Google rival set to rival Google's autonomous cars
Having followed Google's lead into smartglasses development, Baidu is now doing so with autonomous cars. The Chinese search firm says it has successfully tested a fully autonomous car under mixed road conditions, claiming it's the first in China to have done so.
"Fully autonomous driving under mixed road conditions is universally challenging, with complexity further heightened by Beijing's road conditions and unpredictable driver behavior," says senior vice president of Baidu and general manager of Baidu's autonomous driving business unit Wang Jing.
Sick of Ads?
More than 700 New Atlas Plus subscribers read our newsletter and website without ads.
Join them for just US$19 a year.More Information
Baidu has been using a modified BMW 3 Series for the tests with its own "AutoBrain" autonomous driving technology. This includes highly automated driving (HAD) maps, which are said to record 3D road data to within a few centimeters of accuracy of the vehicle's positioning. Baidu says that it hopes to have mapped the majority of China's roads in this way within five to ten years.
In addition, the vehicle features positioning, detection and smart decision-making and control technologies. Its object recognition and environment perception technology allows the car to detect, recognize and follow other vehicles, recognize road lanes and gauge distance and velocity.
So far, the tests have been carried out on one 30-km (19-mi) route, beginning at Baidu's Beijing headquarters, before travelling via the G7 highway, Fifth Ring Road, Olympic Park and finishing back at Baidu HQ.
The tests required the car to carry out a set of driving actions and to respond appropriately to the driving environment. These included right turns, left turns and U-turns, deceleration when vehicles were detected ahead, lane changing, passing other cars, merging into traffic from on-ramps and exiting from off-ramps. The car also hit a top speed of 100 km/h (62 mph) during the tests.
The company plans to develop its autonomous car program through the incremental introduction of new environments, rather than by simply teaching its cars more autonomous actions.