Brain Controlled SmartChair

Where

Zine Research Lab, NIT Jaipur

Objective

A Brain Computer Interface to control a wheelchair, with:

  1. Using signals originating completely from the brain, and not sources of noise such as eye-blinks or facial gestures
  2. EEG headset to be used should be economical so as to make the system affordable
  3. Monitoring patient mental health, to detect emergency situations


Emotiv EEG headset

Methodology

The previous research on Motor-Imagery and P300 EEG were pivotal starting points for this project. Utilizing their software implementations in a GUI constructed using wxPython and MPI, to synchronize the recorded EEG with events on the GUI, a proof-of-concept was constructed. The emokit library was used for recording EEG.

P300 Speller GUI

Results

  1. Accuracies were not promising, ranging from 65-70%. This resulted in a poorly functioning system.
  2. Upon running a number of tests, we realized that the electrode positions were problematic, in addition to the large amount of noise given out by the Emotiv.