It can identify "a normal or abnormal state of the respective user of the respective service", the patent reads, based off training a computer model to look at certain data sets, such as the angle the device is held at, the time the auto is ordered, or walking speed, according to the patent, entitled "predicting user state using machine learning".
The premise of the software is to use a learned algorithm to determine a user's state of inebriation.
It will also draw on the location of the rider and the time of day the request is made.
CNN uncovered the patent, called "Predicting user state using machine learning".
A new patent application suggests Uber is exploring identifying drunk passengers with artificial intelligence.
Uber may be able to tell when you've had a few too many drinks.
For many commuters who have had one too many after work, Uber has been a godsend, allowing over-the-limit passengers to stumble into a cab and get dropped off outside their door at the touch of a button.
It is not known if the information will be used to stop drivers from picking up drunk passengers in order to avoid potential trouble.
If irregularities such as typographical errors, amount of time spent to make the trip request, or the user's accuracy in pressing an interface element are detected, the system would warn the responding driver to be on his toes.
Uber's patents said that the company's VR system will address this issue.
Uber says in the document the need for such a technology arises from the history of "safety incidents that can negatively impact the transport experience".
Ironically, Uber is concerned about drunk passengers assaulting their drivers while it is turning a blind eye to the many cases of sexual assault by Uber drivers.
There are a number of ways Uber plans to identify passengers who may be drunk.