EEG-Based Classification of Emotional State Using an Autonomous Vehicle Simulator

Abstract

Societal acceptance of self-driving cars (SDC) is predicated on a level of trust between humans and the autonomous vehicle. Although the performance of SDCs has improved dramatically, the question of mainstream acceptance and requisite trust is still open. We are exploring this question through integration of virtual reality SDC simulator and an electroencephalographic (EEG) recorder. In order for a passenger to build and maintain trust, the SDC will need to operate in a manner that elicits positive emotional response and avoids negative emotional response. In our experiment, a test subject was exposed to scenarios designed to induce positive and negative emotional responses, quantified by the EEG beta wave to alpha wave power ratio. As predicted, an increase in the beta to alpha power ratio was observed when the test subject was exposed to stress inducing situations inside the SDC simulator. Our results are expected to inform the design and operation of an EEG-based supervisory feedback control module or artificial intelligence (AI) that monitors the emotional state of passengers and adjusts the AI control parameters accordingly.