Yandex, the Russian search giant, recently conducted its first successful test of a self-driving car in Moscow. It’s significant mostly because it’s proof that autonomous vehicles can operate in harsh, snowy conditions.
It’s also a sign that Yandex, commonly known as the Google of Russia, could emerge as an important player in the race to build and deploy self-driving vehicles. According to The Moscow Times, Yandex’s on-demand taxi service first unveiled its driverless prototype in summer 2017. It tested the car’s maneuverability in snowy conditions on a closed course in November, and now the company has moved to testing on public roads.
“The Yandex.Taxi autonomous car safely navigated the streets of Moscow after a recent snowstorm managing interactions with traffic, pedestrians, parked vehicles and other road hazards on snowy and icy streets,” the company said in a blog post. “The drive, which occurred during light precipitation and -6 C (21 F) temperatures, was an advanced test challenging the vehicle with winter weather conditions on public roads.”
In the video, a safety driver can be seen behind the wheel, and yet Yandex described its vehicle as “Level 5 autonomous.” Most experts would associate a Level 5 self-driving car as one without traditional controls, like steering wheel or pedals, that can operate without human intervention in all conditions. It’s not clear whether Yandex’s car meets those conditions.
Last year, Yandex absorbed Uber’s operations in Russia after three years of fighting for market share. The new company was said to be worth $3.725 billion, and span 127 cities in six countries (Russia, Azerbaijan, Armenia, Belarus, Georgia, and Kazakhstan).
Operating self-driving cars in wintery conditions is extremely difficult. Winter weather is particularly problematic for the suite of cameras and sensors that autonomous cars use to perceive the world around them. That is because self-driving cars rely heavily on a spinning laser, called LIDAR, that tracks the objects around the car using laser pings. When snow is falling, the laser can confuse snowflakes with more solid objects.
In the US, most autonomous driving tests are conducted in dry, arid environments like California, Arizona, and Texas. Yet in recent months, companies have begun seeking colder locations: Waymo in Michigan, NuTonomy in Boston, and Argo.ai and Uber in Pittsburgh.