Virtual Reality based test batteries for evaluating Trust and Reliability in Human Machine teams

Abstract

Virtual Reality and Artificial Intelligence studies have interesting loads withinside the issue of all the packages that may be made with it. As they head closer to an excellent greater tech-pushed future, the improvement capabilities are steadily in demand. This has led researchers to find the human dependency and trust on machines in a VR simulation. Human trust is one of the main factors through which they can become more dependent on the machines. If the trust of a person on machines is high, then they can say that humans are more dependent on machines. Physiological sensor data such as electroencephalography (EEG) and galvanic skin response (GSR) are being used to evaluate the trust. In this paper, we proposed an approach which can be used for evaluating the human trust on machines by creating a Virtual Reality based driving simulator in which the player has to follow the path told to them by the machine. In our work, we extracted features using different methods. We did the Fourier analysis to study the spectral amplitude. Time Frequency Representation (TFR) was performed using Morlet Wavelets and Multitaper. Power Spectral Density (PSD) was computed using Welch’s Method and Multitaper. All these methods were analyzed to evaluate the trust in Human Machine teams. Our results indicate that all the frequency bands – delta, theta, alpha, beta, and gamma showed significant features related to trust.