BECAS
PERALTA Maria Florencia
congresos y reuniones científicas
Título:
Home automation by brain computer interface
Autor/es:
PERALTA, MARÍA FLORENCIA; NIEVA, EDUARDO GABRIEL; BELTRAMONE, DIEGO ANTONIO
Lugar:
Huerta Grande, Córdoba
Reunión:
Conferencia; Third International Conference on Advances in New Technologies, Interactive Interfaces and Communicability (ADNTIIC 2012); 2012
Resumen:
In the present work we use the Brain Computer Interface technology to allow to the dependent persons the utilization of the basic elements of their house, such as turning on and turning off lamps, rolling up and down a roller shutter or switching on the heating system. For doing this it is necessary to automate these devices and to centralize its managing in a platform, which constitutes a domotics system. In order to achieve this, we have used the MindWave NeuroSky ® commer-cial device. It is affordable, portable and wireless, that senses and delivers to the computer the electroencephalographic signals produced in the frontal lobe and the levels of Attention, Relaxation and Blinking to the computer. In order to de-termine the efficiency of the obtained signals a Test Software was designed, which verified the operation´s device with different persons. We concluded that the easiest way to control the Attention levels of attention is concentrating in a certain point, and the way to control the Relaxation levels is by closing the eyes. As a second step we developed a software that takes the signal from the EEG (Electro Encephalo Graphy) sensor, processes it and sends signals via USB to an Arduino board which is associated with electronics that complies the different tasks. The user chooses the action by managing the Attention levels. When they are higher than a particular threshold value, the action is executed. In order to disable this action, the user must lower the threshold level and overcome it again. This is the simplest and fastest way to handle but it brings with several problems: if the user concentrates for any other reason and this signal exceeds the threshold, it causes the activation of an involuntary action. To solve this problem, we use a three variables combination that can become independent of each other thru training properly. These variables are Attention, Meditation and Blink. When you comply three simultaneous previously established conditions, the action is executed, and when they return to fulfill the conditions the action is deactivated. The software also has the feature of personalizing its conditions, so it can be best for any user, even a novice one.