Hi guys.
I have a couple questions about your neural interface.
1. There will be only four channels in your gadget. Is there at least one scientific publication that describes motor imagery on 4 EEG-channels?
2. You have placed electrodes in the frontal (AF3, AF4) occipital (О1, О2) zones and possibly reference channels in Т3 / Т4 or А1 / А2. This configuration is suitable for emotion recognition and p300 event-related potentials. How do you plan to register mental motor commands for wheelchair if you don't have any electrodes in the required location of head (motor cortex)? Are there scientific publications that confirm such methods of motor imagery detection?
3. In your video about wheelchair a professional medical-grade multi-electrode EEG-system on wet gel electrodes was used. Motion imagery paradigm is well described for such devices and have a classifier accuracy up to 96%. It is not clear, what is the novelty of your approach?
4. It is not clear which electrodes do you plan to use in your neural interface? Based upon  the sketches of prototypes, this will be dry polymer electrodes. How do you plan to detect a signal with such electrodes in the back of the head, where do people usually have hair? The sensitivity of dry electrodes is much lower than that of wet gel electrodes. This is a serious scientific and technical problem, but in your project there is not even a hint of its solution.