Systems that can process thoughts and translate them into commands to move objects are very useful for people who cannot speak or move, but they have a drawback: they cause mental fatigue.
The Mexican scientist has developed an intelligent interface that can teach up to 90% of the user's instructions in order to work autonomously and reduce fatigue.
The project, Automating the Brain-Machine Interface System, is an initiative of Christian Isaac Peñalosa Sánchez, Ph. D. candidate in Cognitive Neurology in Applied Robotics at Osaka University, Japan.
“I have been working on this project for three years, it is based on a brain-machine interface. Its function is to measure the activity of neurons in order to receive a signal generated by thought, process it and convert it into an order to move, for example, a robotic prosthesis, a mouse or household appliances,”says the scientist.
He explains that this system consists of electrodes located on the human scalp. They measure brain activity in the form of EEG signals. Signals are used to detect patterns generated by various thoughts and mental states of the user.
The system also includes a graphical interface showing available devices or objects that interpret EEG signals and receive user commands.
Promotional video:
In addition, wireless sensors are distributed in the room, collecting environmental data (temperature and lighting); mobile hardware drives that turn appliances on and off, and an artificial intelligence algorithm.
“The latter collects data from wireless sensors, electrodes and user commands in order to reveal the correlation between the environment of the room, the mental state of a person and his activities,” comments Christian Peñalosa.
He adds that in order to relieve users of the mental fatigue and frustration due to high concentration over the long periods of time that are inevitable with such systems, the system must become independent. This is what Christian tried to do.
“We have given the system learning opportunities by implementing intelligent algorithms that gradually learn user preferences. At some point, the system can take over control of most of the devices, leaving the user to focus on another goal."
For example, a person can use it to control an electric wheelchair while moving into a living room using basic commands (forward, backward, left, and right) that the system has already learned. The next time the user wants to take the same route, he just needs to press a button or think, the stroller will take him to his destination.
Once the system works automatically, the user no longer has to focus on managing different devices. However, the system continues to collect EEG data to detect the error signal. It arises when people are alarmed: the system or they themselves did something wrong.
For example, if the room temperature is quite high, the user wants the window to open automatically and the system turns on the TV instead. The human brain registers this action as erroneous. The system receives a signal about the error and tries to correct it.
Peñalosa's efforts led to significant results: in a number of subjects, their level of mental fatigue did indeed decrease after working with the system. The level of learning of such systems has also increased significantly.