The existing concepts of home automation are being rapidly augmented using modern inventions from the field of electronics. Now you can easily find devices on the market which contain interfaces allowing to control them using personal computers, tablets, smartphones or even gestures. But what would be your opinion on even more advanced idea: using your brain to accomplish the same task?
Sounds appealing, doesn’t it? Certainly, gesture-based control may also not require any additional hand-held hardware, while reading your brain signals cannot be accomplished in a wireless manner and still somewhat lacks precision, at least when using portable headsets. But despite current technological limitations the idea of using a brain-computer interface to control home automation devices could be implemented in practice, say the scientists from the Birla Institute of Technology & Science (BITS Pilani), India.
According to the article published online on arXiv.org, the brain-computer interface (BCI) based system could incorporate additional functionality provided by other technologies. For example, bluetooth communications could be used to accomplish indoor user localization and to transmit signals to electronic controllers. But the most interesting feature could be implemented by detecting human eye-blinks. As the authors of the paper note, this approach of controlling gadgets and household appliances with your eyes would be especially useful for a wide range of people, starting with those who are almost completely paralyzed to completely healthy persons.
Certainly, some kind of image recognition technology would be required in order to provide such functionality. But image recognition technologies do exist for many years now and are already used in practice in various applications. However, another issue poses some more challenge: how to identify a particular device that you want to establish a connection to?
For the purpose of device identification for home automation purposes the researchers propose using the technology of steady state visually evoked potential (SSVEP). The name of this technology may sound very complicated but it has been around for several years now and its essence can be described quite simply: our brains generate signals of specific frequency when the light of corresponding frequency hits our eyes. More exactly, such natural response can be generated by a visual stimulus ranging in frequency from 3.5 Hz to 75 Hz. This phenomenon is used widely in human vision research using electroencephalographic (EEG) equipment.
The authors of the article say that only frequencies in the range of 6-24 Hz can be used to induce an ‘adequate’ brain response to the SSVEP stimulation, while maintaining a 0.2 Hz gap between successive stimuli. It is a short interval of frequencies, but it is enough to assign an individual frequency to a large number of devices nonetheless.
“When a user focuses on a light source flickering at a predetermined constant frequency within the aforementioned interval, the evoked EEG signal can be used to determine the device associated with that frequency. The successful selection of this device can be used to trigger the desired action”, the team explains.
According to the technical description, an EEG data sample of 2 seconds at the sampling rate of 512 Hz is completely enough to perform SSVEP-based remote device selection. What to do next? Then user has to blink three times to confirm that this is the device that he intended to select. A 4 second time gap is provided for user confirmation; if no blinking or incorrect number of blinks is detected system considers an error in device selection and the process has to be repeated.
In the article, the scientists describe the experiment they performed to verify the concept feasibility. The test involved four test subjects – three males and one female. The scientists used two sets of table fans and a table lamps placed in two adjacent rooms each. An EEG equipment was used to read the subjects’ SSVEP response. A cluster of six 5 mm LEDs generated the visual stimulus and localization and data transmission was accomplished using USB bluetooth modules.
The tests results clearly demonstrated the robustness of this brain-computer interface approach. The eye-blink detection accuracy was 100%, while the average system response was 5.2 seconds (more results can be seen in the table above). The system does not need any graphical user interface, while its efficiency is on par with or better than the current state-of-the-art, note the developers.
It seems that for us the only thing remaining is to see such system in action – and most likely it will not be limited only to the field of home automation.
Written by Alius Noreika