Uber has filed a patent for technology that would detect drunk passengers. The patent outlines using machine learning to spot passengers who are behaving ‘abnormally’.
While the patent doesn’t go into details about what exactly abnormal behavior is, the language of the report implies drunkenness and fatigue. The system would work by tracking the way users walk and at what speed in addition to the way they are using the app, their typing speed, the accuracy of typing and angle of the phone.
This data is pushed into an algorithm alongside additional information about the user, such as when and where they requested their ride. For example, if a ride is hailed from a well-known nightclub early in the morning and the user is walking slowly or with an uneven gait, the system kicks in and sends the alert of a possibly intoxicated passenger.
Uber Says It Will Prioritize Passenger Safety
The patent application explains once an "abnormal" passenger is identified, several options can take place. For instance, a user can get matched with a specific driver; the driver can get an alert about the possible state of the passenger; and alternate pick-up and drop-off locations can even be suggested.
Uber says this technology will make rides safer and that it will avoid "safety incidents and personal conflict incidents [that] can occasionally occur when users and/or providers behave uncharacteristically."
Uber has a poor reputation for taking driver assault seriously in the past. A CNN investigation revealed that 103 Uber drivers have been accused of sexual assault in the last four years.
In many of these instances, the passengers were intoxicated. Since the departure of former Uber CEO Travis Kalanick, Uber has tried to turn over a new leaf in regards to passenger safety.
An Uber spokesperson said that the ride-sharing company is introducing new safety protocols, including better driver background checks. The new system could be used to protect drivers too, who have to contend with drunk and sometimes abusive customers.