NavEaz 2024
NavEaz is a smartwatch app our team developed to explore early signs of driver impairment. It uses heart rate and electrodermal activity (changes in the skin's electrical conductance) to track patterns associated with fatigue and drowsiness. Our machine learning model combines convolutional and recurrent networks to detect these patterns and estimate changes in attention over the next 5 to 15 minutes, with the goal of providing earlier support to drivers.
Recognition & Outreach
- Selected for presentation at the Transport Canada/NRC Community of Practice workshop in October 2024.
- Received an Innovation Design Award from ICACHI (5,000 RMB).
Core Features
- Track physiological signals through smartwatch sensors.
- Use these signals to estimate current and upcoming changes in driver attention.
- Use generative AI to tailor responses, such as suggesting a rest break or playing music.
Skills
- Python
- Machine Learning
- Real-Time Data Analysis
- Generative AI
- PPG & EDA Analysis
- Electrodermal Activity Monitoring
- Wearable Technology
Keywords
- HCI
- Driver Safety
- Predictive AI
- Real-Time Monitoring
- Wearable Tech
- Generative AI
Team Members
Yuzhe You, Yubo Jiao, Ce Zhang, Michael Brazeau