SSVEP based BMI for a meal assistance robot

Abstract

Meal assistance robots provide disabled individuals the access to one of the important activities in daily living, self-feeding. This paper proposes a Steady State Visually Evoked Potential (SSVEP) based Brain Machine Interface (BMI) for controlling of a meal assistance robot. In the proposed system, the user has the facility to select any solid food item that he would like to eat from 3 different bowls just by looking at the respective LED matrices blinking at different frequencies. The generated SSVEP signals while looking at the LEDs are extracted from EEG signals acquired using OpenBCI EEG signal acquisition system. Extracted SSVEP signals are used to identify the intention of the user and subsequently the detected intentions are used to operate the meal assistant robot. Experiments are carried out to validate the system and results indicate the effectiveness of the proposed method.