Bimodal-Shared Control Interface for Assisted Mobility Application

Abstract

The paper presents (Electro-oculography) EOG- (Electroencephalography) EEG- Radio Frequency Identification (RFID) based Bimodal-Shared Control Interface for mobility assistance application by controlling a mobile robotic arm. EOG-EEG based bio-signal based bimodal interface has been used to move the robot following a predefined path to reach at an object placed at initial predefined position (Zone 1). RFID has been used as shared control interface for object identification and for sending trigger signal to gripper arm to pick the object and place it at another predefined position (Zone 2) automatically. A threshold based algorithm has been developed for horizontal eye movement (HEOG) detection. Minimum Energy Combination (MEC) method has been used to recognize the Steady State Visual Evoked Potential (SSVEP) brain pattern of EEG. Combining shared control with bio-signal based bimodal system has improved the classification accuracy and increased the number of commands without giving extra effort by user. This reduces the chances of fatigue in users due to continuously performing the same task. Classification accuracy and Information Transfer Rate (ITR) have been calculated as performance parameters and compared with previous literatures for evaluation.