Asian Tech Press (Mar. 16) -- Self-driving car technology has been getting a lot of attention worldwide, and recently a news story brought Alibaba's autonomous driving technology into the public eye, sparking widespread discussion in China.
Sources familiar to the matter disclosed last Friday that the self-driving logistics vehicle Xiaomanlv, which Chinese e-commerce giant Alibaba Group Holding Ltd's research division DAMO Academy had previously launched with great fanfare, drove directly into a wet concrete field that had not been filled for long at a university campus in China, and "flopped" on the spot, causing students to laugh and crowd around.
The "crash", which seemed to be caused entirely by technical problems, triggered unrelenting criticism from Chinese engineers.
Some said, "The tiny car is equipped with a leading domestic company's 32-channel LIDAR, but still run at such a level, (I) admire extremely."
An engineer believed that there is an algorithm problem, or that "the algorithm for small cars are relatively simple, not too technical in nature."
Some also helped Alibaba explained, "This can not be blamed completely on the small car, because it is an uncommon corner case, as is the passenger car, which should be processed in a targeted manner. Especially when the cement is dried to 80%, there is no reflection, it is likely and will not be judged as water."
Professional analyses say that errors occurred in all of the following three areas to cause the car to drive directly into the concrete:
1. the camera and ordinary collision avoidance radar couldn't identify the road conditions (it was unclear if the mini logistics vehiclehad a normal radar);
2. the 32-channel LIDAR did not play any role;
3. the map for the area was not updated in time, and corrected according to the actual situation.
There is a potential market for commercialization in the area of self-driving logistics vehicles in China, but it remains difficult to leapfrog the technical difficulties focusing on the human–machine interaction (HMI) level.
China's self-driving technology is pioneered by the search engine giant Baidu Inc., the first local company to lay out its autonomous driving business.
Alibaba started technical investment in the self-driving logistics vehicle Xiaomanlv in 2018, which was unveiled and put into mass production and commercial operation in 2020. During the Singles Day, or Alibaba's "Shopping Carnival" in 2021, Xiaomanlv was introduced to more than 200 colleges and universities and completed up to one million deliveries.
However, unlike commercial vehicles, the number of sensors used by the Xiaomanlv logistics robot is the least in the autonomous driving industry, and the cost of the vehicle is less than a third of the industry average. This means that its research costs would be lower than other self-driving vehicles, and the technical investment would be less.
Therefore, compared to real autonomous driving technology, Alibaba's Xiaomanlv robot may still be lacking in some way.
The ongoing research on autonomous driving technology in China is still at the stage of cash-burning, hardware stacking and technical parameters battling, which requires relatively significant investment.
The DAMO Academy, which was asked by Alibaba Group to be self-sustaining at its inception, has been seeking low-cost deployment and early commercial use in developing its autonomous driving business.
Alibaba's self-driving business has faced more than just Xiaomanlv's sudden "flop", which has put it in trouble. Wang Gang, a former vice-president of Alibaba and head of the e-commerce giant's autonomous driving laba at the DAMO Academya, announced his departure and commencement of his own business in January 2022.
With the "flop" of self-driving product Xiaomanlv and the loss of a tech senior executive in the self-driving business, Alibaba seems to have encountered quite a few obstacles on the race track, and future appears uncertain.
Nonetheless, Alibaba has been continuing to put efforts in the autonomous driving field. Billionaire Jack Ma's e-commerce empire led in September 2021 an investment in DeepRoute.ai, a leading Chinese L4 autonomous vehicle startup, joining other investors to contribute to the self-driving firm's $300 million Series B funding round.
A source has revealed, "Since Alibaba itself has little experience and layout in autonomous driving, that's why it chose to invest in a startup whose valuation is not yet ridiculously high."
As the penetration rate of autonomous driving cars continues to go up, Internet companies like Alibaba may increase their capital investment in the business, aiming for success in the new track.
