I experienced the urban NGP function under XNGP on the P7i in Shanghai, and later on the G6 in Guangzhou. During my second experience, I quickly realized it was a tale of two cities. The time in Shanghai was uneventful, at least one intervention was due to me being disoriented rather than the car. In other cases, I was too cautious.
In Shanghai, the system lost power for no apparent reason only once or twice, while in Guangzhou it happened much more frequently. One possible reason for this is that the torque in the system cannot overcome your hands on the steering wheel, so the system may think you are intervening. Still, in Guangzhou it got stuck behind a parked car and at one point appeared to be heading towards an e-bike waiting to cross the road instead of entering the road it was turning into.
Guangzhou’s two-wheeled transportation seems to present a challenge to the system in general. Unlike Shanghai, Guangzhou’s roads are not well separated between cars, bicycles and mopeds. In China, these road users are unpredictable at the best of times, often with little regard for traffic lights, road regulations or their own safety. XNGP seems to be stuck due to lack of dedicated or isolated channels. Of course, this was last year, and the system has likely been significantly improved since then.
forward thinking
Going forward, data will be a determining factor in the speed of change and system capabilities, and Lee may have a winning edge here. Xpeng Motors’ XNGP is only available on the Max version of four models. Taking NIO as an example, all second-generation cars have the necessary hardware, but users need to pay the equivalent of $530 per month to use the system.
In contrast, Li’s system comes at no charge and comes standard on all L9 and Mega cars. The L7 and L8 are available in AD Max and AD Pro versions, the latter lacking lidar but still offering NOA Highway. Considering that Li Shufu has sold nearly 500,000 second-generation cars, with 50,035 cars sold in December, while Xpeng Motors and NIO sold 20,115 and 18,012 cars respectively, this may help the company with the captured Big data builds leadership.
However, in December last year, NIO launched its first independently developed self-driving chip, which will be used in the ET9 flagship sedan launched in 2025. The 5-nanometer chip, called the NX9031, has more than 50 billion transistors and supports a 32-core CPU, which is said to be comparable to four Nvidia Drive Orin X chips.
In January this year, Li Auto announced that it would use Nvidia’s Drive Thor self-driving chip in its 2025 next-generation electric vehicles as the successor to Drive Orin. Drive Thor is said to have 2,000 TOPS performance, eight times that of Drive Orin.
Finally, apart from the advancement of chip technology and the rollout of China’s independent coverage, it is clear that Asian brands will not be content to stay in their home countries. Last month, Xpeng Motors, which has already entered Europe, confirmed its plan to bring its autonomous driving technology to the world by 2025. “We look forward to making the autonomous driving technology that Xpeng Motors has launched in China available to overseas users,” said He Xiaopeng, the company’s founder and CEO. said the CEO.
Xpeng Motors’ ambitions are not limited to its own cars. In July last year, Volkswagen announced that it would invest US$700 million in Xpeng Motors and acquire 4.99% of the company’s shares. It plans to cooperate with Xpeng Motors in 2026 to develop two Volkswagen brand electric models targeting the mid-size car market in the Chinese market.
The contrast between Xpeng and Apple’s now-defunct Project Titan (both founded 10 years ago) couldn’t be starker.