Evaluating Perception of Acoustic Vehicle Alerting Systems (AVAS) using EEG Brain Waves

This study proposes a novel methodology to evaluate Acoustic Vehicle Alerting Systems (AVAS) tones by integrating psychoacoustic analysis with neurophysiological measurements. The method aims to optimize AVAS tones for urban environments, balancing pedestrian safety with user comfort. The approach begins with a psychoacoustic characterization of AVAS tones using key metrics such as loudness, sharpness, fluctuation strength, and roughness. These metrics are used to estimate Psychoacoustic Annoyance (PA), offering an objective measure of the discomfort caused by different sounds. To assess the emotional impact of the tones, electroencephalography (EEG) is employed to measure emotional states such as attentiveness, stress, and engagement under both idle and moving scenarios. The novelty of this method lies in the fusion of psychoacoustic and EEG data, creating a combined perceptual annoyance model where objective sound characteristics are validated against brainwave-derived emotional responses. This framework allows for the identification of AVAS tones that effectively capture attention without inducing excessive stress. The proposed methodology provides a comprehensive approach for the iterative design of AVAS tones, enabling the selection of sounds that maximize safety while minimizing annoyance, with the potential for future personalization based on individual sensitivities and environmental contexts.