Researchers at the University of Pittsburgh have pioneered a smartphone-based method to detect excessive alcohol consumption, rivaling breathalyzers by analyzing gait changes.
The timeless reminder—"drink or drive, choose one"—could soon be enforced by smartphones. Published in the Journal of Studies on Alcohol and Drugs on August 18, 2020, this innovative system was detailed by University of Pittsburgh experts.
In a four-month study, 22 adults aged 21+ consumed caipirovska (vodka, lime, and sugar) to reach a 0.2% blood alcohol level—exceeding the U.S. legal limit of 0.08%.
Post-consumption, participants followed prompts from the Phyphox mobile app, which captured data from a three-axis accelerometer strapped to their lower back. Tasks included walking about twenty steps back and forth. They later repeated the exercises sober for comparison.
The accelerometer detected gait alterations with 92.5% accuracy. Staggering, a key intoxication indicator, was reliably identified. Researchers aim to refine this precision further.
This technology could transform driver behavior long-term. With apps already enabling vehicle startups, Phyphox could temporarily disable such features if intoxication is detected.
It echoes a 2019 Russian innovation: Yandex.taxi (acquired by Uber) deployed windshield facial recognition to spot driver fatigue via yawns and signs, blocking new rides as needed.
