Robotic Motion Control via P300-based Brain-Computer Interface System

Abstract

BCI have ignited extensive research interest in fields such as neuroscience, artificial intelligence, and biomedical engineering, as they offer an opportunity to interact directly with the external environment  brain signals. Despite the immense potential for applications, practical use of BCI still faces several challenges, including equipment cost and operational complexity. This study aims to develop a Brain-Computer Interface system based on P300 visual stimuli, utilizing a low-cost, user-friendly portable Muse EEG equipment for data acquisition. We designed and implemented a P300 visual stimulator in a 3x3 grid pattern, acquire the user's EEG signals using the Muse EEG equipment, and classify the data using a SVM classifier, ultimately realizing control over robot movement. Offline experimental results demonstrated an accuracy of 84.1% for the classifier under offline stage, while online stage achieved a successful execution rate of 81.2%. These findings substantiate the feasibility and potential of using low-cost, portable devices like the Muse EEG equipment for BCI research, opening new avenues in the field.