Wednesday 17 July 2013

CS2 At Aldgate East - Just One Junction.....

The death of Philippe De Gerin-Ricard on London's Cycle Superhighway 2, near Aldgate East Station on the A11 Whitechapel Road, has been covered by many others, but what will it take to make the area safe?

As this death is still under investigation by the Metropolitan Police's Road Death Traffic Unit, I am not going to postulate on how it happened - I have my suspicions and I know the police do. But, as I do cycle through this area every so often, I can confirm that it is intimidating, scary and not a place to cycle for the fun of it. I have a few ideas which I think would make things safer, but it will mean a reduction in traffic capacity.

The only three-quarter decent bit of CS2 which is westbound, just
west of the Bow roundabout. You are protected from traffic, but there
needs to be a little more protections for pedestrians and the width of
the cycle track is a little tight when busy.
Image from Google Streetview.
CS2 currently runs from the Bow Roundabout and follows the A11 before fizzling out at the Algate Gyratory. As it happens, my usual journey into The City carries on through the gyratory westbound to Fenchurch Street and then over London Bridge. 

But, my return journey is always via Tower Bridge and so I miss out the eastbound gyratory without having realised it before; coming back via Mansell Street (one-way northbound), although the right turn back onto the A11 means I have to cross 2-lanes of left turning traffic first. In essence, this part of The City is a complicated mess of wide one-way gyratory systems which are geared to moving traffic around quickly - that is another discussion altogether.

CS2 at Bow Road going under the "Ferodo" railway bridge. Not even
cycle lanes here and the blue stripes are squeezed next to the bridge
piers. The central hatch area is to create a right turn pocket into
Addington Road which could be accessed from many other locations

and could be closed off. 
Image from Google Streetview.
CS2 is not fun to use. Apart from a tiny bit of cycle track at the Bow end, the whole route is a mixture of advisory cycle lane, bus lane, blue stripes and the odd cycle logo on a blue background. Wherever there is a narrowing, the cycle lane goes and we are just left with blue paint which has nothing going for it in terms of cyclist protection, although I do concede it stops you getting lost.

So, when one travels along CS2 (apart from the nice first bit) one is constantly bugged/ intimidated by traffic passing closely when you are following the blue stripe; one constantly overtakes and is overtaken by buses; there is the inevitable conflict at junctions when wanting to go straight ahead or right because left turning traffic puts you in conflict when going ahead; and right turns often mean crossing two lanes of traffic (yes, it is pretty much a dual carriageway!). There are advanced stop lines at the signalised junctions, but they are only really of use for those wishing to carry on ahead or turn left and that depends on when you arrive in them.

Mile End Road, westbound approach to Grove Road/ Burdett Road
junction. Do you fancy turning right here on two-wheels?
Image from Google Streetview.
This on-carriageway arrangement carries on along the whole route and in both directions. The route (which opened in 2011) doesn't even follow the massively out of date London Cycle Design Standards which was published in 2005. The scheme was rushed in and it is absolutely clear that the brief was that the scheme should not affect traffic flow at all. Mind you, this kind of approach can be found all over London, not just on CS2.

Of course, those with a memory which can cope with a year's passing will remember the little sports day we had in London last summer. Part of the deal with the Olympic movement was that London would shove in some VIP lanes for the Games people. There was outcry and a massive amount of planning, but in the event, the traffic disappeared and it ran very well. This proves that it is possible to make decisions to remove traffic lanes when it suits.

A11 westbound.
Image from Google Streetview.
So, back to the scene on the death. The junction in question if the A11 Whitechapel Road where it meets the A1202 Commercial Street and Leman Street. The junction is signalised with staggered pedestrian crossings (green men) on all arms, except Leman Street which has a straight across green man.

The two A11 approaches have 3 lanes; left/ahead, ahead and right. There are no cycle lanes, but Advanced Stop Lines (ASLs) are in place on all arms except Leman Street which is a two-lane one way street going away from the junction and for the A11, there are blue stripes running through the junction itself (which are seen all over London). 
In Commercial Street looking over at the two lane Leman Street
which is one way going away from the junction. The City is to the
right and Stratford to the left.
Image from Google Streetview.

Commercial Street has two lanes, left and straight on with the right turn banned, except for buses, taxis and cycles. There are yellow boxes in the junction and so coupled with the one-way Leman Street and Commercial Street's banned right turn, this is a classic attempt to maximise traffic flow. Oh, all the traffic lanes are narrow which is another symptom of capacity problems.

If we go back to the basics of the process used in casualty-reduction schemes, the highway engineer is looking to reduce conflicts in space and time. By this, I mean that different traffic movements are given their own space (lanes) or conflicts are separated by time (green man versus green traffic signal). If we move beyond casualty-reduction to casualty-prevention, we need to be looking at how people need to pass through the junction and separate out the hazards in space and time from the outset.

Leman Street outlined in red with the "yellow" main roads picked out
in green. Leman Street is actually part of a larger network bounded
by really main roads which are very wide and should be taking the
traffic through the area.
Image Google Maps.
The first thing to look at is Leman Street. It forms part of a larger network of one-way and two-way streets which are bounded by the A11 Whitechapel Road, A13 Commercial Road, A1203 The Highway, Butcher Road (Limehouse) and the Mansell Street/ The Minories and Aldgate gyratory complex. In short, a white road in lots of yellow roads on the map. This area has mixed usage such as dense residential, commercial, local shops, schools and indeed part of Cycle Superhighway 3 on Cable Street. The obvious answer is to close Leman Street and make the junction 3-arm which removes several movements and conflicts from the off. There are other crossroads within the larger area and so these need to be made 3-arm, but on the opposite side (if that makes sense) as we still need access into this local area.

Now we have an area without a side road, a protected cycle track can be built which bypasses the traffic signals and provides a link into Leman Street (filtered permeability). The Commercial Street arm is more tricky. We might consider traffic turning left and right again as the ahead movement is gone and other junctions will have a knock on back to this if they are reworked. The cycle track can allow all movements (and a connection to Leman Street), but the issue is dealing with pedestrians. 

We are not allowed to use zebras with traffic signals and besides, a zebra over the cycle track and then a pelican over the road is confusing for pedestrians. We are also not allowed to use any flashing amber signals to allow modes to mix. We could use toucans (with separation) to cross pedestrians and cyclists, but if we give an all round green, there is conflict and do cyclists on a superhighway really want to use a push button? 

This is the bit I struggle with in terms of converting European arrangements to the UK. I am sure another blogger will know how cycles are detected elsewhere, but UK cycle detection is always rubbish, but I think we need some narrow field detection on the cycle track in advance of the junction which may trigger demand or we go for fixed timings which are improved and changed as time goes on and hopefully cyclists make up a greater proportion of movements.

One possible layout, but things are getting complicated!
Anyway, I have had a go at trying to draw up an idea of how the junction could be reworked. It is fairly idealised, but I think it can be squeezed in as the layout does not have the staggered pedestrian refuge areas in the pelican crossings. Blue is CS2 separated from pedestrians by a kerb, with a hard strip between CS2 and the carriageway (somewhere to stick traffic signals?). Green is a "local" cycle track (protected) and buff is footways. The red 'L's are pelican crossing points. I have not shown any cycling road markings.

Leman Street is closed, but allows cycles to turn left in and out, with a speed table to assist pedestrians crossing the cycle track as it is an uncontrolled crossing point. Westbound (towards The City) maintains 2 ahead lanes and has a right turn lane. Eastbound maintains 2 lanes with the left lane being ahead/left. In terms of signal stages for traffic, this would run with 2; the first is the A11 arms running together with eastbound getting a red before westbound so that right turners can clear the junction. Then Commercial Street would run. Right turns by cycles would be two-stage, so for a right turn into Leman Street, one would cross Commercial Road and then the A11.

I have assumed that pedestrians would get a green man as a single stage allowing all 3 arms to cross at the same time. With cycles, it gets trickier and I have not got to grips with the sequence yet, but cycles could all get green at once (using low level signals?) and have to give way to each other through the junction, or there could be pedestrians and cycles running together, but over two stages; first A11 cycles would run when pedestrians cross Commercial Street, but have a cycle red on Commercial Street. Then Commercial Street cycles can exit with pedestrians crossing the A11 on both arms, but cycles are held on the A11. Of course, in the second arrangement, cycles would get a green on cycle routes not crossing a road when traffic runs.

And this is the point of the blog. This is just one junction on one direct route into The City. The A11 has lots of junctions like this to deal with and I am not sure we can maintain the two traffic lanes in each direction and have right turn pockets at each location. If we lose traffic lanes, cyclists and pedestrians might gain, but bus passengers won't. All of this needs to be decided and it starts with a policy; that is, a plan agreed by politicians and not just the Mayor.

We are at the end of the era where paint 'n' signs make a cycling scheme, we need to build stuff with the same vigour as we used to have for catering for private cars (and may do again soon with the current lot running the show). We will have to accept that this work needs the kind of budgets spent on revamping town centres, station forecourts and the like and this is not going to be isolated junctions, this will be entire routes and will need serious investment - think about schemes on a scale like Exhibition Road, but everywhere and for years. In terms of removing traffic capacity - should anyone need to be driving so close to the centre of London other than those moving goods or many people?

There will be more deaths, more protests and more debate before things start to change, but in my view, unless we are radical, we may as well use the paint to squeeze even more car lanes into what space is left. I think the Mayor should spend less time on planning his island or looking to bury roads (Really? Does he know how much tunnels cost these days?) and think strategically on cycling. 

I will leave you with a view of Commercial Street and the question: what will this area look like in another 100 years?

Commercial Street, just a little way back currently and c.1907.
Over 100 years the tram was clearly the transport mode of choice.
Image from Google Streetview and Wiki Casebook.


  1. In you photo of the Ferodo bridge, the right turn pocket is actually for Addington Road I think, not Kitcat Terrace, which is before the bridge. Addington Road is not a dead end, and could be closed off to Bow Road

  2. I agree with the general thrust of your blog, which is that roads such as the A11 need to be put on a 'diet'. As you say, "I am not sure we can maintain the two traffic lanes in each direction and have right turn pockets at each location." I fully concur.

    However, the fact that Leman Street is marked on the map in white, and Minories in orange, is a bit misleading.

    According to Google maps, Leman Street / Dock Street is actually the A1202, whereas Minories, though nominally the A1211, is more like a side road. The only motor traffic which routinely uses Minories are buses.

    In practice, the gyratory which links Tower Bridge with Commercial Street works as per this map.

    Regarding your point that we are at the end of the era where paint 'n' signs make a cycling scheme, it's a shame we didn't first get to the point where a comprehensive, city-wide cycle network was made to function.

    It's either network first, and then a separation of functions; or isolated pieces of quality infrastructure first, and then join up the pieces. Top-down or bottom-up? It's pointless asking you to supply any evidence that the bottom-up approach would work better, since there isn't any. So we'll just have to agree that one approach is evidence-based, and the other approach is informed by ideology (i.e. wishful thinking).

  3. We do appear to be stuck between a rock and a hard place in the UK when it comes to traffic controlled junctions. As you say, if we want/need to have two straight on lanes on our main routes, then we're going to be pushed for space to fit in dedicated turning lanes that provide the desired temporal separation of motor vehicles and bikes/pedestrians.

    The Dutch system when it comes to designing these junctions says, start with a big junction and if it doesn't fit, make some compromises by allowing sub-conflicts where the risk is acceptable or by limiting traffic movements (closing junction arms, restricting turns, introducing one-way).

    The problem we have in the UK is that our system doesn't cater for allowing sub-conflicts between motor vehicles and bikes/pedestrians. As you say, we're not allowed orange flashing arrows or combined zebra and pedestrian lights (and we have no concept of a bicycle zebra).

    I'm not sure if there is a solution that fits within the UK framework. Without allowing sub-conflicts the only solutions are:

    * Limiting traffic turning movements
    * Restricting the number of through lanes to provide space for turning lanes
    * Separating motor and bike/pedestrian light stages temporally rather than spatially (bikes go on motor all red and vice versa)

    None of these are a very satisfactory solution for every situation. Without allowing sub-conflicts we're always going to have a sub-optimal set of solutions to pick and choose from, but it's going to be a tough nut to crack.

  4. mjemmeson - quite right, I have updated the comment. Thanks for pointing it out. My usual journey is a blast along the A11 and I don't venture into side roads that often!

    bikemapper - the whole area is a mess of one-ways and gyratories and the 'A' road classification is pretty arbitrary. This is purely some thinking on my part about an area I don't know well (other than the junction) and one potential reworking of a junction as a 3 arm junction is easier to deal with than a 4-arm.

    In practice the whole corridor and indeed parallel corridors need looking at. Evidence is always difficult, but the old LCN approach did not increase levels of cycling much as it concentrated on A-B routes and not how other side routes fed in or out (the dense filtered permeability).

    Some early figures from a scheme built under the Sustrans Connect 2 banner (cannot tell you where as I worked on it!)is showing an increase in trips using the route (19,000 new trips attracted to the route projected for 2013 compared to 2009. A modest start, but cycle count data shows the increase in trips and face to face interviews suggest that the sections of the route which are away from traffic either through parks or on cycle tracks are very popular.

    We need the dense networks, but we need direct protected routes as well and my post was to explain the difficulties at one junction, let alone a whole route or indeed city!

    Paul James - All you state is spot on. The only ideas I have (which need a change in law) are cycle signals (see my recent post!) or perhaps zebra stripes on cycle tracks having the same meaning as regular zebra crossings. Needs a change from the government and a cultural shift.

  5. "Evidence is always difficult, but the old LCN approach ..."

    I have asked you before not to talk to me about the old LCN. I do not regard it in any way as relevant to the issue under discussion, which is top-down or bottom-up?

    When I ask for evidence that the bottom-up approach is the correct one, what I am really asking for is quotes (see here). Such-and-such a person said such-and-such a thing about the value of pursuing a piecemeal approach, for example.

    But to say that finding this evidence is "always difficult", is another way of saying, "There is no such evidence."

    Just to be clear, I totally accept your view that we need direct protected routes as well as a dense network. But which comes first?

    With a top-down approach, the dense network would come first; with a bottom-up approach, the direct protected routes would come first.

    I am saying, "Network first, and then a separation of functions." What are you saying?

  6. bikemapper - the issue here is how we rework junctions to protect cyclists and how much work will be required to roll the idea out across London - yes that will include a dense network.

    As you know, I have to have a foot in both camps and I cannot ignore LCN as in my neck of the woods (Outer London), it is spine for direct travel by bike.

    What we need to do is perhaps use the routes which won't be mentioned as boundaries to smaller areas which need to be transformed for access only to vehicles (no through traffic), but dense filtered permeability.

    The trouble is, much of the money for the boroughs comes from TfL and they call the shots in terms of providing evidence on use which on the whole doesn't exist.

    The Sustrans approach was interesting as they convinced the funders (Big Lottery being the main one) that if we build protected routes (which may be a bit more leisure in Outer London)users will appear and that is where the data (physical counts) is showing it to be true.

    As I have stated before, CS2 is simply an extension of LCN12 and both provide direct routes between town centres, but it's quality is poor with very little protection. The bit of LCN12 where I live is not busy like in The City, but I would estimate that where it is on the carriageway, even with advisory cycle lanes, two thirds of cyclists are on the footway - they do not like the traffic.

    1. Ranty, I am preparing a blog about the Cycle Superhighways, and I was looking to see what other people had said, and I came across this blog, and I saw that I had commented. I must say that I hardly recognised myself in the comment I made above: I came across as very aggressive, and that's not really me at all. I apologise.

      Hopefully you will recognise that my attitude was born of total frustration. We have got such a long way to go in London, and as Dave Horton says:

      "We need to develop our visions and move beyond the shame of speaking them. Find our voice. Of course we must compromise – to make cycling big requires working with others, and that inevitably entails compromise. But unless they know what we really stand for, those others can’t know by how much we’re compromising."

      And then, as he goes on to say:

      "Just one example – the conversion of two lanes of a dual-carriageway’s four into top-notch space for cycling. Howls of protest, obviously. But the prospects of such change have to be higher the more people see them as forming part of an ongoing societal project to re-design our cities away from cars towards bicycles. The more people can see and understand the bigger picture, the more supportive they will be."

      I genuinely cannot understand how anyone would think to do differently to what the Dutch and Danes did in the 70s, and the Germans did in the 90s, to wit, plan a network, study it, and then get it up at running by doing as much as possible at least bureaucracy first (that is, says Cycling: the way ahead, all of the installation measures which call for little planning and which are needed to get a network to function may be adopted automatically, "without major risk of error or loss").

      Alas! we ain't never going to do it that way for as long as out strategy is informed not by evidence, but by ideology.

      I was pushing you hard because obviously your opinion would have some sway. I didn't mean to push you quite so hard as I did, but even so, I notice you still didn't answer my question, which is, very simply, top-down or bottom-up?

    2. Apologies for the lack of reply. First, you didn't come across as aggressive as you had reasoning behind the points - for me aggression is where the person has no reasoning!

      Top top or bottom up. To be absolutely honest, I don't know. I am a kerbs and tarmac person, not a planner or indeed a well-informed campaigner.

      Things have moved on apace in London since this post and as explained in the "about" section of this blog, I am learning! The N-S and E-W superhighways are of course important spines, but I would be the first to accept that they are not making a dense network.

      What they (hopefully!) will do is deal with the majority of capacity and technical layout issues in one big hit which will lead to people realising that if it can be done on those streets, it would be a doddle elsewhere.

      On the other hand in the day job, I have a good feel of which way the political wind blows and there is definitely a case for bottom up to get little things in which get people use to interventions so when you proposed something more radical, it seems less radical.

      Yes, so on the fence to some extent because it depends I guess!

    3. Thanks for your reply, Ranty.

      You say that the two new Superhighways would "deal with the majority of capacity and technical layout issues in one big hit which will lead to people realising that if it can be done on those streets, it would be a doddle elsewhere."

      Alright, okay. I don't much agree with this, but I am not going to argue about it. Okay. But having proved this point, what's on the agenda for the next mayoralty, I wonder? Prove it again? And then, probably, prove it again somewhere else? And so on, and so on, until eventually?

      As to your second point, consider some of the comments from aseasyasriding's blog about the Perne Road roundabout:

      "No-one ever seems to make the first step towards proper safe cycling infrastructure in the UK or properly designing cities to avoid problems like this from occurring. Rather, designers attempt to resolve large-scale problems by altering one junction at a time."

      "It’s about time road designers were held up to the highest professional standards backed up by the scientific method, not whatever passes as the current party political whim."

      If the political wind is blowing in the right direction – as it is in Bristol, say, or Brighton – good luck to you. If not, tough titty. As a professional, how would you feel if the political wind was against?

      For myself, I would sooner trust to proven strategies than to political winds. The case is that every town and city in this country could at least have a *functioning* cycle network.That wouldn't be difficult.

      There is definitely a case for bottom-up, as you say. But why not do bottom-up within the framework provided by a functioning cycle network?

      The designer of the Perne Road roundabout said that they could have provided a segregated cycle track around the perimeter, but that there was no off-carriageway infrastructure to link into and no prospect of providing any in the foreseeable future.

      If there was a functioning cycle network in place – which there easily could be in Cambridge – easily – he would have been laughed to scorn had he used a similar argument.

      There is definitely a case for bottom-up, as you say. But what is the *evidence* for it?


  7. Middle of the Road segregated cyclist CENTRE lanes avoid left turn deaths, opening car doors, taxi pick-ups, bus stops, parked police cars, hgv & van deliveries. Placing the cyclist in the middle of the road asserts an esteemed visible dominance that best seperates the responsibilities and procedures of motorist and cyclist. It solves the dangerous manouevre of the cyclist moving from the nearside kerb across the road looking over the shoulder at any oncoming traffic from behind in order to make a right turn! MIDDLE of the road CENTRE lanes use unused dead space taken up by chevrons or 3 metre wide unused pavements as in Piccadilly or Regent St High Holborn Holloway Rd Kensington High St Streatham etc etc. Claim 50 cms of both pavements on TOWER Bridge and create a middle of road segregated CENTRE Lane north going from 4 am to 13.00 and South going 13.01 to 3:59 am. CentreLanes avoid shoppers and pedestrians so politically they won't be so opposed by retail groups and other powerful bodies? CentreLanes use less space than go dutch schemes and are inherently safer In many respects.

  8. adam kingdon - trouble is, how will cycles get in and out of the many side roads they want to use - we cannot put traffic signals in at every junction?

  9. "In terms of removing traffic capacity - should anyone need to be driving so close to the centre of London other than those moving goods or many people?" Very good point. I would also add infrequent private car trips requiring short-term, relatively expensive car parking, and maybe some fleet parking for essential services. But catering for regular car trips with all day cheap car parking? Those days should be over.

    Two other points:

    Firstly, "traffic" needs to be re-defined away from its implied car-centricity and be applied to people e.g.

    Secondly, transport engineers should be now be focussing on maximising *people* throughput and not solely motor vehicles. I suspect that a road like The Overtoom in Amsterdam (dual tram tracks, dual one-way cycle tracks, footpaths and one lane in each direction for private motor vehicles) has a higher people throughput capacity than the four-lane road traffic control layout that could be fitted in the same space. The Dutch Crow "Design Manual for Bicycle Traffic" Table V19 implies that a 2.5 metre wide cycle path can carry at least 750 bikes per hour, which is a significant fraction of the nominal 1,800 motor vehicles per hour for a single motor vehicle lane (assumed speed 60km/hr which is probably too high for urban areas).

  10. Of course, there are other needs for cars in cities - some disabled people for example have no option but to drive.

    We have the Traffic Management Act which does define traffic as including cyclists and pedestrians, but is often taken as a proxy for vehicular traffic; but politicians can change who gets priority of course!

    Interesting link and one we could all learn from in writing committee papers, especially as we should be giving neutral and factual advice.

    1,800 vph is about right, even on a single carriageway road in town - not nice to cycle within or to try and cross.