Modelling drivers aberrant responsiveness to yellow interval: case study of durban signalized intersections

Abstract

The paper explored the extent of drivers’ aberrant responsiveness to yellow intervals at signalized intersections in Durban, South Africa. A signalized intersection primary light pole located at the stop-line has three different kinds of lights; green, yellow and red. It's a given that red means stop, green means go and yellow means when you enter the intersection during the entire yellow interval, it is permitted to proceed and clear the intersection safely, however in a situation where you can neither enter nor be in the intersection on red, you must stop upon receiving the yellow interval. At the onset of yellow, action taken is at driver’s discretion; probably explaining why drivers’ aberrant responsiveness is prevalent at signalized intersections. In a stochastic study carried out at four selected signalized intersections in Durban, a binary logistic model was used to estimate the probability of red-light running given; speed at a distance from stop line (51m, 70m) and acceleration variables. The results showed that for both the distances; 51m and 70m, the probabilities of red-light running were quite low at low speeds and increased with increase in both speed and acceleration. In the 51m distance, the aberrant behaviour was observed on average in the last second of the 3 second yellow interval. In the 70m case however, the behaviour was observed in the first 1.5 seconds. The paper concluded that driver’s aberrant responsiveness during yellow interval emanates from the absence of prescribed speed boundary information before the traffic stop lights.