Optimization of Driver Behaviour Profiling Using K-means Clustering Algorithm with Environmental Context

Abstract

Advanced Telematic systems have transformed modern vehicles into centralized data collection systems. Contemporary vehicles nowadays are equipped with sensors to collect data and transmit it to the Electronic Control Unit (ECU) of the vehicle. The data gathered from the sensors can provide insight into the driving behaviour and driving patterns. However, the high dimensionality and complexity of the vehicular data hinder its effective utilization for improving road safety; the data requires an advanced method for in-depth analysis. This study applies the K-means clustering algorithm to analyze and categorize driving behaviour patterns, and driving profiling was conducted based on the characteristics of each cluster. To account for external influences, environmental contexts through weather conditions were integrated into the analysis. The findings reveal that K-means effectively identifies six behavioural clusters; two clusters
were identified as good driving behaviour, while four clusters exhibit bad driving behaviour. Comparative analysis showed that environmental factors significantly influence driving styles, with more cautious behaviour observed under rainy conditions. The findings highlight the effectiveness of K-means clustering in profiling driving behaviour and emphasize the importance of considering environmental factors for accurate risk assessment. This research provides crucial information to the driver’s awareness, gives insight into policymakers and law enforcement, and thus improves road safety.