Intelligent Adaptive User Interfaces for BCI Based Robotic Control

摘要

The inherent lower bandwidth of a two-class BCI is still a major challenge in making BCI practical for day-to-day use. In this paper we present an intelligent adaptive user interface (iAUI) to facilitate enhanced robot control via BCI. The iAUI is a user-centric graphical user interface (GUI) that facilitates multiple controls needed in controlling a powered wheelchair or a robotic device using just a two-class motor imagery (MI) brain-computer interface (BCI). The iAUI offers a continuously updated prioritized list of all the control options (in the form of forward, left, right, backward and start/stop) for selection via a BCI by utilizing the information obtained from the sonar sensors mounted on the robotic device. Results on multiple participants in player-stage simulation as well as the physical robotic arena show that the iAUI addresses to a large extent the inherent lower bandwidth of a BCI for effective control of a robotic device.