BrewWave: Brain-Controlled Radio App
GitHub · Devpost · Demo · NatHACKS 2023 — NeurAlbertaTech
A hands-free radio web app controlled entirely by brain signals, built and tested with real BCI devices at NeurAlbertaTech’s 48-hour hackathon where I served as team lead. Designed to integrate with existing clinical software at Glenrose Rehabilitation Hospital, giving individuals with severe motor disabilities independent control over their radio experience.
Users make selections through a motor imagery EEG paradigm — imagining a physical movement produces a distinct brain signal, which gets mapped to a keyboard action and triggers the corresponding radio control. No physical input required.
How it works
- User runs the app alongside existing clinical BCI software (i.e. Emotiv)
- User imagines a movement — the EEG headset captures the signal
- Clinical software classifies the motor imagery and maps it to a keyboard keystroke
- BrewWave intercepts the keystroke to navigate categories, select a station, and control playback
Development Details
As lead, I scoped the project around real clinical constraints and drove full-stack development across the team:
- Designed the UI flow around the motor imagery paradigm — minimal selections, large targets, unambiguous feedback
- Built front-end (HTML, CSS, JavaScript) and back-end (Express.js, Node.js) for radio streaming and keyboard-event-driven control
- Ensured compatibility with clinical software by relying on standard keyboard mappings rather than custom drivers
- Provided our own implementation of real-time classification of EEG with LDA
Team: Ayan Abbas, Elena Arnold, Ian Gauk & Kaitlyn Stark.