Self-driving Xiaomi SUV sets Nürburgring lap record yet a human race car driver is still faster

Xiaomi took just two years to turn a profit from its nascent EV business and is now making an effort to elevate the autonomous driving capabilities of its electric cars to compete with giants like Tesla or BYD.

Stunts like being the fastest autonomous vehicle at the legendary Nürburgring racetrack would certainly help, and Xiaomi didn't miss the opportunity to deliver a record self-driving lap there.

First self-driving lap record at Nürburgring

Xiaomi has set what it calls the world's first autonomous-driving lap record at Germany's storied Nürburgring, completing the 20.8-kilometer circuit in its YU7 GT SUV without a human at the wheel. The time clocked at the end of the circuit was 10 minutes and 29.483 seconds.

While that really is a milestone for autonomous racetrack driving from a production car, back in May the very same YU7 GT, with an actual professional driver in the seat, posted a lap of 7 minutes and 22.755 seconds to set a production SUV record. This makes the autonomous system more than three minutes slower than a skilled human race car driver, which is quite a sizeable gap on one of the world's most demanding circuits.

To its credit, Xiaomi framed the autonomous lap as "a new starting point rather than an end point," which is about the right way to read it, as it can only improve the self-driving lap times from here. "On the day of the test, when an unmanned car drove on the track, everyone present stopped to watch, and the only consensus among them was that the future had arrived," added Xiaomi Auto's CTO.

As a reminder, Xiaomi's top YU7 GT SUV trim packs a 1,003 hp dual-motor setup, hits 0–100 km/h (0-62 mph) in 2.92 seconds, and its max speed is 300 km/h, so there is a lot of potential still to be realized by Xiaomi's ADAS.

The electric SUV can also add 250 miles of EPA range in 15 minutes thanks to the 900 V powertrain architecture platform, so the EV's specs are really not the bottleneck here but rather the autonomous driving software's abilities to react and take risks.