Your smartphone could tell if you are drunk by analysing your walk

pints of beers

Your phone might realise you are drunk even if you don’t

SolStock/Getty Images

Smartphones may soon be able to tell whether someone is drunk based on the way they are walking, even if the phone user is unaware of the fact.

Brian Suffoletto, now at Stanford University in California, and his colleagues at the University of Pittsburgh in Pennsylvania, wanted to see if they could take advantage of the accelerometers embedded within most smartphones to detect changes in walking patterns that occur when people are drunk.

Suffoletto and his team recruited 22 volunteers and gave each an hour to consume a mixed drink with enough vodka to produce a breath alcohol concentration of 0.2 per cent, well above the legal limit for driving in the US of 0.08. They then strapped a smartphone to each participant’s lower back. Every hour for the next 7 hours, the volunteers were breathalysed and then asked to walk in a straight line for 10 steps, turn around, and then walk back 10 steps.

Advertisement


More than 90 per cent of the time, the researchers could use changes in a person’s gait, measured by the accelerometer on the smartphone, to accurately predict when breath alcohol concentration exceeded 0.08 per cent.

Suffoletto says the next step will be to determine whether it is possible to achieve similar accuracies when the phone is placed in different positions, such as when it is held in a person’s hand or is in their pocket.

“It could be used by individuals who want an alert when they are showing signs of impairment,” says Suffoletto. Previous research by his team found that people don’t realise they are impaired up to 50 per cent of the time when they are drunk. “It could alert someone who may not recognise the impairment and prevent them from driving their car when they are [drunk],” he says.

Since data related to smartphone use and sensors, including accelerometers, is widely collected, it is possible that this information could be harvested by third parties to try to determine whether a smartphone user is drunk, says Suffoletto.

“If someone wanted to go through the effort to process and analyse it, they could probably make inferences about changes in walking patterns,” he says. However, Suffoletto says that without more evidence, “it would be a leap” to conclude someone was intoxicated based on their walk alone.

Journal reference: Journal of Studies on Alcohol and Drugs, DOI: 10.15288/jsad.2020.81.505

More on these topics:

source: newscientist.com