Using only your brain activity (aka read your mind), the system can control various home devices.

Overview

More than 5 million people in the United States are living with paralysis. Reports suggest that 12,000 to 15,000 people in the United States have amyotrophic lateral sclerosis (ALS) and approximately 17 million people have cerebral palsy, globally. The emergence of Brain Computer Interface (BCI) technology presents an interesting avenue for improving the lives of those that suffer from these debilitating conditions. BCI monitors brain signals from a user, extracts relevant features, and translates them into a desired action. It has shown tremendous potential towards helping those with neuromuscular disorders or injuries that cause serious motor disabilities and hindering their normal communication with the outside world. The engineering goal of this project is to develop a home automation system which is controlled only by brain activity to support people with neuromuscular disorders. The P300 event related potential (ERP) embedded within the electroencephalogram (EEG) signals can be used to analyze and extract key data from brain signals. P300 signal is seen on an EEG as a rapid single potential change in response to a sensory, cognitive, or motor event. Detection of P300 signal requires the subject to correctly recognize the stimulus event to generate a strong and observable P300. A low cost EEG device, Emotiv EPOC+ with 14 channels, is used to record the EEG data while accessing the system using only the brain activity. A Python interface is developed to stream the EEG data for real-time application. The application will enable the user to select the home device control commands by using only the brain activity through P300 signals. The recorded EEG data is preprocessed and the extracted features for the target and non-target commands are used to train the classifier, Linear Discriminant Analysis (LDA). The performance is calculated according to the number of samples correctly classified into the target (P300) and non-target (non-P300) commands with the classifier. The classifier was able to predict more than 95% of the commands correctly. The participants performed real-time sessions to select commands related to home-automation tasks such as Light ON and Music ON simply by looking at a target command word from a display of 12 words on the user interface. The trained classifier identifies the user’s intended command. The system then provides visual and auditory feedback of the identified command to the user. The selected command is then wirelessly sent to a Raspberry Pi using network communication. The Raspberry Pi translates the received command to appropriate response and controls the respective home device. The system is currently able to control home devices like lights, music, television, and air-conditioner. People suffering from neuromuscular disorders have limited or no communication capacity. The proposed system would be very helpful and easy to use for the patients suffering from debilitating neurological disorders, allowing them to execute daily tasks with ease and improving their quality of life. The Emotiv headset is affordable as compared to other devices and the setup time for the device is short and can be used by anyone without expertise.