September 25, 2018
In recent years, it has become increasingly difficult for people to obtain driver’s licenses in China. However, an increasing presence of unmanned vehicles on the road is bringing new hope to those who can’t get a license but still desire to drive independently in a vehicle a chance to make their dreams become true.
Compared to traditional automobiles, unmanned vehicles feature driverless operation, automatic navigation, automatic obstacle avoidance and much more. Will autonomous vehicles transform the way people travel? That’s the topic the Chinese television show “Smart AI” recently explored.
In a Smart AI episode that was shown in September, the unmanned vehicle “XINGJI,” developed by Idriverplus and equipped with Velodyne LiDAR technology, tackled for the first time the challenging road conditions of Tianmen Mountain, located in the northwestern part of Hunan Province, under extreme weather. The autonomous vehicle drove successfully across a mountain road that has 99 bends, handling multiple difficult natural obstacles along the way.
From its appearance, the unmanned vehicle looks nearly the same as an ordinary vehicle. But thanks to Velodyne LiDAR’s long-distance and high-precision perception capability, the unmanned vehicle can be driven automatically after a user inputs the destination into the navigation system. The self-driving vehicle can automatically identify road lanes, traffic signs and more after being processed by Idriverplus algorithms. During driving, the unmanned vehicle will upload information about road conditions through the sensing equipment and perform real-time positioning analysis that leverages this large amount of data, thus determining how to execute travel direction and speed safely.
In the “Smart AI” episode, the unmanned vehicle was challenged to drive through a cake-like, winding route on Tianmen Mountain in 30 minutes. The route was 10.77km long, with its altitude rising sharply from 200m to 1,300m. The most dangerous section has a 180° curve, with mountain terrain on one side and a great cliff on the other side, creating an extreme road condition for any vehicle.
The unmanned vehicle had not previously experienced a similar driving situation, and there were only 48 hours for the system to learn its new environment. On Tianmen Mountain, “Smart AI” designed three tests for the unmanned vehicle. The first was automatic obstacle avoidance – detecting whether the vehicle could actively judge and avoid other vehicles parked in the road. The second was passing while on a curve – passing a bus at the most dangerous bend on Tianmen Mountain that has a degree of about 180°, width of only 6m and gradient of 35°. On this curve, the unmanned vehicle could only guarantee safety if it planned a route within 200ms. The third was to test whether the unmanned vehicle could drive through dark tunnel under the extreme weather of rain and fog.
“The challenge seems to be simple, even unsurprisingly, to the common people, but it is already very powerful for the industry. Every step on the hill can be fatal, and each step made by the ‘XINGJI’ is the fruit of the company’s ongoing efforts,” said Zhang Dezhao, CEO of Idriverplus.
“For today’s auto industry, the greatest challenge is not algorithms or products, but the organic combination of algorithms and products. In the next two to three years, the entire industry will rely more on the core advantages of technology to bring innovation. The cooperation strategy between Velodyne and Idriverplus will be an ideal solution for AI companies to achieve the integration of products and algorithms,” said Weng Wei, Velodyne LiDAR Asia Pacific Director.