Chinese L2 Self-driving and Robotaxi vehicles and Tesla FSD Review

1) She tested L2 self-driving of major brands like Huawei, Li Auto, NIO, Xpeng, and Xiaomi and they exceeded expectations overall. The rides were not overly cautious and handled complex situations quite well considering the road conditions in China are very challenging.

1) She commented it was nothing compared to Tesla's approach and saw imitation learning/end-to-end as the only effective approach for self-driving. While the Chinese peers performed well on main roads, they struggled on frontage roads due to reliance on high-precision maps and rule-based methods e.g. cars stopped in the middle of the road where there was no clear white lining.

3) Chinese EVs' self-driving capabilities are far ahead of those from US and EU brands but she doubts any Chinese players can profit from the L2 self-driving, not because it’s not useful, but because it is hard to differentiate as price dominates the market.

Note: You can watch the video she uploaded in the post and the L2 self-driving screen looks almost the same. I doubted that their capabilities are far ahead of those from US and EU but the take-up rate is.

4) Chinese consumers and regulators seem much more receptive to self-driving. For the insurance, L3 cars, Original Equipment Manufacturers (OEMs) bear responsibility for incidents hence OEMs avoid labeling cars as L3.

Note: This is the same approach used by Tesla FSD (Full Self Driving).

5) She tested robotaxi of major brands like Pony, Didi, and Baidu and rated Pony equal to Waymo which was ahead of its peers. The same issue applies here: user experience is nearly perfect in Yizhuang, Beijing but expansion is the real question.

6) Chinese robotaxi companies are very sophisticated. While the rest of the world focuses on technology, Chinese peers treat it as a product, considering unit economics, operations, mass production, etc. Interestingly, most companies expressed a preference NOT to operate the fleets themselves. They aim to be asset-light and let fleet managers handle operations.

Note: I take it robotaxi is outsourced and  operated by a separate company.

7) China's policy supports autonomous driving. It has a very clear approval process which is based on data such as autonomous driving distance, fully driverless distance and intervention rate, passenger ratings.

My Take:

1) If you check the writer's posts in X.com, Freda frequently posted reviews on technology and does not give investment advice. As someone who travels frequently in the US and China, I think the points she raised are worth taking note of.

2) Someone always claimed that Tesla FSD is only L2. In fact Tesla FSD is capable of L3 but declares as L2 which is strategically a smart move. It allows the vehicles to be used by everyone and almost everywhere in the US - no restriction or geo-fence. In return, Tesla collects data to be used to perfect the neural-network autonomous software.

3) Someone said the take-up rate of Tesla FSD is very small. According to Teslarati, Tesla announced in its Q1 2024 that FSD has broken the 1.3 billion-mile mark, and during the succeeding earnings call, Elon Musk said that the advanced driver-assist system has been pushed to about 1.8 million vehicles. From this number, Musk stated that about half are using FSD. This suggests that at the time of the first quarter 2024 earnings call, there were about 900,000 FSD users. The size of users which is huge compared to other EV brands is going to accelerate in the next quarter.

4) Nvidia CEO Jensen Huang believes Tesla’s FSD is the most advanced system out right now.
“Tesla is far ahead in self-driving cars,” Huang said in an exclusive interview with Yahoo Finance. "One of the things that's really revolutionary about version 12 of Tesla's FSD is that it's an end-to-end generative model," Huang added.

Source: