Last year, Uber announced a new feature called “RideCheck” that utilizes the GPS, accelerometer, gyroscope, and other sensors inside a smartphone to detect trip irregularities, like a vehicle crash or an unexpected long stop. After a year of piloting and refinement, RideCheck is now live in the US, with other countries soon to follow, the company announced.
The system alerts both riders and drivers when something out of the ordinary is detected. A notification will pop up asking if everything is okay, which leads to a list of possible answers, including the option to call the authorities at 911 or Uber’s safety hotline. If Uber verifies that there has been an accident, the rider will be prompted to call 911. Uber’s team of safety operators may also reach out to ensure the rider is safe when the feature is triggered.
The feature doesn’t require any new permissions because it’s linked to the driver’s smartphone, rather than the riders. Drivers have the Uber app on more frequently than riders, who typically keep the app on in the background during trips.
To avoid any false positives, like a dropped phone, Uber uses machine learning to account for multiple factors in addition to the input from the phone’s sensors. Riders and drivers can also say they are okay in response to the initial notification, which leads to the RideCheck system getting better over time.
To be sure, using smartphones to detect and respond to auto accidents isn’t new technology. Other companies and startups, like Zendrive, have been working on similar crash detection technology for several years now. Other systems, like GM’s OnStar, use sensors embedded in the vehicle itself to detect crashes and automatically alert the authorities.