[? Billion euro guide? ]? The key turning point has not yet arrived, and the problem remains unsolved.

Author Zhang Nan

Editor Zhang Yan/Yang Yaru

Under the epidemic, all trades were depressed, but driverless technology unexpectedly gained attention.

Due to the infectivity of the epidemic, the unmanned business model was hotly debated for a while. In this "anti-epidemic", the value of unmanned vehicles has gradually emerged. In Wuhan Fangcang Hospital, unmanned vehicles shuttled back and forth. These vehicles have remote control devices, which can be used at any time to meet the needs of hospitals. In addition, the provider of the vehicle arranges engineers to monitor and respond at any time if necessary. ?

for a while, it seems that unmanned spring has come. Even in the cold winter of capital and the downturn of the industry, Pony.ai and Yushi Technology have successively announced the completion of new financing. The former announced that it has obtained financing of up to 462 million US dollars, and the leading investor is Toyota Motor Group; Although the latter has not announced the financing amount, the investor is also a well-known parts company Bosch. ?

On the world stage, China's autonomous driving technology seems to be gaining recognition. ?

Another good news is that on February 27th, in the "219 Autopilot Road Test Report Card" published by DMV of California Vehicle Administration, Baidu's MPI value (Miles? Per? Intervention, the average mileage traveled between every two manual interventions, surpassed Waymo, the "big brother" of autonomous driving, for the first time, ranking first in the list of 36. Among the top 1 companies, four are from China. "Autonomous driving has entered a new era in which China and the United States coexist at the same time." AutoX? CEO Xiao Jianxiong told Yiou Automobile.

after the bursting of the bubble in 219, the self-driving industry in China really needs good news.

but it may be too optimistic to say that spring has arrived like this.

Waiting for the "critical turning point"

For the autonomous driving industry in China, the "critical turning point" has not yet arrived. ?

in the 199s, domestic universities and research institutions began to develop autonomous driving technology. In 1992, National University of Defense Technology developed the first truly self-driving car in China, CIAVT-I, and the self-driving industry in China really started. ?

However, for more than 2 years, the autonomous driving technology still existed only in university experiments. Even until 29, in the "China Smart Car Future Challenge" sponsored by the National Natural Science Foundation of China, almost all the participating teams came from university campuses.

the first turning point was in 214.

In that year, Baidu, which was in urgent need of transformation, began to drive automatically, and Andrew Ng, known as "one of the four donkey kong in the field of artificial intelligence", joined in.

In order to meet Andrew Ng, Li Yanhong canceled several important meetings. After a three-hour dinner, Andrew Ng said to Li Yanhong, "I hope to help you." So I joined Baidu. Since then, he has become the chief scientist of Baidu, responsible for the leadership of Baidu Research Institute-including the autonomous driving business. During that period, Baidu absorbed a group of autopilot talents such as Lou Tiancheng, Han Xu, Peng Jun, Wu Xuebin, Ni Kai and Han Xu.

in September p>214, Baidu started cooperation with BMW in the field of autonomous driving. Although the three-year cooperation between the two parties ended early in 217, at that time, the strategic cooperation with German luxury brands attracted a lot of attention for Baidu.

The joining of Internet giant Baidu has accelerated the industrialization of autonomous driving in China. More and more people see the commercial value of this technology, including investors. ?

From p>216 to 217, it became the first explosion point of autonomous driving technology. ?

Many technicians left Baidu and set up their own autonomous driving companies one after another, such as Xiaoma Zhixing founded by Peng Jun and Lou Tiancheng, Heduo Technology by Ni Kai, and Main Line Technology in Zhang Tianlei. ?

For a time, the auto-driving industry in China was "full of flowers", which attracted investors from all walks of life to "throw money". This stimulated the development of the industry to a certain extent, and subsequently, relevant laws and regulations were introduced one after another. In this process, companies such as RoboTaxi have gained more capital favor and become the most concerned enterprise type in the industry. ?

but those crazy days are over. From the second half of 218 to 219, the development of China autonomous driving industry slowed down. Now, even if there is good news, there is no substantial breakthrough in many aspects. The key turning point of the new industry development has not yet arrived.

Difficult problems to be solved

Many difficult problems, such as commercialization, capital, supply chain and policy, which punctured the bubble of autonomous driving industry in 219, are still unsolved. ?

from the technical difficulty, the unmanned vehicles running in the shelter hospital are not the same as the L4 and L5 automatic driving that people expect. The realization of "open road scene" autopilot is much more difficult than that of closed/semi-closed scene autopilot-the former has complex scenes, many participants and extreme cases (corner? Case) has a high probability of occurrence.

At present, the immaturity and security risks of autonomous driving technology are also important factors that restrict the commercialization of technology and the development of the industry. ?

Gartner, a consultancy, published a "technology maturity curve" in 219. Unfortunately, the technical direction that many autonomous driving companies are striving for-L4 autonomous driving is in an "disillusionment period". Gartner believes that its technology will take more than ten years to realize. In addition, L5 autopilot is in the "expected expansion period".

Some insiders are more pessimistic. ?

"This technology is really difficult. It will take decades for self-driving cars to be popularized on the road." Waymo the end of 218, John Kolafik (John? Krafcik) once said. His voice did not fall, and another Silicon Valley tycoon-Apple co-founder Steve Wozniak (Steve? Wozniak) also expressed his opinion: "I don't trust self-driving cars."

considering the difficulty of technical breakthrough, Baidu Apollo, which has always praised the intelligent bicycle route, has also begun to increase the technical direction of "vehicle-road coordination". At the end of 218, Baidu opened the vehicle-road coordination code, and carried out the RoboTaxi passenger trial operation with V2X equipment on the streets of Changsha for several months. In 219, Baidu put 45 RoboTaxi vehicles into trial operation in Changsha test section, and at the "Apollo First Ecological Conference" at the end of the year, it highlighted the vehicle-road collaborative solution.

however, in limited scenes, the difficulty in realizing autonomous driving technology can not be ignored.

In the relevant areas, there are quite a few unexpected situations in the driving process of vehicles-vehicles and people walking at will, parking indiscriminately, etc., which will cause troubles for unmanned vehicles.

on the other hand, even in the same scenario, the technical adaptability of enterprises is not high at present. Unmanned vehicles that can run smoothly in A mining area may make mistakes frequently in B mining area. Enterprises still need to further polish their technology to meet the needs of large-scale commercial landing in the future. ?

Even if the MPI value of self-driving enterprises in China is equal to that of the United States, it does not mean that self-driving enterprises in China can achieve large-scale commercial landing. In addition, large-scale commercialization still needs to consider the cost issue.

in terms of capital, although two financing information have just been released, it is not optimistic to look at the whole industry. ?

The huge capital demand of each enterprise is in contradiction with the unexpected technological development and the increasingly cautious investment strategy of institutions, and the industry has begun to shuffle.

in 219, the number of RoboTaxi financing enterprises decreased, even the former star company RoadStar.ai was liquidated, and the American autopilot company Drive.ai, which Andrew Ng joined after leaving Baidu, also collapsed, which made it easier to realize the limited scene of application landing. Autopilot enterprises gradually attracted attention. ?

But after the inflation period, the valuation of autonomous driving companies has been raised. Under the epidemic situation and the cold winter of the industry, it is more difficult to raise funds in the primary market and the capital is more rational. Only enterprises with commercial signs are more likely to gain investment favor. "Investors will choose a landing scene that is just needed and has the ability of buyers (to invest)." Chen Longru, CEO of Huituo, told Yiou Automobile. In this case, autonomous driving companies that have not yet been able to make blood by themselves are struggling.

Zhang Dezhao, CEO of Zhixingzhe, told Yiou Automobile that there may be some large-scale landing of limited scene autopilot next year, but the cost performance is still the key to restricting large-scale application.

In addition, China's self-driving supply chain is immature and its policies and regulations are backward, which still need time to solve. ?

taking lidar as an example, even though many manufacturers have lowered the price of their products to below $1,, they still do not meet the standards for large-scale commercialization. At present, China only allows self-driving cars to carry out road test in some areas, which is far from California's policy of allowing self-driving cars to drive on the road, which is not conducive to technical iteration.

Although the sudden epidemic has attracted the attention of some autonomous driving companies, the uncertainty of the development of technology companies has been amplified because of its negative impact on the automobile industry.

Winter will always pass, but it is still too early to talk about spring.

Editor: Zhang Yan/Yang Yaru

This article comes from the author of ChejiaNo. in car home, and does not represent car home's standpoint.