I am familiar with the round-a-bout/traffic light pros and cons, however, I've been tasked with deciding on a solution for the given parameters below.

Currently, there is a two-way (four lane) road with a speed limit of 35 mph. Rough traffic counts for peak hours are 580 vehicles in the morning (primarily North bound), 660 vehicles mid-afternoon (North/South split), and 400 evening (primarily South bound).

Access to a new facility will create a need for an access road resulting in a 3-way intersection. Access road will be to the East and for the most part only North bound traffic will make right turns to enter the facility. Less than 10% of the North bound traffic is expected to use the facility at any given time. Once finished at this facility, it is expected that the traffic will leave via the same access road and make right hand turns to continue North bound. However, as much as 20% of this traffic will need to make a left turn to return South bound.

Assume a community with some, but very little experience with round-a-bouts, and a primary desire to keep North and South bound traffic flow steady.

Giving your opinions and reasons, would you consider a round-a-bout or traffic control device utilizing a camera system for the left turn lane on the access road to be better?

  • 4
    $\begingroup$ It's hard to visualise what you're trying to describe here. A picture would help, also knowing if the traffic drives on the right or left would help. $\endgroup$
    – jhabbott
    Feb 7, 2015 at 21:58

1 Answer 1


A round-a-bout introduces a disruption in any traffic, no matter which direction. In your case the only two "collision" directions - two left turns - are the least used ones, so traffic lights working in a preference program with induction loops on the two possible left turns, plus uncontrolled right turn lanes would be optimal.

enter image description here

The crossing would stay in green light as long as there's no need for vehicles to take the turn. Only detectors would activate the phases for left turns (and only one would block both the north and the south road). You can either limit the rate at which the side exits can be activated by detectors on the main direction (better, more expensive) or by timing it, so that there must pass at least X time between the turn activations.

If real estate is a premium, you can create about all of the infrastructure for this crossing by reducing the main road to one lane in each direction, dedicating the remaining lane to turns, inclusion, and the "islets".

Here's an actual map from one of our controllers with a similar situation. This one includes two zebra crossings as well. It uses more detectors to determine optimal timing of green light, and change signals even before given vehicle would have to stop.

enter image description here (an interesting thing about this one, it uses detection loops on the main road but videodetection on the side road - the road was still due to be finished when the traffic lights were installed, and so we couldn't put induction loops in the gravel - currently induction loops would be better there (both more reliable and cheaper) except videodetection is already in place, so replacing it would only mean more cost.

Also, if your investor is rich and wants something prestigious and memorable (and truly optimal for the traffic), get a round-a-bout with a viaduct.

enter image description here


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.