Modelling of Motorcycle Accidents at Non-Exclusive Motorcycle Lane Junctions in Malaysia

In Malaysia, motorcycles constitute more than half of the total registered vehicles and contribute to more than 60% of casualties (death, hospitalised and slight injury) to total traffic accidents. In the 1990-2000 period, almost 3,000 motorcyclists were killed every year in traffic accidents. To overcome such problems, the exclusive motorcycle lanes have been constructed along the major trunk roads in Malaysia. However, not much work has been done to address junction accidents involving motorcycles. As such, a detailed study on this area has been carried out to allow traffic engineers to establish appropriate junction treatment criteria specifically designed for non-exclusive motorcycle lane facilities.A total of 104 junctions in Hulu Langat, Klang, Kuala Langat and Petaling districts with 1,095 injury related motorcycle accidents in the period 1997-2000 were included in the study. The generalised linear modelling with Poisson distribution was used to develop the model. The variables of the model were examined using univariate and multivariate analyses. The final models revealed that traffic flow entering the junction, approach speed, lane width, number of lanes on major road, shoulder width, junction control and land use were significant in explaining motorcycle accidents. Meanwhile, pedestrian flow, number of lanes on minor road and number of intersecting legs were not significant in explaining motorcycle accidents. Non-motorcycle flow on major road had the highest effect on the probability of motorcycle accidents at junctions. The final models allow traffic engineers to decide the appropriate intervention levels for junction treatment with respect to motorcycle accidents. Using the final models, design parameters for junctions may be changed to achieve the appropriate safety levels for them. The decision on whether to allow motorcycles to pass through a junction without treatment to it or the need for special end treatment to minimise motorcycle conflicts at junctions can be objectively carried out based on the model, and this can easily be done using the software developed in this study. Apart from the software, a series of design curves relating major road and minor road flows at junctions with typical shoulder widths of 0.0 m, 1.0 m, 1.5 m, 2.0 m and 2.5m have been established. The design curves were developed based on the number of Personal Injury Accidents (PIA) of 1-PIA per year. These design curves also enable traffic engineers to decide the need for special end treatment of the junctions