Drunk User Interfaces - Determining Blood Alcohol Level through Everyday Smartphone Tasks

Explore a groundbreaking conference talk that introduces "drunk user interfaces" - innovative smartphone interfaces designed to measure alcohol's impact on motor coordination and cognition. Discover how researchers developed the "DUI app," which utilizes machine learning models trained on human performance metrics and sensor data to estimate blood alcohol levels. Learn about the week-long longitudinal study involving 14 participants that demonstrated the app's ability to estimate blood alcohol levels with high accuracy. Gain insights into this novel approach that could potentially provide a more accessible alternative to traditional breathalyzers for assessing inebriation.