Public transit driver behaviour analysis under heterogeneous road traffic conditions

Abstract

In regions with diverse traffic conditions, such as Sri Lanka, understanding and improving driver behaviour is essential for enhancing road safety and traffic management. This research paper addresses the critical need to study and model driver behaviour comprehensively. Traditional methods like surveys and observations are limited, expensive, and timeconsuming. In contrast, this study employs readily available GPS and speed data to analyze driver behaviour. The literature review highlights a scarcity of research, particularly in the case of public transport bus drivers navigating through heterogeneous traffic conditions. Previous studies have explored driver behavior among private vehicle drivers, revealing valuable insights into categorizing drivers based on behaviour. The methodology section outlines the data preprocessing and visualization techniques used to analyze driver behaviour. The process involves extracting and organizing vital information from GPS data and incorporating advanced visualization methods to provide clear representations of driver behaviour patterns. The research findings reveal distinct driver behaviour categories when examining speed patterns, dwell-time analysis at bus stops, and speed within zones between bus stops. These insights can help authorities develop effective strategies to prevent accidents, ease traffic congestion, and ultimately improve road safety and the daily commute experience. This research aims to contribute to saving lives and making roads better in regions with heterogeneous traffic conditions.