VP of Autonomous Driving for LeEco, Ni Kai, said that to have autonomous cars run in China, developers must take local Chinese characteristics into account. His comments were made on the weekend at the 2016 World Robot Conference in Beijing.
“Can an autonomous vehicle developed in the U.S. run on roads in China? My answer is negative,” said Ni. “Transport in China is extremely complicated, and we must take more into consideration when we design these sorts of products.”
Ni said that both perception and planning are crucial. Other than using sensors and high-definition maps to improve the car’s own perception, the autonomous car must be good at planning – for example, deciding whether the car in front will cut in the road.
For LeEco, its super car must first be electric. More than an autonomous car driving on its own, it must also be a smart car that is well-connected to the internet of things (IoT) and interact well with its users.
According to Ni, when it comes to artificial intelligence, China is among the forerunners – at least not far behind the U.S. when measured by the number of published thesis papers that have been cited. However, China does lag behind without its own CPU, GPU, or operating system (OS).
Having its research and development of autonomous cars based in both China and the U.S., LeEco is looking to welcome talent from around the world to take part in its technological research, including in machine vision, vehicle controlling, high-definition positioning, and perception planning.
According to Ni, LeEco has obtained its Autonomous Vehicle Testing Permit in the U.S.
In addition, LeEco, together with the LeEco-backed car-hailing app Yidao, have been working on building charging points and parking places for electric cars.
Whether these strategies will give LeEco an edge over Tesla in the China EV market remains to be seen, but if so, Tesla would certainly not be the first tech company to lose out by misunderstanding the different conditions of the Chinese market.
(Top photo from Baidu Images)