Uber is no stranger to controversy. From regulatory disputes to employment issues, the company seems to push the envelope continually and, as a result, shape the way we go about our day-to-day lives for better or for worse. Surprisingly, their most recent controversial action appears to us in the form of a patent application they filed on December 1st, 2017 that published in June 2018. Uber filed a patent application titled “Predicting User State Using Machine Learning” (Application No. US20180157984A1) that describes using artificial intelligence to determine whether an individual requesting a ride is intoxicated – a new application of AI and use of user data that is ripe for debate.
Uber’s Problem and Patent-Pending Solution
For a ride-hailing business, such as Uber, transporting drunken passengers can be both costly from a business standpoint and dangerous. For example, when a drunk passenger vomits in a vehicle, it takes the Uber driver out of commission until the vehicle is cleaned appropriately. Not only does this mean that the Uber driver is unable to earn money during that time, but it often results in Uber picking up the tab for cleaning charges. Another challenge with intoxicated passengers is their unpredictable behavior, which could inhibit a driver’s ability to pay full attention to the road and to keep everyone inside and out of the vehicle safe.
To address this issue, Uber has filed a patent application covering a “computer-implemented method of data processing for predicting user state.” The patent application describes using app data to determine a sober baseline for how users typically walk and interact with the application. Then, the patent application describes a way to assess a user’s impairment by analyzing typographical errors and how app buttons are pressed, a user’s walking speed and swaying motions, and when and where a ride request is made. How Uber will treat riders they have deemed intoxicated is not discussed in its patent application – whether they will charge a premium rate or refuse service altogether. Both possibilities and an ongoing concern for privacy have driven conversation about their described method.
Uber’s Past and Future Patent Prosecution
Whether Uber will ever implement its technology is to be determined. Many large, innovative companies have patent portfolios containing many patents that are never used, and Uber is no exception. Uber Technologies, the official assignee name on their patent application, has filed over 625 patent applications since 2004.
Uber Patent Filings Over Time
Impressively, Uber has been granted 90.8 percent of their patent applications that have been fully prosecuted (226 patent applications are pending to date). Their patent application for determining user impairment has been assigned to Art Unit 2122, an art unit responsible for computer architecture, software, and information security patents that typically allows 71.1 percent of the patent applications they examine to grant. It is difficult to determine this application’s fate looking at their assigned patent examiner’s patent data because the application has been assigned to a Supervisory Patent Examiner who has examined an unusually low number of patent applications. However, give the patent application track record for Uber and the general patent statistics of their assigned art unit, the odds are good that Uber will eventually be granted protection over their controversial method.
Patent data and patent statistics have been provided by LexisNexis PatentAdvisor®. PatentAdvisor™ patent prosecution tools arm patent practitioners with the ability to develop informed patent strategies and make patent prosecution more predictable.
Get to know your examiner better with more context and a deeper understanding of your examiner’s behavior than ever available before.
With your free trial, you will gain instant access to:
Examiner Search allows you to search by examiner name for a filterable, examiner specific dashboard of patent analytics, including rejection specific statistics, appeal statistics, prosecution statistics, interview statistics, a backlog of RCEs and timeline.
QuickPair easily replaces the USPTO Public PAIR by providing the most robust application details anywhere, including examiner timeline, examiner allowance rate and the average time and number of office actions to allowance.
PatentAdvisor, the first-ever data-driven patent strategy tool, provides a systemic approach to crafting an effective prosecution strategy. Understand why certain patent applications take longer than others to reach allowance—then use that knowledge to devise better patent prosecution strategies.