Desarrollo de un sistema interfaz cerebro-computador basado en respuestas de potenciales visualmente evocados para control de dispositivos iot.

Abstract

Steady-State Visually Evoked Potentials (SSVEPs) are natural responses of the brain to visual stimuli when staring at a signal between 3.5 y 75 Hz. These signals can be detected through the classification of brainwaves gathered through an electroencephalography device that is connected to the user. Then, the classified SSVEP responses can be transformed into commands to control electronical devices. With this background, in this study an online Brain-Computer Interface system to control IoT devices based in four SSVEP stimulus generated by LEDs blinking at distinct frequencies was built. The brainwaves required to generate to generate SSVEP responses were gathered through an electroencephalography device called OpenBCI and were generated when the user stares to the visual stimulus. Then, were classified using Canonical Correlation Analysis (CCA) algorithm and transformed to commands to control the user interface of the application to control IoT devices. Once obtained the commands, were send through sockets to the user interface that allows to control a list of IoT devices. The IoT devices control from the user interface was made through the MQTT protocol. During the tests made to the final prototype, with fifteen volunteers, an average time of 9.7 seconds was obtained to turn on a smart bulb and 16,68 seconds to turn it off. And the classifier algorithm got a success rate of 97.61%.