Data mining of driver characteristics to spatial and temporal hotspots of single vehicle crashes in Western Australia

Abstract

This paper presents innovative methods for identifying the characteristics of drivers involved in single vehicle crashes (SVCs) in Western Australia when viewed spatially and temporally. Spatial and temporal hotspots of SVCs can be defined as clusters of SVCs that have exceeded the expected value over a certain time period and at certain locations. The EM (Expectation-Maximisation) algorithm was adopted to identify the characteristics of driver involved in vehicle crashes. Drivers were divided into different segments based on their socio-demographic characteristics, i.e., age and gender and driver related crash factors such as, drive license's types, alcohol consumption, speeding, fatigue and inattention etc. The spatial hotspots of the SVCs were identified using the Kernel Density Estimation method. Comap was used to integrate space, time and characteristics of drivers into one view to better understand the nature of the SVCs. The main conclusion draws from this study are as follows: • Characteristics of drivers subject to SVCs Change over time with different contributing factors. • Some subgroups of drivers at greater risk were identified. Females aged around 19, with BACs around 0.03%, on probationary licences were more likely to be involved in crashes between the hours of midnight and 2:10am and require hospital attention. Males aged around 61, on full licences were more likely to be involved in crashes between the hours of 7:55am and 10:55am because of inattention and require medical or hospital treatment. Males, and some females, aged around 26, on full licences, and with BACs around 0.07% were more likely to be involved in crashes between the hours of 2:10am and 4:55am because of alcohol, inattention, and speeding, and require hospital treatment. Males, and some females, aged around 36 were more likely to be involved in crashes between the hours of 10:55am and 17:40am because of inattention and require medical attention. Males and females aged around 17 on probationary licences were more likely to be involved in crashes between the hours of 17:40pm and 20:55pm because of inattention, and require medical attention. • Alcohol and speeding were found to be the main contributing factors in SVCs at night or early morning, and inattention was the main factor in SVCs during the day. • Severity of SVCs is generally higher at night than during the day. • The distribution of hotspots of SVCs varies over different time periods The results of this research will help road safety management authorities better understand the underlying nature of why crashes happen and where and when they do so that remedial action can take into account the characteristics of drivers.