According to a report from VentureBeat, Russian tech giant Yandex is testing a proprietary system for its Yandex.Taxi ride-hailing service that it says will combat drowsy and dangerous driving.
The system uses Yandex’s SignalQ1 Camera to monitor 68 points on a driver’s face, and uses machine learning algorithms to detect when the individual is tired or distracted. It looks for indicators such as blinking or yawning, and then uses the rate of occurrences of these indicators to assign a sleepiness and distraction score.
Currently being tested on a limited number of vehicles in Moscow, the system as it is now simply emits an audible beep to notify the driver when they are exhibiting potentially dangerous behaviors, but in the future Yandex aims to link the data directly to the driver’s Yandex.Taxi account, allowing the company to take preventative actions if it deems the driver unsafe.
“Whenever the driver gets tired, [they] will be notified and suspended from receiving further [ride] orders until [they get] some rest,” noted Aram Sargsyan, Yandex.Taxi’s regional general manager for Europe Middle East and Africa (EMEA) and the Commonwealth of Independent States (CIS), at the Move 2020 mobility conference in London this week.
Yandex.Taxi was started in 2011 in Russia and has been expanding its operations ever since. It now operates across the CIS and EMEA regions, and in 2017 merged its operations with Uber to launch a new joint venture that targets Eastern European countries.
As Yandex.Taxi has continued to grow its competitors have faced increased scrutiny in North America and Europe with regards to passenger safety. Uber recently lost its license to operate in London following a Transport for London (TfL) review that found a number of practices that TfL deemed were placing passengers at risk.
Among those charges levelled against Uber by TfL was the ease with which drivers who have not had background checks can use another driver’s account to pick up passengers. Uber’s response to this has been to implement a facial recognition-based check-in service for its drivers, similar to what is done in the U.S.
Uber is also attempting to address concerns over passenger safety by limiting its drivers to 12-hour shifts behind the wheel, after which a six-hour break must be taken.
On its end, Yandex is looking to avoid the bumps in the road that Uber is currently facing, and according to VentureBeat, it is currently working on developing its own biometric authentication system for its drivers.
Yandex’s system will differ from the one used by Uber in that it will have a multimodal approach to biometric authentication, requiring the driver to not only use their smartphone to verify their identity with facial recognition, but to also pass a voice-matching step in the authentication process for an added layer of security.
Yandex has also been working on implementing other safety technologies, including a speed control system which notifies drivers when they are driving too fast, and a system that tracks driving styles and can identify erratic or aggressive behavior.