Lambeth Bridge shows us that TfL still needs a fundamental shift in design philosophy

Catching up with the latest TfL consultations, I was a combination of delighted and exasperated to see the proposals for Lambeth Bridge.

Delighted, because they reveal the political will to provide for cycling and walking, even if it means sacrificing a tiny bit of motoring capacity at these junctions.

Exasperated because even when the will is there, the technical capacity clearly isn’t. Just look at it.

Click to download fullsize diagram from TfL (PDF)

Look at that splurge of tarmac.

TfL had a blank canvass here. They’re completely rebuilding the junction, which is currently a roundabout. The kerblines are going, the trees are moving, the drainage will have to be redone. It’s not like any compromise is being made to work within the constraints of what’s there now.

It reveals a lot about how TfL’s designers are approaching these problems. They had a blank canvass and they filled it with tarmac. Then they asked where they needed to fit some kerbs into it. Walking and cycling is still being designed to fit around the edges of motoring, which remains the natural rightful state of our streets.

And look at the absurd mess that has led to.

Look at the entry to Lambeth Bridge. They’ve painted an arrow on it to tell motorists to get out of the cycleway. How on earth did that happen? According to the diagram, there is only ever one lane of traffic able to head onto the bridge at any time. How did they get themselves into a situation where people might be driving in the cycle lane and need to be told to get out of it?

Look at the entry into Albert Embankment. What is that lane merge arrow painted on that grossly distended carriageway space for? Again, there is only one lane of traffic entering this street at a time.

These kinds of tarmac oceans, with arbitrary shifting kerblines and unclear routes through them, are why British junctions are confusing, stressful and dangerous — whatever mode of transport you’re using. There is a large area of tarmac here that in theory should never ever be used, even by the largest and longest vehicles, because there are no legal manoeuvres that would use the space, or at least, that wouldn’t normally and more safely and sensibly be performed using some other space. But what that tarmac does do is provide opportunities for people to make mistakes, to perform illegal manoeuvres, or to behave dangerously. The gaping wide entries into Lambeth Bridge and Albert Embankment do nothing for a law-abiding motorist except create confusion about where they’re supposed to be, but they do invite idiots to use the turn lanes and cycle spaces to jump queues or to make illegal turns.

What might this junction look like if designed with a less motor-centric philosophy?

Instead of washing an undercoat of carriageway all across the blank canvass, you’d start from the opposite default. The canvass is blank and the only carriageway you add to it is the carriageway that’s needed to accommodate the manoeuvres that are possible here. Who are the users who need to move through the junction, and where do they need to move from and to? Draw the paths that they will need to take through the junction.

Draw the paths that motorists would take through this space if following the turn restrictions and if following the most rational routes. Work out how you’d cycle around it, and how you’d walk around it. The result looks a lot like Dutch crossroads do.

Click to embiggen. (This is just a 5 minute scribble, exaggerated in places, to illustrate the concept. Please don’t tell me the turning circles are too tight for HGVs or the cycleways too narrow or there are white lines in the wrong place, I’m not trying to propose this as a working plan for the contractors.)

This kind of design enables all of the same turns that TfL’s design legally allows, and has exactly the same number of motor traffic lanes feeding into and out of each arm of the junction. Capacity for legal manoeuvres should be the same (but the illegal turns, lane misuse and simple mistakes encouraged by TfL’s original design become much more difficult). I’m not saying necessarily that this is what I’d want the junction to look like, because I wouldn’t necessarily accept the same demands of motor capacity as TfL, but if you do accept the demands TfL are working to, this is roughly how the junction would work if cycling were properly designed for.

The one impact I haven’t addressed with my crayons is that they would have to address turn conflicts across the cycleways — something that wouldn’t be an issue if we had normal priority rules. But there are multiple options available to solve that problem TfL could call upon — indeed, they already solve it in their own design, for traffic/cyclists exiting Albert Embankment, with a separate turn signal.

The cycle lobby: Andrew Gilligan messes it up

Andrew Gilligan accuses “the cycle lobby” of thinking only of themselves and not “putting themselves in the heads” of non-cyclists. In-fact, failure to think as a non-cyclist is exactly why the policies of Boris Johnson are such failures.

Despite the “cycling mayor” image he encouraged early on, after four years in City Hall, Boris has been getting a beating from folk who cycle in London. His flagship scheme for cycling was meant to be the Cycle Superhighways, intended to “transform” London, “boost safety” and — independently of all other initiatives — contribute to modal shift to the tune of 120,000 more daily journeys:

“I’m not kidding when I say that I’m militant about cycling, and these Superhighways are central to the cycling revolution I’m determined to bring about. No longer will pedal power have to dance and dodge around petrol power – on these routes the bicycle will dominate and that will be clear to all others using them. That should transform the experience of cycling – boosting safety and confidence of everyone using the routes and reinforcing my view that the bike is the best way to travel in this wonderful city of ours.”

Kulveer Ranger, said: “Cycle Superhighways form a key part of the Mayor and TfL’s target to increase cycling in London by 400 per cent by 2025, compared to 2000 levels. From cycling the proposed routes myself, and speaking to a whole range of cyclists, I’m sure that these routes will prove a hugely welcome addition to London’s cycling infrastructure – giving many more people the confidence to ride”.

But this hyperbole soon backfired on Boris when it turned out that the Superficial Cycleways were, except for sections of existing dedicated infrastructure taken over on CS3, little more than £100 million paint on the road — paint that dances and dodges around petrol power, does nothing to transform the experience of cycling on the capital’s busy arterial roads, and does nothing to boost the confidence of the would-be and wanna-be cyclists that Boris claimed would be attracted by the novel hued bike lanes. Although TfL have been able to claim that there has been a large increase in bike traffic on the Superhighways, they don’t really appear to be doing much to enable or encourage non-cyclists: at most, some existing cyclists have been tempted out of the backstreets and onto the main roads; few new cyclists have been created. The most common question Londoners have about the Superhighways is: are they joke?

Since people started dying on his Superhighway at the Bow junction on the East Cross Motorway, Boris has taken the emphasis off the dozen radial routes which were once “central” to his cycling revolution, and when he does talk about them these days he will tell you that the blue paint is a navigational aid — no mention of excluding “petrol power”, boosting safety, or transforming experience. What were originally sold as part of a cycling revolution which would enable and encourage people to take to their bikes have turned out to be, at best, something to help existing cyclists find their way to the square mile.

This is why Boris has failed on cycling: he’s trying to drive a cycling revolution — more people cycling for more of their journeys — by providing for existing cyclists. Hilariously, Gilligan is so clueless about the substance of the disagreement between Boris and “the cycling lobby” that he attributes this problem exactly backwards:

“Cycle lobbyists need to put themselves in the heads of a non-cyclist or politician most of whose voters aren’t cyclists, asking why we should arrange the streets for the 2 per cent who cycle rather than the 98 per cent who drive or take the bus.”

Go Dutch, and The Big Ride, are precisely the product of the London Cycling Campaign “putting themselves in the heads of non-cyclists”, and calling for streets to be arranged for the 98 per cent who currently would never dare to cycle on them. The Go Dutch campaign was squarely pitched at the non-cyclist, showing everybody how, with a determined leader, London’s busy roads could be transformed into places where anybody and everybody can use a bicycle, and share in all the benefits that come with cycling. Gilligan seems to think that the campaign and ride was a demand by existing cyclists that they must be pampered and privileged in their niche activity. Far from it. The point that The Big Ride made was that the “cycling revolution” that Boris Johnson promised will not be delivered so long as he continues designing cycling policies and “Superhighways” for the 2 per cent who already cycle. Indeed, many of those who rode with us on Saturday are, on any normal weekday, part of the 98% themselves.

As part of the two per cent willing to — no, no, as part of the one per cent happy to — cycle on the streets of London as they are, Boris is the last person who should be appointed to lead a “cycling revolution” aimed at enabling the 98 who don’t cycle to take it up. He boasts that “scooting down Euston underpass” and around Hyde Park Corner are “no problem” when you’re “used to it”, and his now infamous comments about the Elephant and Castle being “fine if you keep your wits about you” tell you everything about how far he has penetrated the minds of ordinary non-cycling folk.

Boris’s “cycling revolution” seems to be designed around the premise that there is a large population of Londoners who are just on the cusp of taking up cycling and who just need lessons in “keeping their wits about them”, or blue paint and hire bikes to help them to “get used to it”. Boris understands how his 2% cycle so he designs policies for more of it. But the conclusion of last year’s Understanding Walking and Cycling project (admittedly primarily based on research in England outside of London) was that there is no such substantial section of the population just waiting to take up cycling in traffic, ready to be nudged in by one cheap and simple little thing. The Understanding Walking and Cycling project — which has informed and given urgency to infrastructure campaigns like Go Dutch — “put themselves in the heads of non-cyclists” and found that the 98% will not cycle so long as they expected to keep their wits about them and get used to the Euston underpass. There are very few waiting to join the 2% cycling in heavy and fast traffic: if you want a cycling revolution, you have to try something new and different. The 98% look at the policies of the Cycling Mayor and see irrelevant “Superhighways” which they presume must be good for Cyclists but on which they would never dare to cycle themselves. They look at Go Dutch and see civilised dedicated space on which they might. And Gilligoon thinks it’s the latter who are out of touch and appealing to the minority on cycling.

Boris even came close to showing signs of understanding all this when he talked of not having to “dance and dodge around petrol power”. But like so much about Boris, that turned out to be all waffle and no substance.

The problem with Boris and his cycling revolution, and the many reasons why he has messed it up on cycling, obviously go far far wider and deeper than his inability, as a contented member of the 2%, to understand why the 98% are so reluctant to join him. But I’m not sure I can bring myself to write about, or even think about, it any more. Please, just make it stop.

Pick and mix and the Hierarchy of Provision

The marvellous Mad Cycle Lanes of Greater Manchester has posted an extract from a history of provision for cycling in the Netherlands:

the government decided to develop a large array of measures to promote cycling, walking and traffic calming, such as:

– Reducing car access to city-centres and create car-free areas;
– Making parking in city-centres more expensive;
– Constructing cycle paths and reducing road space for cars;
– Facilitating cycling through cycle network planning, road design, signalling, parking and enforcement;
– Reducing maximum speed on the majority of urban roads to 30 km/h or less;
– Promoting cycling to encourage the use of bikes and discourage car-use.

But MCLGM draw a rather odd conclusion from this, which is not supported by the quoted text:

Note how most of the measures are about reducing and removing access by motor vehicles.

The key to increasing cycling is reducing motor traffic, the cycle tracks are the follow-up measures.

I tried to post a comment, but of course because MCLGM is hosted on Blogspot it disappeared with a server error page somewhere in the gazillion-step commenting processes. (Seriously Blogspot bloggers, it is way past time you came and joined us on a platform that works.)

I’m going to post the comment here, because it seems to be a very common mistake, and one that needs to be laid to rest.

Er. Not quite. All of the measures are “key”; none of them are “follow-ups”. They are the different solutions that apply to different situations. The British seem desperate to put things in a hierarchy. It’s completely the wrong approach, and it’s certainly not the Dutch approach.

In some places — residential streets and city centres — we need to reduce traffic speed and remove vehicular access. In other places — main arterial roads and other places with high traffic volumes and/or large vehicles — we need to reallocate space for cycle tracks. You start from the specific problem, not the preferred solution.

That is the key to increasing cycling. Knowing what needs to be done in each situation.

I’ll use this opportunity to expand on this point about putting solutions into hierarchies a little.

In Britain we have a “Hierarchy of Provision”, which recommends some types of cycling provision as preferable to other types. It was developed by the DfT and CTC in 1997, is still endorsed by many cyclists (sometimes enthusiastically so) as well as officials, and is part of the design guidance for cycling infrastructure — Local Transport Note 2/08 — upon which the nation’s Crap Cycle Facilities are modelled.

Specifically, the Hierarchy of Provision states that one should “consider first” reducing the speed and volume of motor traffic, and “consider last” shared use footways. (Note, not: “never ever consider”, the Hierarchy of Provision thinks shared footways should be “considered last”.)

This is, as my lost comment says, approaching things the wrong way around: bringing a set of pre-ranked preferred solutions to a road and trying each one in turn to see which one fits. The correct approach — the one that the Dutch apply — is to start with the purpose and properties of a road: whether it is the main A-to-B road, or a little residential or access street; whether it needs to carry big dangerous trucks and buses; and so on. Once you’ve answered those questions, there is no need to try different solutions on for size: when you understand the problem, the appropriate solution follows.

For many years in the UK there was a Hierarchy of Provision way of thinking, which led, and still leads, to some absurdities, such as the idea that in the Netherlands cycle tracks are a “follow-up measure”. When the Cycling Embassy of Great Britain came along and made the then heretical statement that, actually, for some kinds of roads, we should be asking straight-up for Dutch-style cycle tracks instead of trying to apply one preferred top-of-the-hierarchy solution to every road in the country, this entirely sensible position was often misunderstood. Unable to think outside The Hierarchy, the embrace of cycle tracks was interpreted as an attempt on the crown: to put cycle tracks at the top of The Hierarchy and to campaign exclusively for the construction of segregated cycle paths alongside every inch of road, lane, street and cul-de-sac in the country.

It seems that this misunderstanding is still spread, and by those who should know better:

@carltonreid Carlton Reid
135,575 km of roads in Netherlands. 29,000km of segregated tracks. So, in UK, there’s a need for separation on *every* road? #cyclesafe
7 hours ago

@jonsmalldon Jon Smalldon
@carltonreid Nobody has said or is saying that though?
7 hours ago  Favorite Retweet Reply

@AsEasyAsRiding Mark
@carltonreid Who’s saying that?
7 hours ago Favorite Retweet Reply

@carltonreid Carlton Reid
@jonsmalldon @AsEasyAsRiding David Arditti coming very close. http://bit.ly/yf7JwZ
6 hours ago  Favorite Retweet Reply

(Readers can follow Carlton Reid’s link that to the essay in which David Arditti explains very clearly and at length, in 3,000 words and two diagrams, his ideas on the interventions which might be appropriate solutions the problems that cyclists face in different situations, placing it in the context of David’s other articles, and can draw their own conclusions about whether Reid is right to sum up David’s argument as a call for segregated cycle tracks on every public road and street in Britain.)

Campaigners seem to be growing more comfortable advocating the full breadth of solutions in the Dutch “Sustainable Safety” package, but even when doing so in the Xmas edition of their magazine, the London Cycling Campaign can’t stop thinking in hierarchies: “the Dutch mantra is mix if possible, separate when necessary,” they say, as if the Dutch have a hierarchy of “consider first: mix”, “consider last: separate”. But as the LCC magazine says elsewhere, the Dutch “start by deciding the function of a road, and from this flows the design.” On a busy trunk road they don’t first consider mixed traffic.

But not everybody has quite come to terms with it all yet, and in attempting to reconcile the comprehensive and detailed Dutch solution with the crude and uncomplicated British hierarchy approach, it seems that many have had to resort to what MCRcycling calls the Dutch pick and mix: we’ll pick the filtered permeability to put at the top of our hierarchy, and campaign for that one, thanks.

Over the past couple of years, the breadth of the successful Dutch approach to street design and cycling has begun to be explained and appreciated in this country, thanks initially to the likes of David Hembrow’s blog, Mark Wagenbuur’s videos, and occasional articles like this, and recently spread even further by the Dutch Cycling Embassy and London’s Go Dutch campaign. The activity stirred up by The Times CycleSafe campaign will, I hope and expect, lead to its much wider dissemination. I imagine that attempts to arrange the continent’s engineering into a hierarchy of provision will keep people occupied for a while yet.

Tour du Danger: Waterloo

This is part of a series gathering thoughts on what is wrong with London’s road junctions.

If the junction of the A301 Waterloo Road/Waterloo Bridge and the A3200 York Road — the BFI IMAX roundabout — were just a roundabout it wouldn’t be very interesting. It probably wouldn’t be quite so dangerous for people on bicycles, either — not that the standard British design three-lane large-radius roundabout is at all safe or inviting, it’s just that this junction is made extra special by the Waterloo Bus Station beside the railway arches:

The bus station is on the northbound side of Waterloo Road, and the northbound section of the roundabout is, perhaps uniquely, split in two, both in space and time: the buses exiting the bus station have their own phase of the lights, and their own dedicated carriageway for this quarter of the roundabout. In amongst the sprawling traffic traffic lanes and bustling bus station, a few scraps of stop-start “cycle facility” are provided.

Heading north on a bicycle on Waterloo Road, when past the left-turn into the bus station, one is given the option of a tight-left turn into a narrow fully segregated unidirectional cycle track (best illustrated by this architect’s drawing of a hypothetical bus shelter). There is no such cycle track for bicycle users entering the roundabout from any of the other arms, but they are provided with a way into this cycle track once they have negotiated the roundabout all the way past the Waterloo Road arm. Except that by then the cycle track has already run out and instead become, for the final few yards, an on-street cycle lane, within the bus station, on the right-hand side of the buses. That leads you up to the advance stop box and the traffic lights in this video.

I can’t think of any reason why anybody would use this facility, except out of cautious lack of familiarity with the road, in the mistaken belief that a cycle track will safely lead them to where they need to go. The timing of traffic lights on the roundabout and on Waterloo Road are such that if you chose to use the facility then the traffic lights for the exit from the bus station will almost always be turning red a few seconds before you reach them, and if they’re not, well, that’s even worse — you’re deposited in a narrow and soon to expire lane on the wrong side of a line of buses all racing to get through the short cycle of the lights.

And see how they race! In fact I observed a few rounds of the lights before I erected the conspicuous camera and tripod, and I’m sure the drivers must have noticed when they were being filmed, for, before the camera was set up, on every single round of the lights a bus would crawl all the way to the front of the advance stop box and the drivers, clearly intimate with the signal timings and watching the behaviour of the neighbouring traffic streams, were experts in setting off a full second before their own light turned green — every time. You wouldn’t want to be the on a bicycle directed up the narrow lane on the right-hand side of those.

There are several fundamental problems with the ideas behind this junction. One is that it’s good to put bicycles and buses together. British engineers are told by the DfT’s cycle infrastructure guidelines that cyclists like sharing with buses and so engineers should plan them into the bus spaces. But to say that cyclists like sharing with buses is either a misinterpretation or misrepresentation of the research — a survey which found that existing cyclists preferred lanes where they only had to deal with buses over general traffic lanes where they had to deal with buses and fast cars and big trucks. (The same survey also found that cyclists and bus drivers have a low opinion of one another.) Bus lanes are less awful than no bus lanes; that doesn’t mean that most people like riding bicycles in them or that they can bring about mass cycling, and it certainly doesn’t mean that a bus station, where a lot of buses are stopping, waiting, and pulling in and out of tight spaces, is a safe, sensible or attractive cycle route.

But the most fundamental issue is perhaps that both the cycle facility and the bus station is weird cheap improvised one-of-a-kind crap stuffed in where it won’t get in the way of the very important people who drive cars and hire cabs in central London. The de facto hierarchy and prioritisation of motorised modes is a familiar problem, to the point that it is barely interesting when considering this junction. Rather, it’s the cheap improvisation that makes this one stand out. These weird ad lib facilities, which stop-and-start, merge and diverge, and abandon you in unexpected places, apart from being unattractive and unlikely to be much help in bringing about mass cycling, make everybody’s behaviour unpredictable, and that leads to mistakes being made.

Certain British cyclists look at the Netherlands, see it all working smoothly, and conclude that the laws must be beating everybody into good behaviour — the fear of insurance claims under “strict liability”, perhaps. But they’re overlooking the many ways that the Dutch control behaviour through engineering — not merely physically preventing bad behaviour by designing out speed, but also engineering out mistakes by making things obvious and predictable. Roads are built differently to streets and lanes, for example, and it is therefore obvious which you’re on and what is expected of you. Roundabouts in particular are made to be predictable places: there are few designs, with minor variations between them. People on bicycles get their own dedicated space, everybody understands that this is so, and it is made very obvious where bicycles and motor vehicles could come into conflict and which gets priority in those places.

British roundabouts follow no such rules. Some of them are a single lane, some of them two, three, four, or even five or six, depending on how much space the engineers had to play with. Sometimes there are lane markings, sometimes these are concentric circles that you veer across as you proceed, sometimes they spiral around to carry you all the way to your exit, and sometimes lane markings come and go several times in the course of your gyration. Sometimes there are traffic signals on the roundabout, sometimes there are traffic signals only on the entrances, and sometimes a roundabout will have a mix of signalised and non-signalised segments and entries. Sometimes there are signalised pedestrian crossings, sometimes there are informal traffic-island crossings, sometimes there are zebra crossings set back by the statutory distance, sometimes there are underpasses, and sometimes there is nothing and nowhere for people on foot. The huge variety comes from the obsession with eking out every last drip of traffic flow capacity at the expense of safe and intelligible standardised layouts. And the result is confusing and stressful enough for users, even without their having to worry about what sort of bizarre ad hoc cycle facility hack has been woven around the edges.

If you want people to ride bicycles where there are big, fast, complicated roads like this, you need a cycle track, but a proper one, put in the right places, continuous and predictable.

Under Ken Livingstone, when street space was still being reclaimed in the post-CCharge introduction period, a new design for the IMAX roundabout was proposed.  There was one of those world-of-their-own architect’s mockups, big on shared space, which imagines that a stone surface makes trucks disappear and leaves a big plaza full of happy pedestrians (and, even more bizarrely, imagines a totally new IMAX building). South Bank organisations are still promoting the plans (passively, at least, through a website last updated two years ago), but there can’t be any chance of anything changing here with Boris in city hall.

Do you have any observations of the Waterloo junction to add to the Tour du Danger dodgy junctions dossier?

On country lanes

This post is part of a series, starting at “So what would you do here?”, on making utility cycling attractive in rural Britain.

I was sent a CPRE book last week. It reminded me of these lines from Betjeman:

Let’s say goodbye to hedges
And roads with grassy edges
And winding country lanes;
Let all things travel faster
Where motor car is master
Till only Speed remains.

A common objection to dedicated cycling infrastructure is the size of the British road network. You can’t put cycling infrastructure on every road in the country, therefore you shouldn’t put it on any. I disagree, not just with the conclusion, but with the premise too. We can put cycling infrastructure on every road in the country. It just depends on the definition of “road”.

Country lanes like the one below, a narrow 4km long link between the north-south A357 and the east-west A30, are currently “roads” — used primarily to get motor vehicles from one place to another. How could you separate bicycles and motor vehicles here without demolishing buildings, cutting down hedges and paving over the countryside?

It starts as a little street squeezed between two old shops in a tiny town centre:

On the village high street

This post is part of a series, starting at “So what would you do here?” and On rural main roads, on making cycling an attractive mode of transport in rural Britain.

Big busy fast main roads are the major barrier to making journeys by bicycle for most of the rural British population, and proper high-quality cycle tracks are the most plausible solution to that problem. It would be nice to be able to reclaim all the roads from the fast cars and big trucks, make them places where people can happily ride bicycles again, but that doesn’t seem likely for a long time. At the moment, providing dedicated inviting space for bicycles alongside them is the only proven solution.

In some cases the same solution could be applied to rural settlements: enabling cycling by reallocating some of the roadspace from general traffic to bicycles, as in this village on a busy road and intercity bicycle route between Arnhem and Nijmegen in the Netherlands:

But more often the reverse is true of rural villages. Not only is it difficult to provide cycling space because of the constraints of very old streets, but the political will is the opposite within villages than without: speed is unpopular and most residents want the cars tamed and the streets reclaimed for people. It is just about possible to calm these streets, with the right engineering, and the will to do it exists — not because of anything to do with cycling, but because people live and work and shop and raise their kids in these places.

That’s part of the Dutch model, where they recognise that roads, whether we like it or not, have become routes for motor vehicles, while streets must be places for people but where, with the right engineering, a limited number of motor vehicles can be accommodated.

On this country road near Assen the speed limit drops to 20mph when it enters a village…