Data-Driven EEG Model for Predict the Risk during Roundabout Maneuvers

The behaviour of drivers is significantly influenced by their perception of risk, which can have a profound impact on the transportation environment. This can potentially undermine road safety and efficiency. This study addresses this crucial concern by introducing an algorithm that forecasts driver-perceived risk using data obtained from electroencephalogram (EEG). The algorithm employs a Support Vector Machine (SVM) to develop a strong and predictive model that can forecast perceived risk levels. This model can then be used to inform the implementation of preventive safety measures. The efficacy of the algorithm was evaluated through the use of driving simulations, which involved three participants utilising the SCANeR Studio driving simulator. The simulations involved traversing a two-lane roundabout filled with vehicles and allowed the participants to make decisions during the entry and navigation stages. The results demonstrated the effectiveness of this approach even with a limited dataset with respect to a Pattern Recognition Neural Network (PRNN). This research offers valuable insights into the potential for neurobiological data-driven strategies to enhance driver safety.