A System for Multimodal Assistive Domotics and Augmentative and Alternative Communication

Abstract

In this paper, we present an new assistive system called SMAD (System for Multimodal Assistive Domotics). Through this SMAD, a user with motor disabilities is able to control home devices from a wheelchair by means of biological signals captured on muscles (sEMG) and eyes (EOG/VOG). Our SMAD focuses on the following end-user features: ease of setup and use, low cost and customizability. The functional design of the interface has been driven by a user-centered performance analysis. Moreover, the system has been conceived to be adaptable to the user's degree of disability. The physical communication between wheelchair and devices on the environment employs radio-frequency (RF), infrared (IR) and Bluetooth (BT). Two scenarios are presented, using respectively facial gestures and eyes movements through Emotiv EPOC and Eye Tribe Tracker. Experimental results show fulfilling performance for the first scenario. The second scenario is an evolution of the first one, in which preliminary results indicate greater usability of the control interface; however the system must undergo testing by a larger number of people for better accuracy performance. During offline and online sessions, users succeeded in employing our SMAD to operate home devices, simulating common needs of people with disabilities, with accuracy above 90% and information transmission rate of up to 53 bits/min. The final goal of our SMAD is to provide the end-user with a practical daily assistive platform, as close as possible to a product that works out of the box.