What is the progress of autonomous driving technology from 2018 to 2019?
The answer given by waymo, the industry leader, is: officially launch the official app & ldquo; waymo one & rdquo; in the Apple App store, users can directly & ldquo; call & rdquo; self driving taxis in Phoenix City after downloading and registering. So far, waymo’s robotaxi is more and more close to the ideal appearance, similar to the existence of Uber and didi.
Even if the current users who download and register need to wait for the company’s consent to use the service, this does not prevent waymo from being regarded as an important “milestone” in the industry for the commercialization of automatic driving;. In the same period of 2018, waymo launched the world’s first automatic driving commercial car Hailing service for the passenger side, and only a year later, the company has completed the transformation from an open trial operation to the application of IOS system for the public online & mdash; & mdash; automatic driving technology seems to have developed very smoothly.
But if you look back from abroad, you will find that this technology has produced a series of unexpected changes in 2019. Compared with the industry situation in 2018, these changes may have caught many practitioners by surprise.
Capital has the most acute sense of smell, so it is the first to change its attitude. From 2018 to 2019, the market of self driving venture capital gradually changes from “busy” to “calm”;.
In terms of the number of financing companies, in 2019, a total of 25 Chinese auto driving parts and solutions suppliers announced the completion of financing, 8 fewer than the 33 in 2018.
The result of the decrease in the number of financing enterprises is directly reflected in the total amount of financing. According to the statistics of yiou think tank, in 2018, the total financing of China’s auto driving parts and solutions suppliers reached 16.23 billion yuan, three times that of 2017 (5.37 billion yuan), even 20 times that of 2016.
But the upward trend did not last. In 2019, the total amount of financing in this field dropped to 10.71 billion yuan, a year-on-year decrease of 34%. The type of enterprise with the highest amount of financing changed from L4 level full stack solution supplier to subdivided industrial chain technology supplier. From the perspective of single round financing amount of start-ups, few enterprises can break the threshold of “a round exceeds 100 million yuan”.
The main reason is that in 2019, capital’s focus on emerging companies of autonomous driving has been significantly reduced. According to the statistics of yiou automobile, of the 27 financing events of automatic driving in 2019, only 2 financing rounds are angel or seed, accounting for 7.4%. In 2018, the figure was 21, accounting for 28.9% of the total.
Compared with 2018 (8.5%), in 2019, the financing events of more than B rounds in the field of automatic driving accounted for 22.2% of the total events, showing a significant increase, which shows that capital pays more attention to the industry head companies. It can be predicted that the trend will continue in 2020, and the survival difficulty of small and medium-sized auto driving enterprises will increase.
Unlike before, the new regulations cut off the path of bank funds entering the venture capital industry and capital market, which to a certain extent led to a sharp drop in the number and amount of funds raised in 2019. The overall profit-making effect of China’s venture capital industry is not obvious, which makes the “lack of money” of VC and PE firms more serious. They began to hold on to the “money bag”;. This has a great impact on enterprises such as autonomous driving, which have a long path of commercialization and are still in the embryonic stage of exploration & mdash; & mdash; capital has no patience to accompany the growth of start-ups, waiting for the maturity of high-difficulty technologies, and it turns to pay more attention to enterprises that are easy to achieve commercialization.
In a sense, this environment is also changing the way of commercialization of automatic driving technology.
Scene changes & nbsp;
Before 2019, the first-class enterprises in China’s auto driving circle are L4 level auto driving enterprises committed to building the robotaxi team.
At that time, there were several “despise chains” in the driving circle;. For example, in terms of the place where the enterprise is founded, the Silicon Valley in the United States despises the domestic establishment; in terms of the R & D scene, the open road despises the closed scene; in terms of the goal of the automatic driving level, the L4 level despises the L4 level or below; in terms of the speed, the high-speed automatic driving despises the low-speed automatic driving; in terms of the model, the passenger car despises the commercial vehicle … all of these, aiming at building L4 class self driving passenger car enterprises in the open road scene, is undoubtedly standing at the top of the chain and overlooking the existence of all living beings.
At that time, capital and related enterprises believed that robotaxi would come one day, and they could all share a share in it;.
Therefore, even though there are many emerging self driving enterprises in 2018, it is the enterprises planning to build robotaxi that can get large amount of financing. Among the self driving related enterprises that completed financing in that year, the two enterprises with the largest financing amount belong to this type, they are Xiaoma Zhixing and roadstar.ai respectively. The former completed two rounds of financing successively in one year, with a total amount of US $214 million, while the latter set a record for the highest financing amount of round a of China’s automatic driving enterprises with a single round of US $128 million.
The change took place in 2019, and the enterprises aiming to build robotaxi team began to & ldquo; fall down the altar & rdquo;. In that year, only a few head companies completed financing, and compared with 2018, the amount is not too large. For example, Xiaoma Zhixing has completed financing of US $50 million, and Wenyuan Zhixing has received tens of millions of US dollars. What’s more tragic is that the former star company roadstar.ai failed due to internal strife, which is lamentable.
In contrast, self driving start-ups that were previously at the bottom of the “ disdain chain ” in low-speed, closed or semi closed scenarios began to emerge frequently “ in 2019 ”. For example, it aims at the scene of the mining area, such as Xingge intelligent travel, huituo intelligent, focuses on the main line technology of the port or high-speed logistics scene, Tucson future, etc.
All of a sudden, things seem to be getting easier & mdash; & mdash; businesses that are most likely to commercialize in the shortest possible time can get financing.
The gradual tightening of capital & ldquo; money bags & rdquo;, in particular, has led some self driving enterprises that planned to build L4 passenger cars before, to change the target scene from an open road to a closed or semi closed location, or to change the target model from a passenger car to a commercial vehicle, in an attempt to enhance their hematopoietic power.
The huge investment even makes Baidu, a technology giant, overburdened, open a multi-channel parallel mode, and develop L4 level automatic driving solutions, as well as autonomous parking solutions in the parking lot scene.
At the same time, China began to walk out of a characteristic road of automatic driving & mdash; gradually turning from single vehicle intelligence to vehicle road coordination.
The concept of vehicle road coordination is not new. It has a history of several decades since it appeared. It refers to the overall planning of the information of people, vehicles and roads through the transformation of vehicles and road terminals, with the help of wireless communication and the new generation of Internet technology, so as to achieve the effective coordination of the three, reduce the occurrence of traffic accidents and improve the traffic efficiency.
However, since the birth of vehicle road collaboration, it has been tepid until 2019, when many enterprises were blocked in the way of single vehicle intelligent exploration, vehicle road collaboration began to “explode”;. For a while, & ldquo; we need smart cars, more intelligent roads & rdquo; and so on.
This is largely due to the current environment & mdash; & mdash; capital tightening, enterprises waiting for blood transfusion, both of which show extreme desire for commercialization. The 5g technology, which is close to the outbreak point, and the local governments which pay more and more attention to the automatic driving technology, are the important driving forces for the change of China’s automatic driving master strategy.
On the other hand, from a technical point of view, in the past decades, generations of autonomous drivers have focused on the research and development of single vehicle intelligence, but even so, autonomous vehicles still encounter a variety of unexpected situations (corner cases), many problems are difficult to be solved. For example, there are blind areas of vision and high-precision map can not be updated in real time.
In addition, the development and application of this technology are restricted by the delay in reducing the cost of refitting, which makes the single car intelligence can only be demonstrated as demo at this stage and can not really achieve mass production.
This needs the assistance of vehicle road coordination technology. By rebuilding the road infrastructure, many autonomous driving vehicles can share the intelligence of roads. From a technical point of view, adding sensors on the roadside can form a “God’s perspective”, eliminate the blind area of vision, and at the same time, consider the operation of all vehicles in the area, plan the best driving path and make the best decision. In terms of cost, the vehicle road collaboration technology can reduce the number of sensors and reduce the cost of the system for the automatic driving vehicle &ldquo, &rdquo.
In China, a strong government has made it possible to significantly transform infrastructure. Baidu, Huawei, Ali and other technology giants have focused on vehicle road collaboration technology. Although they claim not to build cars, they still want to take a place in the field of smart cars.
The visible future & nbsp;
In 2019, China’s automatic driving technology and pattern have quietly changed at many levels. The development of technology is not as expected. Both the capital and the self driving enterprises begin to pay more attention to the commercialization.
The activity of venture capital market is far less than that of that year, and the industry’s view on automatic driving technology has become pessimistic. As the technology enters the “calming down period”, the self driving enterprises also begin to “shuffle”, and the cold winter trend will continue for some time. In this process, the capital continues to concentrate on the head companies. The capital no longer believes in PPT, but only the start-ups with favorable development, hematopoietic capacity and prospects can survive.
It can be predicted that in the next 2-3 years, the automatic driving technology in low-speed or closed scenes will take the lead in commercialization, and the automatic driving technology in high-speed or open road scenes will experience a trial operation period. Since then, its dependence on the government and vehicle manufacturers will gradually strengthen.
With the help of the government, the vehicle road coordination technology with Chinese characteristics will flourish, which may drive the development of infrastructure related enterprises and generate many investment opportunities.
Nowadays, the situation of China’s auto driving industry is more like a quiet period after revelry. Sudden changes have caught many enterprises unprepared, and “low key” has become the theme. All enterprises are hibernating for the winter, waiting for their spring.