BrainBotics is a university group at the University of Technology Munich (TUM) administrated by young and enthusiastic students.
Our overall goal is to control a robot arm by merely analysing human thoughts. Thus our aspiration for the future is to restore a certain independence for paralysed patients without risking a dangerous invasive surgery. In our first stage we are going to use an EEG to tap the motor cortex of the brain. The signal will be processed to distinguish which movement a patient is thinking of and then sent to a robot. In a next step, we want to add further signal sources. One example are cyber glasses to track the vision of the patient, so that objects can be focused. In general, we aim at developing a system that is easy to use based on open source components, where possible, and enable a project that is nursed by next student generations to obtain continuous improving results.
Our goal until October 2014 is to develop a moto-imaginary multiclass closed loop online BCI with feedback control based on an EEG System or Emotive Epoc. We want to classify the classes imagined left hand movement, right hand movement against the rest state. The correct classification rate of the two classes should be at 70%.
To achieve this goal we have determined several milestones: In the beginning, we will concentrate on a one class offline classification, e.g. only imagined left hand movement. As soon as we have found good classifiers we want to extend our system to an online, live system before expanding to a multiclass online application.
In our future implementation can be used to control a wheelchair or robot arm by using merely thoughts. In addition we also have small computer games for entertainment in mind.