Analysis of the influence of traffic variables on lane change manoeuvres decision-making in real-world traffic scenarios.
This paper presents an analysis of lane changing manoeuvres under real traffic conditions, focusing on how drivers adapt their behaviour across different driving regimes. Using naturalistic driving data collected on a high-capacity road in Madrid (Spain), the research explores the role of traffic context in decision-making focusing on gap acceptance, vehicle dynamics, safety metrics and motivational factors. To analyse these dynamics, drivers were exposed to six controlled traffic configurations designed to reproduce representative real-world highway scenarios with different surrounding vehicle distributions, inter-vehicle distances, and relative speeds. Driver behaviour was analysed across three distinct manoeuvre phases, corresponding to car-following, anticipation, and lane change execution. The results show that traffic density and the surrounding environment significantly shape driver strategies. Higher traffic complexity leads to longer anticipation phases, greater dispersion in timing, and reduced safety margins. While Time-to-Collision (TTC) offers general insights into risk, gap-based metrics proved more stable and informative for predicting the actual execution of the manoeuvre. This study contributes to the understanding of context-dependent lane-change behaviour using controlled naturalistic driving scenarios. The findings support the development of predictive models and the design of advanced driver assistance systems (ADAS) and automated vehicles, enabling safer and more adaptive lane-change algorithms that better account for driver intent and traffic dynamics.