.

Thanks,
Andy

_______________________________________________
Talk-GB mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/talk-gb
Reply | Threaded
Open this post in threaded view
|

Re: What is needed for something to be classified as a 'cycle route' (London)

Simon Still


> On 15 Dec 2020, at 17:39, Andy Allan <[hidden email]> wrote:
>
> es.
>
> * Not all bike paths are part of a larger signed cycling route.
> * Not all bike lanes are part of a larger signed cycling route.

But any cycle infrastructure that DOES exist - eg short sections of protected cycleway - will be picked up by routing engines and prioritised for routes



_______________________________________________
Talk-GB mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/talk-gb
Reply | Threaded
Open this post in threaded view
|

Re: What is needed for something to be classified as a 'cycle route' (London)

Robert Skedgell
In reply to this post by Simon Still
On 15/12/2020 15:06, Simon Still wrote:

>
>
>> On 15 Dec 2020, at 14:35, Robert Skedgell <[hidden email]
>> <mailto:[hidden email]>> wrote:
>>
>> If 1057 is used on a carriageway
>> rather than on a lane or track, it presumably indicates a route,
>> although TSRGD 2016 does not elaborate upon this - is there an LTN which
>> does?
>
> Not by any means.  1057’s are the ‘go-to’ way to DO SOMETHING for
> traffic engineers.  
>
> - Cyclists getting hit by cars at a junction? Paint some 1057s across it
> ‘to alert drivers that there may be cyclists there” (though of course
> drivers should be conscious that there could be cyclists on any road)
>
> - can’t work out how to get cyclists around a bus stop or parked car?
> Paint a 1057 to indicate road position. >
> OSM Wiki Cycle_routes <https://wiki.openstreetmap.org/wiki/Cycle_routes>

The wording in TSRGD 2016, however is "Cycle lane, track or route". If
it is on the carriageway and is not part of a lane, the assumption would
be that it indicates a route.
https://www.legislation.gov.uk/uksi/2016/362/schedule/11/made#tgp2-tbl2-tbd1-tr28-tc2

Luckily there is further guidance on the use of 1057, which is a bit
more detailed than TSRGD's description.

LTN 1/20 10.5.4 includes "Providing road markings to highlight the
presence of cyclists to other road users, such as cycle symbols to TSRGD
diagram 1057, lines to TSRGD diagram 1010 and advisory cycle lanes, as
well as coloured surfacing". 10.7.35 allows this on the approach to a
mini-roundabout: "Cycle symbols to TSRGD diagram 1057 may be placed in
the primary position to guide cyclists and to alert motorist to their
presence."
https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/906344/cycle-infrastructure-design-ltn-1-20.pdf

Chapter 6 of London Cycling Design Standards 6.2.5 includes this:
"Diagram 1057 cycle symbol markings should be selected according to the
width available: usually medium-sized, but small for cycle tracks and
large for ASL boxes. They are used, orientated in the direction of
travel for cyclists, in three distinct and well recognised ways:

• For conspicuity: alerting other road users to expect the presence of
cyclists
• For positioning: suggesting a recommended line of travel for cyclists
• For wayfinding: indicating a route, particularly at a decision point
Any use of this marking should either meet all three functions, or
positioning and conspicuity without an explicit wayfinding function.

The cycle symbol should never be used for
wayfinding where it compromises the positioning
function, particularly at junctions and past parking
and loading bays."
http://content.tfl.gov.uk/lcds-chapter6-signsandmarkings.pdf

You are correct in stating that 1057 outside a lane on a carriageway
does not necessarily indicate a route as I had previously thought.

> "Cycle routes or bicycle route are named or numbered or otherwise
> signed route”

However, if a cycle route in London has 1057 at 150-200m intervals (on
local streets), or 20-30m intervals on a main road route (LCDS fig.
6.2), it's an "otherwise signed" route. It's not a particularly well
signed route and not using 1057.1 for the route number is unhelpful, but
as some CS and Q route numbers appear to be changing to C route numbers
they could be out of date anyway (unless Will Norman changes his mind,
or Shaun Bailey has them all ripped up).

> I would argue that a ‘route’ marked with nothing but 1057 symbols is not
> useful in any way and doesn’t meet that definition
>
> I have similar issues with London’s Q network - sections of
> un-numbered quietway.  However, these should indicate a certain level of
> service - ie that they meet TfL s quality criteria in terms of traffic
> volumes etc - but also have a point.  Q sections are supposed to be
> feeders for the strategic cycle network of QW and CS routes - ie follow
> a Q and you should soon get to a main, destination signposted, route.
>  (though again, naming and numbering being revised and all routes that
> meet *latest* quality standards will be C numbered)  


_______________________________________________
Talk-GB mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/talk-gb
Reply | Threaded
Open this post in threaded view
|

Re: What is needed for something to be classified as a 'cycle route' (London)

Simon Still


On 15 Dec 2020, at 18:18, Robert Skedgell <[hidden email]> wrote:

You are correct in stating that 1057 outside a lane on a carriageway
does not necessarily indicate a route as I had previously thought.

"Cycle routes or bicycle route are named or numbered or otherwise
signed route”

However, if a cycle route in London has 1057 at 150-200m intervals (on
local streets), or 20-30m intervals on a main road route (LCDS fig.
6.2), it's an "otherwise signed" route. It's not a particularly well
signed route and not using 1057.1 for the route number is unhelpful, but
as some CS and Q route numbers appear to be changing to C route numbers
they could be out of date anyway (unless Will Norman changes his mind,
or Shaun Bailey has them all ripped up).

Coming back to this thread/discussion.  

I work for London Cycling Campaign as an infrastructure campaigner and have been thinking about mapping and wayfinding a lot in recent months and there is a real issue with different generations and standards of route signage.


The recent history of cycle routes from LCN (theres not really anything I’m aware of that predates LCN that wasn’t re-signed as part of it) LCN+, then CSH and QW and now just Cycleways. 

These three images show the challenge.  The first is an attempt to show ‘good safe routes’.  Tfl’s is obviously of limited use as there is a lot more out there that *is* useful. But OSM contains a lot of stuff that, in my view, doesn’t actually exist in a useful or followable form - which means its broken for the end user. 

“Safe cycle routes” - TfL cycleway routes, good quiet ways, good safe older infra, legal shared paths, parks https://ibb.co/W09wSxL

Transport for London - only what meets our newest quality criteria, has been assessed and signposted - https://ibb.co/8sPkMrq 

OSM - everything that has ever been marked as a route on a map or could possibly be interpreted as a route in the most liberal interpretation https://ibb.co/Xsxjgr4

In my local area around Brixton OSM includes/included  
- stuff that appeared on a map many years ago (but not on subsequent map releases) 
- a ‘planned’ route (Q5 across Clapham Common) that has still not been officially designated 5 years later with no evidence ‘on the ground’ 
- ‘routes’ on old maps (but not more recent) where some signs are still present but most are now missing. 
- ‘routes’ that consist solely of 1057 symbols on the ground, with no signage, that have never appeared as a route on ANY map that I’ve found. 


That raises a few questions for me 
- what is a route  *for the purposes of OSM* and are we ensuring that stuff marked as routes is *useful to the end user* being key the key ones.  

So - 

1) at what point does a route exist?
- the planned route across Clapham Common is interesting but not useful to anyone trying to navigate.  Would a note or link on the way to source be better than showing a line that may never actually exist be better for a future mapper to refer to?

2) at what point is a route considered to *cease to exist* for the purpose of OSM (so that it should be removed)?
TfL don’t consider any of the LCN routes to be ‘current’ based on the map above.  But some of them obviously *are* still useful and possible to follow

However, some LCN routes have definitely fallen into ‘disrepair’ so that they are not, in my experience, possible to follow on the ground.  https://www.openstreetmap.org/#map=16/51.4502/-0.1198&layers=C 
The LCN route that runs through the middle here - Leander Road - has no signage that I can find north of Christchurch Road (where there is a finger post to “Brixton”). The spur off to Brockwell Park has no sign on Leander Road (only on the estate road once you’ve turned) so that would be missed.  These LCN sections haven’t appeared on borough maps for many years. 

Likewise Battersea Park - Wandsworth council removed all LCN signage a few years ago. Cycling is permitted on the carriage drive but I don’t think it should be marked as a route (it should just be cycling=yes)

3) what’s the minimum requirement for a route?
The original discussion here.  I was removing sections of ‘route’ that had nothing more than 1057 markings, had never been on any maps 

Refers to ’named numbered or otherwise signed’ but it’s a wiki and it appears the people who’ve edited it are the same that have added these only 1057 symbol routes in the first place, which means it’s a bit of a circular discussion! 

I wouldn’t consider a 1057 only route to be signed.  Discussion with transport planners and engineers is that 1057’s are only useful as ‘additional reinforcement’ as part of a larger signing strategy for a route. 





_______________________________________________
Talk-GB mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/talk-gb