Targeting speeders on Perth roads

Abstract

Speeding on the road is not only a concern for developed countries i.e. USA, Europe and Australia, but continuing to be of alarming concern for many countries in Asia, Africa and the Middle East. This paper aim is to focus on highlighting speeders characteristics on40 km/h on a busy urban road with high pedestrian movement. Thus, authorities decided to drop the limit on that road to 40 km/h from 60km/h and introduced an electronic variable sign as an engineering measure. Speeding data was utilised from the WA (Western Australian) Police to examine the factors that may contribute to speeding. This data will be applied to compute the logit from the independent variables such as age, gender, time of day. Detection types are either ‘on the spot’ type where infringement notices are delivered on site to offenders, or ‘roadside’ type where notices are posted to offenders. Four speeding levels were recognised (low, medium, high and excessive) that was employed as a dependent variables. The low level speeding was considered as a reference category. All the above variables were modelled using multinomial logistic regression. A stepwise procedure included all the selected independent factors.Results showed that the likelihood of detected speeding does decrease with age and it also showed that this is a slightly stronger indicator for more dangerous speeding. Thus, younger drivers are slightly likely to drive with high speed. Age ranges between 17-83 years. Female drivers are less likely to be detected speeding than male drivers. As for the time of detection, it seems that drivers are prone to be detected during afternoon time more than morning time. This was consistent with road crashes between 2007- 2011 road crash data on the same study road. Significantly, it was found that after the period of the installation of the electronic 40km/h electronic flashing signs, drivers are less likely to speed compared to before the installation adding to the model. The period of testing the signs benefits in reducing speeding may be short and may need more time to determine. Further to the above, the model needs to be simulated for further testing.