Friday 13 January 2012

Bow(ing) to traffic flow?

In my previous post, I was cheered by TfL's response to the issues at Bow for cyclists. There appeared to be additions of cycle only lights to separate cycles and traffic wishing to turn onto the A12, as well as the consideration of cycle lanes over the flyover with light controls to aid cyclists on and off of the flyover.

So far, so good. For those who use the flyover, a way of being able to access it without trying to control a very wide lane of traffic splitting between the roundabout and the flyover, and for those who prefer the roundabout some lights to safely allow cyclists to proceed past the exit slips to the A12.

I was pleasantly surprised that these measures were being implemented, especially on the roundabout where previous suggestions at "Toucan" crossings for pedestrians and cyclists were dismissed by TfL as being too disruptive to "traffic flow". It looked like "traffic flow" (meaning, presumably traffic with an engine as opposed to traffic without) was taking less priority to safety at last.

However, I might be being cynical, but I am suddenly becoming very sceptical of these new plans, after looking at the plans and videos on the TfL site. Specifically these two videos

Now, the text indicates that the improvement is :

A cycle 'early-start' phase at the traffic signals on the eastbound and westbound entrance to the Bow roundabout. This would provide a dedicated green light phase to allow cyclists to travel ahead of other traffic 

But is this actually tying up with the videos (which is the entrance to Bow Roundabout Eastbound)?

I say this because the change appears to be a set of cycle lights in the cycle filter lane which controls cyclists entering the (larger) ASL. Once in the ASL, the cyclists are controlled by the standard lights which are green for both vehicles and cyclists.

So what would be the point of that?

Well, I suspect that the cycle lights go red when the standard lights are green to stop cyclists moving into the ASL and progressing through the junction with the traffic.

So conflict is being stopped by only allowing cyclists to progress at the green traffic light who are already ahead of the traffic in the ASL. If this is the case, then cyclists arriving at the junction at the green traffic light will be held in the filter lane by the red cycle light until the general traffic lights go red. Then the cycle lights will go green to allow the cyclists to proceed to the ASL where they will have to wait for a full rotation of the standard lights before being able to proceed when the standard traffic lights go green.

No wonder TfL isn't worried about "traffic flow" - the disruption is to cyclists, and not motorised vehicles. Also there are a few issues on top of the fact that, if this is a correct interpretation of the scheme, it will be massively inconvenient to cyclists :

1) The "early start" phase doesn't look like an early start at all. It looks like a standard head-start given by the gap in a deeper ASL. Slower cyclists will surely still come into conflict with vehicles if the vehicles are driven aggressively and the cyclist is slow. How big is the ASL to give a decent enough head-start? How fast will a cyclist need to pedal to avoid agressively accelerating cars?

2) Won't the ASL simply be populated with the normal assortment mopeds, taxis, and cars as per many other ASLs? In which case there won't be any advantage given to cyclists at all.

3) What happens if the roundabout gets backed up (as it often does)? Not only will this fill the ASL with vehicles, but the vehicles already on the roundabout may end up conflicting with the "early start" cyclists from the ASL.

I hope I have misread the video. I have watched it a few times and cannot see how else the scheme works. If I am correct in reading the scheme, then TfL will be making very little safety improvements, at the expense of delaying cyclists for a couple of traffic light cycles. It doesn't seem like much of a step forward.

I hope I am wrong...


  1. Unfortunately you are right. It's the same blue-wash we've seen before.

    ASLs don't protect cyclists because they are not obeyed by car drivers, and they only protect those that are able to get to it.

  2. I must have been reading Hembrow's blog too much, but I'm not impressed either.

    There's a cute raised curb between the car lanes and the bike lane. Which is fine, as the cars can't easily block the filter lane. But why does it end so early? Make it longer, all the way to the next exit so left turning cars can't squeeze cyclists. Just leave a gap big enough for the cars to get through. The exit lanes look very wide to me, allowing the cars to drive fast.

    The cyclists would wait very near the next exit on their own lane, and they could get across the gap pretty quickly if they continued on the circle. Left turns would just continue on the exit lane. They wouldn't have to stop at the lights at all.

  3. Looking at the videos I see there are two sets of traffic lights: one to stop cars before the ASL and the other just after the ASL. When the cars got green, the second set of lights seemed to turn off, not green.

    Are the lights before and after the ASL simultaneous (repeaters) or are they independent: can the first lights be red while the second ones are green? And which ones do the cars follow?

    If that was the case, with clever timing cyclists could get green to enter the ASL, and when they got moving they'd get green on the second lights too, without stopping. Only then the cars would get the green. This would give the cyclist an early, flying start.

  4. Yeah, it looks like you're right. Basically, cyclists get stopped at every potential turn conflict. We'd be better-off cycling in the regular traffic lane in 'primary position'.

    And what happens if a cyclist fails to observe his green light quickly enough and gets started late? He gets the same turning conflict as if the extra set of lights didn't exist.

  5. Has anyone else noticed that the blue paint on the 'cycle superhighways' is virtually impossible to see at night, and therefore redundant?