Uber aiming to patent AI-driven technology that identifies intoxicated passengers

According to Uber, AI-driven identification of intoxicated passengers will be done by evaluating factors like the angle at which a smartphone is held, the number of typos made while adding destination and other such factors.

Published Date
13 - Jun - 2018
| Last Updated
13 - Jun - 2018
Uber aiming to patent AI-driven technology that identifies intoxi...

In a bid to make commuting safer, Uber has developed an artificial intelligence-based algorithm to identify intoxicated passengers looking for a ride. While not particularly new, the patent that was filed in 2016 is among one of the many technologies that the company has stated it is working on to "help improve the Uber experience for riders and drivers". The company also does not have any immediate plans to implement such technologies, with a company spokesperson highlighting that the application for the patent was filed to the United States Patent and Trademark Organisation almost two years ago.

Questions have often been raised pertaining to behavioural safety risks that intoxicated passengers often pose, both to drivers and co-passengers. It is a wide, grey area that may not instantaneously have concrete answers on how to deal with such situations, and it is also not wholly clear how Uber intends to use this technology to its benefit, if at all. According to the patent, the technology would use an AI engine to read the nature in which the Uber app is being used by a person. Metrics such as time taken to book a cab, the number of typos made while entering destination, and even the angle at which the phone is held and the number of taps that are made on the display will be evaluated by this AI engine that will presumably be equipped with machine learning algorithms and preset decisions to judge whether a person is in an inebriated state. This will potentially help drivers refuse such passengers, although this leaves a lot to chance and misjudgement as well.

Critics around the world have also suggested that the same may help identify frantic or vulnerable passengers, which in turn could be helped in speeding up the booking and arrival time of a cab. Uber has often faced much flak in the way it has handled data, including reports of the service live-tracking locations of drivers and passengers, and even a data breach that exposed private data of thousands of driver-partners. While such AI-driven technology leaves a lot of room for subjective manipulation, it may actually be used to streamline services like UberPool during fringe hours. As Uber states, such technologies will help "take an action to reduce undesired consequences". A lot of precautions and on-road training would be necessary, though, for such technology to make effective judgements.

Souvik DasSouvik Das

The one that switches between BMWs and Harbour Line Second Class.