-
-
The prototype headset.
-
The electronics behind the electrodes.
-
Our mobile app showcasing real-time stress information from the headset.
Inspiration
We were so stressed trying to come up with an idea for this hackathon that we decided to help individuals reduce crashouts. As students, we know how overwhelming exams and deadlines can be—so we created a solution to help others avoid burnout and stay mentally balanced.
What it does
Gauge uses four EEG electrodes placed at TP9, TP10, AF7, and AF8 (with A1/A2 as a reference) to detect real-time stress levels. The data is processed and sent to a mobile app, where users receive instant stress assessments and insights.
How we built it
- Trained an SVM model using a public stress classification EEG dataset
- Built a hardware frontend using ADCs and Arduino to sample EEG data
- Designed a mobile app using Flutter and Dart
- Connected EEG hardware to the app over USB for real-time predictions
- Synced stress predictions to Firebase Cloud Firestore for live app updates
Challenges we ran into
- Operational amplifiers lacked sensitivity to small differential voltage signals
- High ADC noise, mitigated through software filtering
- EEG signals were noisy and inconsistent, making classification difficult; we experimented with feature extraction techniques to improve accuracy
Accomplishments that we're proud of
- Built a fully functional mobile app that tracks and displays stress levels
- Used high-resolution 16-bit ADCs and digital filters to reliably sense EEG signals
- Delivered an innovative, end-to-end wearable stress monitoring solution
What we learned
- Never commit Firebase Admin credentials to a public repo :)
- Don't follow the crowd and wait three hours for food
What's next for Gauge
We plan to refine our model with better training data to boost accuracy and are exploring funding opportunities to develop Gauge into a full-fledged mental wellness tool.
Log in or sign up for Devpost to join the conversation.