Robustness of roundabout metering systems (RMS)

A simple explorative study shows that roundabouts are too small to be equipped with permanently operating traffic lights. The conclusion of a recent, more detailed study was not fundamentally different to this: at a small two-lane roundabout with four legs and with leg-by-leg control the shortest green time is 11 to 13 seconds, irrespective of the volume of traffic on the least busy leg. This results in a great deal of unnecessary waiting time. Leg-by-leg control at a four-lane roundabout is therefore not regarded as a robust solution. In view of this, at small roundabouts incidental metering signals are used. At single lane roundabouts, metering signals provide a more balanced distribution of the waiting time, whereas at two-lane roundabouts, the metering signals also have the potential to improve traffic flow. This was shown as early as 2003, in research conducted in collaboration with the author. This article explains the theoretical background to the research. In addition, general design principles for roundabout metering signals (RMS) are deduced. In 2011, research carried out as part of a Master's thesis supervised by the author studied the robustness of RMS for various traffic loads and roundabout types. Two types of roundabout were studied: • a standard turbo roundabout and • a spiral roundabout. A simulation model was used that underestimated rather than overestimated the effect of an RMS. The results show that, even under less than optimum conditions, traffic performance of the whole roundabout was improved by approximately 10 %. Traffic performance in the relevant (saturated) leg improved by more than this, namely by 15-45 %. The reduction of time loss at all legs taken together is even greater: -20 % to -50 % in total, and as much as -70 % for the saturated leg. Notably, altering a dominant load pattern can even reduce waiting time in the leg with the metering light.