I found that piercing the hobbyist Brain-Computer-Interface(BCI) realm to be diffiuclt. Firtly, it is expensive - specifically because an EEG device can cost anywhere from 200 dollars to more than a few thousand. Second, it requires quite a wealth of experience - Signal processing, Neural Networks, Electronics, etc. This project is my attempt at creating a useful, relatively functional BCI at relatively low-cost and minimal experience.
This is a neurofeedback system that detects Mind-Wandering during driving: it reads your brain electrical activity, detects if you're Mind-Wandering, and beeps if you do.
Phase 1: Preparing the EEG headet - Hacking the Mindflex headset
Phase 2: Data Collection Module
Phase 3: Data Processing: Process data and train classifier
Phase 4: Make the device: Software
Phase 5: Make the device: Hardware
Phase 6: Real-Time testing
All details about getting started with the project + details about specific phases of the project can be found on a series of articles on my website : https://joshetasrinivasan.com/distracted-driving-detector-intro/
- MindWanderDataCollect.ino: Arduino code for Data Collection Module: reads breain data and prints onto serial
- EEGData_mindWander.py: Python script to save Arduino Serial contents into CSV file
- SampleData.csv: Sample csv file after data collection
- circuitElements.pdf: Circuit details for Data Collection Module
- EEG_NN_Classifier.ipynb: A jupyter notebook that trains a Neural Network to predict Mind Wandering
- MindWanderRealTime.ino: Arduino code for the final product!
- hardwareLayoutMindWander.pdf: Hardware layout for the final product!
You can contact me at [email protected] in the case of any questions or suggestions.
All resources used is either documeted in the code or my website(https://joshetasrinivasan.com/distracted-driving-detector-intro/) where the walkthough for the project is detailed.