The view from the top of a double-decker bus

This article was published in 2014, in Newsletter 117.

It isn’t about the fact, that because of a recent operation, I’ve been having to use buses, with consequent doubling of trip times, but that in the distant past part of my job involved recording flows and incidents at experimental roundabouts. We had a modified old double-decker bus, which we used as a base, often plugged into a lamp post to power kettles and a very early desk-top computer, an Olivetti Programma (anyone else remember these?).

More relevantly, we could often use the upper deck as an observation point. I worked as an observer at a number of places where mini roundabouts or even multi-minis were being installed. Of course in the early 70s it was not about cycle safety, and I don’t think we even recorded cycles, but it was maximising the throughput of motor vehicles and minimising delays that counted, and experimental changes were made to achieve that.

This all gives me a different view from some others when looking at changes to roundabouts in Cambridge. We need to consider (1) the consequences of changes to layout and to normal driver behaviour, and (2) whether significant reductions in motor vehicle capacity will cause chaos.

One crucial thing about small roundabouts is to get the ‘deflection’ right, so as to slow traffic, such that entering traffic could risk a smaller time gap safely. Some found it counter intuitive that by reducing the size of both the central island, and the circulating space (then called ‘weaving sections’), that throughput increased and serious crashes were reduced.

Of course roundabouts are difficult for those on cycles or on foot… but many experienced users would not wish them to be changed to traffic signals, as they know that this can increase delays.

At the Radegund Road roundabout the challenge was to make it easier and safer for those at the two adjacent secondary schools, who should have just achieved ‘Level 2 Bikeability’, to use this junction. This of course with a limited budget. If we can get teenagers onto bikes, and this means convincing parents to permit them to cycle, it improves their independence, as well as reducing cars on the school run.

We need a location for a full ‘Dutch-style’ roundabout, with an outer ring cycle route where motor vehicles need to give way to cycles both on approaching the entry, and leaving the roundabout, but that is not here. Apparently, given the very heavy peak hour flows along the ring road, even the Dutch would not give those on cycles priority over Perne Road, but would use signals. Some suggest they would cure the problem by building an ‘Outer Ring Road’ to divert motor traffic from a residential area. I’m sure such money would be better spent on ‘distributed improvements’ that would advantage cycling and result in a significant modal shift and hence reduce flows on Perne Road.

So what has a compromise due to the lack of infinite funds achieved?

Those on foot or who lack confidence when cycling no longer need to make a significant diversion to the light-controlled crossing. Such users have a slower stream (and it is a stream in the peak!) of motor vehicles to cross, with wider refuges as part of the changes to the ‘splitter islands’. It is also legal to cycle the short stretches of shared use. Because of the number of crossing points and different routes for those walking and cycling any attempt to segregate would increase conflicts and lead to hundreds of the dreaded tactile paving slabs required by disability regulations.

On the road, numbers of people have said that traffic speeds are significantly lower. There is no longer an almost ‘straight through’ path that encourages speed. The narrower road and slower speed mean that those confident enough to cycle on the road should probably ‘take the lane’ if going straight on or turning right, to reduce the risk of a motor vehicle turning across their intended path.

Some have said the cycle lanes on Perne Road should run right to the ‘Give Way’ lines to aid in queue busting. I am strongly against this, as I feel it is unsafe, and saves little if any time. In December 2002 I wrote in the Campaign Newsletter about how to approach typical traffic lights.


Basically, if you are stationary alongside a motor vehicle at a Stop or Give Way line you are especially vulnerable. You may be in a blind spot and even if not, the first driver will be concentrating on either the lights changing or getting the first available gap. They won’t be looking around, and may even be turning left having failed to indicate.

Stay just to the rear of the first motor vehicle but in good view of the second. Unless you are turning left, don’t remain on the left. As the first vehicle moves off, you can ‘take the lane’, and if you are always observant and quick, even at busy times, you may be able to grab the same gap as the vehicle in front. The driver behind will have seen you, and if they also take the same gap, they should be less likely to turn left in front of you. Continuing the cycle lane right up to the Give Way gives several disadvantages:

  • it encourages those on bikes to be in the wrong position at the Give Way
  • it makes drivers think that is where bikes should be, and be less sympathetic (or even aggressive) to those who correctly ‘take the lane’

Although it may save a second, going right to the front and staying on the left is just risky. At busy times a queue-busting lane cycle lane such as on Perne Road may have already saved you several minutes. Be content. If you want to see the disadvantage of a cycle lane that goes right up to the Give Way, just go to the Fendon Road approach to the Addenbrooke’s roundabout. There much of the motor traffic turns left, yet most on cycles are going straight ahead to Addenbrooke’s. It tempts those on cycles into the worst position, and makes drivers think that is where they should be! This error is perpetuated in recently released plans for an upgrade of this roundabout.

Jim Chisholm