Uber wants to know
if passengers are drunk before picking them up.
The ridesharing company has filed a patent to use artificial intelligence to identify how drunk a passenger is before picking them up.
“The system uses a computer model to identify user and trip characteristics indicative of the uncharacteristic user states,” says the patent application. “The system uses
the data about past trips to train a computer model to predict a user state of a user submitting a trip request.”
Ride-requester activities, such as how many typographical errors are
made requesting a ride, the amount of time it takes for the person to interact after receiving new information via the mobile app, the angle at which the phone is being held, movement of the device
during the request entry or a user’s travel speed can be factored in.
The system then analyzes the data and generates a prediction about the state of the user, such as does the person
appear to be drunk. The system can compare past trip behaviors of the person to identify any deviation.
Based on the results, the system could match the person with certain drivers and modify
pickup or drop-off locations.