Sunday, 28 October 2012

A strategic cycle network

"What is needed," the LCC's Go Dutch advocate explained in a video that was presented to the membership, "is a strategic network of routes that look really attractive to cycle on." This is true not just in London, but in every town and city in the UK. The question is, firstly, Do you accept this? and secondly, How would you set about developing such a network?

I explained in some considerable detail how best to set about this task in an article that was published on the Movement for Liveable London website. The long and short of it, however, is thus:

1. Think in terms of a network.

Only by studying a cycle route network will it be possible to truly grasp the situation. (p.40, Cycling: 
the way ahead)

2. Plan the network.

Analyse journeys — origin/destination (headcountsstatisticsinterviews). (p.56, Cycling: the way 

Design from patterns to details.

3. Study the feasibility of the network.

Studying the feasibility of a network is of a similar importance to setting up a cycling unit or appointing
 a cycling coordinator. (p.57, Cycling: the way ahead)

4. Introduce the network.

The level of minimum functioning is a prudent course to follow. (p.56, Cycling: the way ahead)

5. Develop the network.

Strengthen the network further on the basis of priority interventions and a timetable. (p.56, Cycling: the 
way ahead)

The key here is sustained investment, progressively reprioritising the urban environment in favour of more sustainable forms of transport.

* * *

Who is the investment in cycling infrastructure for, and why do we make that investment? Is it for those who are already cycling, or is it for those who are not? 

Those of you that keep a close eye on the cycling blogs will recognise that these questions were posed recently by Freewheeler, who himself was quoting the Cambridge Cyclist. For myself, I am of the view that the introduction of a functioning cycling network should be undertaken with existing cyclists in mind first and foremost. Let me explain why.

The important thing is to get the network up and running, not least because a network is more than the sum of its parts. (For an alternative view on the power of networks, you might like to watch this video.) Making it safe enough such that unaccompanied children can use it, say, would come afterwards. Indeed, there is absolutely no reason why this work couldn't be started immediately afterwards (
Step 5 demands that there is a timetable in place). But first things first. 

The minimum requirements for a functioning cycle network are that the routes should be:

(i) meaningful (i.e., they go to the places that people want to get to);

(ii) direct;

(iii) pleasant (wherever possible, but never to the detriment of the first two-listed factors); 

(iv) joined up (a network, by its very definition, is inter-connected);

(v) functional in both directions; and,

(vi) effectively waymarked.

I am not aware of any town or city in the UK that can meet these minimum requirements. I am currently staying in Reigate, and I have just cycled to the shops. No problem for me. I am much more of a fair-weather cyclist than the Cambridge Cyclist, but even so, I am not put off by a bit of vehicular cycling. However, coming back from the shops, I found that I couldn't reasonably make this journey without breaking the law! How absurd is that? According to a Radio 4 audience, the bicycle is the greatest invention of all time. In which other countries in the world, and in what circumstances, could riding one of these beautifully simple machines ever be regarded as a criminal offence?

One commentator has recently said that "the screaming-out need [...] is to have the majority of the population onto bikes for everyday trips." No. That's our destination: in the short-term, there are more pressing demands. 

Mikael Colville-Andersen of Copenhagenize has said, "'Badly-behaved' cyclists are usually just cyclists with inadequate infrastructure" (quoted not so long back by AsEasyAsRiding). Providing 'adequate' infrastructure for existing cyclists is, I would argue, "the screaming-out need". 

Who is the investment in cycling infrastructure for, and why do we make that investment? Is it for those who are already cycling? Yes, to begin with, why not? They're the ones who are being killed and seriously injured! They're the ones who break the law just to get back from the shops! 

* * *

Cambridge is the cycling capital of the UK, but even so, it does not have a functioning cycle network. Indeed, the cycle map [pdf] produced by the local authority shows only the routes in that city.

If the City of Cambridge is minded to develop a network, I would recommend that they follow the five-step plan indicated at the top. Step 2 requires that the network is planned, designing from patterns to details. Since a significant number of routes have already been identified in Cambridge, I thought I would see for myself if it were possible to join these routes up to form a network.

* * *

Compass colours code routes according to a direction of travel, as can be seen here. (This map, and all the ones that follow, are best viewed with the terrain box (top right) ticked.)

As you can see, all I have done so far is to lay down one route in each of the five compass colours (plus the circular route, which is coded in yellow), but still, you can easily see how this would give the network its identity.

Next I coded the other primary routes on this network, as can be seen here (don't forget to tick the terrain box). Actually, there are a few new sections of cycle route which appear on this map but which do not appear on the Cambridge cycle route map as produced by the local authority. These are:

(i) The top half of Elizabeth Way (A1134) (coded in cyan);
(ii) Northampton Street (A1303) (coded in orange);
(iii) Mill Road / Parkside / the bottom half of Parker Street (coded in orange);
(iv) A farmer's track between Long Road (A1134) and Grantchester (coded in red).

In each case, I can see that there are alternatives to the new bits, but just to say, I haven't incorporated them into the proposed network for my benefit. The fact is, the new sections provide for more meaningful and more direct journeys, and if the authorities in Cambridge are serious about cycling, this is the sort of thing that they would need to pay attention to.

After this, I coded what I call the secondary routes, as can be seen here. My proposed design for a strategic cycle network in Cambridge is now complete. 

As before, I have added a couple more new sections:

(i) Chesterton Lane / Chesterton Road (A1303) (coded in red);
(ii) an alternative course for NCN Route 11 south of the Botanic Gardens (coded in cyan).

There are some 'blue' routes on the City of Cambridge map which I haven't been able to code. See here.

This proposed design comprises 16 routes. With relatively so few routes (my proposed design for London is made up of about 70 routes), there are other coding options available, the most obvious being one colour per route. But we are not very far from the absolute limit of this strategy, in the sense that once you get to about two dozen routes, you start to run out of colours. 

* * *

I would say, get the network up and running, and do that as quickly as you can. Or else, why would you not? 

Churchill has said: "The maxim, 'Nothing but perfection' may be spelled 'paralysis.'" Dom Nozzi expressed this idea more forthrightly: "If you are an elected official lacking in courage and leadership, and you face even a peep of opposition to a project, fall back on perfectionism to find the flaws so that you can shoot down the project. Perfectionism leads to paralysis."

Once the network has been introduced, it would be further strengthened on the basis of priority interventions and a timetable. I imagine that some local authorities might want to sort out one or two junctions first, even if the works done serve only as an interim measure, before introducing the network in its entirety. That would be quite understandable. But the essential thing is to make that commitment.

* * *

There are two big advantages with using compass colours. Firstly, as the city expands -- as it inevitably will -- you can add further routes without compromising the simplicity and elegance of the design. Secondly, it is useful even without a map. One comment on the Cycle Lifestyle website reads: "I occasionally park in Hyde Park and cycle to Holborn [...] just know the general direction and take off." 

Thus, a network coded using compass colours gives cyclists the opportunity for spontaneity. The big disadvantage is that it is not always possible to code little bits of route. Still, if we're wanting to create a strategic cycle network, then we can't always be too concerned to code every single bit of cycle route, particularly if they have limited strategic value. 

The cycle route map produced by the City of Cambridge also shows what they call local links. These can be seen here

* * *

I was prompted to start work on a design for a Cambridge Cycling Network primarily in response to the Cambridge Cyclist's questions about who for? I do take his point about Gilbert Road. It seems to me that the hardest decision has already been taken, to wit, to remove the on-street parking. 

Gilbert Road (Photo by Richard Taylor)

A reallocation of carriageway space is self-evidently not going to be enough for some of the people who ride bicycles; they are going to need more than that, and if the local authority is really serious about the bicycle as a mode of transport in its own right, they need to recognise this.

But I would also urge all towns and cities in the UK to have an eye on the bigger picture as well. And that means thinking in terms of a network.

Once you have a functioning cycle network, then you have a solid base, and as anyone who has ever tried to build on sand would tell you, that makes all the difference in the world.


  1. Interesting, thanks for directing me here!

    I agree with some of this, but I'd argue that Cambridge almost has a 'network' for us; so many of our roads do have cycle lanes already. Yet we're not seeing the increase in cycling we need to see. When folk won't cycle down Gilbert Road and up Milton Road to get to the shops, its obvious that more of the same won't seen an increase in cycling numbers. And without that increase in numbers why would we ever see better facilities constructed? Note, this isn't asking for perfection. Its asking for just about good enough.

    Therefore I don't agree that we should be building such a network primarily for those who cycle already, as by definition we're already happy enough to take to the roads and other 'facilities'. From the ground up, we need to re-design the bad facilities we have already, and construct new ones, with the purpose of attracting new riders. Besides, it strikes me that constructing cycle routes badly before then doing it well is a waste of resources.

    But there are ideas here that are bang on - yes, consider cylcling not from the perspective of individual routes, but as a network which has a greater value than the sum of its parts. And your suggested routes seem entirely sensible.

  2. The problem with your suggested network of routes is that there are not nearly enough of them. The Dutch did this research over 30 years ago. They tried city scale experiments of both a sparse network similar to what you have drawn and a much tighter network constructed so that no-one lived more than a couple of hundred metres from a high quality cycle route.

    The sparse network had very little effect because it was only of use to people whose journeys lined up more or less exactly with the routes provided. The tight network on the other hand provided heightened subjective safetyfor all journeys and lead to a permanent increase in cycling.

    For that reason, a high density network of high quality routes remains current policy in the Netherlands.

    And that's what has been happening here ever since. You rarely hear of hyped up exciting sounding ideas from the Netherlands because the Dutch have been getting on with steadily and quietly improving the quality of the infrastructure that everyone uses all the time. For instance, no less than half of the streets and cycle-paths in Assen, where we live, have been improved in the five years that we have lived here. All the improvements have improved conditions for cycling. This is what has lead to the Netherlands continuing to grow its cycling modal share from an already high base despite facing a severe demographic challenge. This is why this country has neither flat-lined at a low level as has the UK nor experienced the same sharp drop in cycling that we see in Denmark.

    At the end of the day it just comes down to spending money wisely. There is no place for doing a bad job and no place for a two track system with low quality infrastructure for beginning cyclists. Those who are least keen to cycle or who find it most difficult to cycle need the very best quality to encourage them.

  3. Thank you both for your comments.

    Gnomeicide, you say that Cambridge "almost" has a functioning network, in which case, it ought not to be too difficult to complete this work. In developing a network primarily for existing users, the authorities would hardly need to spend any money on it. Literally the only cost would be to ensure that the routes are properly waymarked.

    I am not asking for more of the same. I am simply suggesting that, as a first step, wherever you are in the country, networks are introduced to a minimum level of functioning, this being "a prudent course to follow". Once you have a functioning network then you have a solid base, and the task of "re-designing the bad facilities we have already, and constructing new ones, with the purpose of attracting new riders," becomes more manageable, particularly if this work is undertaken "on the basis of priority interventions and a timetable".

    David, you say that my proposed design does not include enough routes. It occurs to me, firstly, that you have not taken on board the fact that I am talking about a strategic network; secondly, that my idiosyncratic use of the terms 'primary' and 'secondary' routes may have misled you; and thirdly, by using colours to distinguish one route from the next, it gives a slightly false impression as to the density of the network.

    Does this map strike you as having the right sort of density? If it does, then I hope you would be comforted to know it describes exactly the same network as this map.

    In truth, this map probably describes much more nearly the most "ideal" strategic network, but the main point is, whichever signing strategy you use, it would be good practice to study the feasibility of the whole thing, rather than simply to try and build a network from the bottom-up.

    To both of you: it strikes you, Gnomeicide, that constructing cycle routes badly is a waste of resources; and for you, David, there is no place for doing a bad job: those who find it most difficult to cycle need the very best quality to encourage them.

    I agree with both of you, and I just ask you to please consider properly what it is that I am actually saying.

    1. Bikemapper, I think you couldn't be further from the truth when you say that it should be simpler to make a wider 'network' thats inoperative good enough such that everyone would want to use ut. It seems much harder to get improvements in cycle infrastructure than it is to get them constructed in the first place; look at the 10 year failure to get a good enough upgrade on Gilbert Road. Once we've got a facility its monumentally difficult to get it upgraded, the perception is that we've already accepted the bad one, this further generates a perception that narrow, poor quality gutters are good enough.

      I would argue that we need to do precisely the opposite; lets stop accepting these marginal gains that are, within our lifetimes, not getting us to where we need to be. As I believe I once stated, facilities like Gilbert Road do represent victories, but they're not tactical victories. Ultimately they hold us back.

  4. Thanks for your reply, Gnomeicide. I do actually agree with a lot of what you're saying, but I feel as though we are not properly understanding each other.

    Let me put it this way. If you wanted to build a house, once you start the construction, you would lay the foundations all at once, right? All I am saying is that, in developing a cycle network, you would do exactly the same thing. That's all. From this solid base, you would then build upwards.

    I am specifically not saying that a network introduced to a minimum level of functioning would be good enough for everyone to be able to use it, but ultimately, of course, that is what I would like to see happen.

    I would also argue that the introduction of a functioning cycle network is a tactical victory. (To be clear, there isn't a town or city in the UK that has a functioning network, although I am quite happy to accept your view that Cambridge "almost" has one.)

    In the case of Cambridge, once the network is up and running, the authorities should put in place a timetable. Within five to ten years, say, all of the routes on this network should be able to be used by, for example, an unaccompanied twelve year-old child.

    As regards Gilbert Road, I think it would be possible for the authorities to revisit this scheme and provide properly segregated infrastructure.

    If you would like to discuss this further, please write to me at I am sure we can work this out.

    1. P.S. I have just read Freewheeler's latest blog. He writes: "A genuine Cycle Action Plan should, to my mind, have two basic features. First and foremost it needs to build a network of primary segregated cycle track routes across the borough; secondly it would close all residential areas to rat-running."

      These primary routes would be "safe, convenient, attractive segregated cycling routes (which obviously would require priority over all side roads and dedicated cycling-only green phases at major junctions)." I wholeheartedly accept this, but I would just simply ask, unless this work can be completed quickly, what should happen in the meantime?