Influence of Roadway Curvature on Two-Wheeler Rider Kinematics and Safety in Urban Environments: A Case Study

抄録

Rapid urbanization is degrading urban mobility due to increased traffic congestion, pollution, and pressure on infrastructure. Two-wheelers contribute significantly through their high volume, frequent lane changes, and CO₂ emissions in dense areas. The combination of challenging road geometries (e.g., sharp curves, narrow lanes) and the inherent limitations of motorcycle design and rider behaviour amplifies risks for two-wheeler riders. In this context, the study aims to investigate the influence of road geometrical elements—specifically, curve radius, deflection angle, and curve length on the kinematic behavior and trajectories of two-wheeler (TW) riders in urban environments. Three machine learning models RFR, XGBoost, and SVR were used to model the non-linear relationship between velocity and kinematic variables on curved road segments. RFR and XGBoost showed high accuracy (R² > 0.94), while SVR performed moderately (R² = 0.69). The study results offer valuable insights for policymakers and highway designers to develop targeted interventions in infrastructure design aimed at improving two-wheeler safety.