BruBotics researchers from the Vrije Universiteit Brussel develop non-invasive brain-computer interface for everyday applications
Led by Prof Kevin De Pauw, researcher Arnau Dillen from the Human Physiology and Sports Physiotherapy (MFYS) research group at the Vrije Universiteit Brussel (VUB) has developed a brain-computer interface, in which the brain controls a robotic arm. This innovative project was facilitated through cooperation between MFYS and the BruBotics consortium at the VUB with the ETIS laboratory at the Université Paris-Cergy, in the framework of the EUTOPIA European university network.
“A brain-computer interface is a connection between the brain and a computer, through which brain activity is measured and converted into commands to control devices,” says Arnau Dillen. “Unlike invasive methods like those of Elon Musk’s company Neuralink, in this research we use electrophysiological measurements – electro-encephalography or EEG – that are completely non-invasive. This makes the system much more accessible for daily use without surgical intervention.”
“To enable this non-invasive monitoring, we use motor imagery, a technique that measures brain activity while a person thinks of a movement, without actually performing it,” says Kevin De Pauw, professor in the MFYS research group. By analysing specific frequencies of EEG signals, the system can determine which movement the person is thinking about. This data is then used to direct a robotic arm.”
“This required us to first develop a method to collect calibration data and use it to train machine learning – artificial intelligence – models,” says Dillen. “The system has to be recalibrated for every user, but with this method the process is considerably faster.”
A unique aspect of the research is the integration of augmented reality (AR) using Microsoft HoloLens. Dillen:
"These AR glasses use eye-tracking to select objects in the environment, while the user directs the robotic arm with their thoughts. This combined system provides an intuitive and efficient way for users to perform complex tasks."
The next phase in the research is to develop the system into a software product that can be brought to market. Dillen is already seeking postgraduate opportunities and industrial partners to refine the technology.
Contact
Arnau Dillen
Arnau.Dillen@vub.be
Kevin De Pauw
Kevin.De.Pauw@vub.be
+32 (0) 2 629 27 54
