Bi-modal Human Machine Interface for Controlling an Intelligent Wheelchair

Abstract

This paper presents a bi-modal human machine interface (HMI) alternative for hands-free control of an electric powered wheelchair (EPW) by means of head movements and facial expressions. The head movements and the facial expressions are detected by using the gyroscope and the cognitiv suite of the Emotiv EPOC sensor, respectively. By employing the cognitiv suite, the user can choose his/her most comfortable facial expressions. Three head movements are used to stop the wheelchair and display the turning commands in the graphical interface (GI) of the HMI, while two facial expressions are employed to move forward the wheelchair and confirm the execution of the turning command displayed on the GI of the HMI. By doing this, the user is free of turning his/her head while the wheelchair is being controlled without the execution of an undesired command. Two subjects have tested the proposed HMI by operating a wheelchair in an indoor environment. Furthermore, five facial expressions have been tested in order to determine that the users can employ different facial expressions for executing the control commands on the wheelchair. The preliminary experiments reveal that our HMI is reliable for operating the wheelchair.