The end game of autonomous driving depends on China and the United States.
Author | Liu Jane
Edit | Wen Liang
Autonomous driving has become a hot word in the hearts of the people.
In the past year, the government has issued a number of policies related to autonomous driving and new energy vehicles, all of which have promoted the healthy development of the industry.
On Friday, CPPCC members and people's congress deputies from all walks of life gathered in Beijing again to submit relevant proposals and suggestions, reflecting the problems in the development of the industry.
Among them, in the field of autonomous driving/smart cars, Baidu Robin Li, Xiaomi Lei Jun, 360 Zhou Hongyi and other big names from the technology industry have put forward valuable opinions, such as:
Robin Li believes that although mainland autonomous driving technology has led the world, it still faces many qualifications and responsibilities to clarify problems, and it is necessary to further break through the policy bottlenecks that are incompatible with the development of technology and industry, and stimulate innovation capabilities in the field of automatic driving; especially for large-scale commercialization of unmanned automatic driving.
In fact, the suggestions and suggestions of the representatives and members of the "two sessions" have profound significance - they are a microcosm of the current development of various fields in China, and some of the proposals will even affect the formulation of laws and policies.
Public information shows that the Ministry of Science and Technology, the Ministry of Industry and Information Technology and other ministries and commissions have seriously considered some of the proposals of last year's two sessions, some of which have been related to new energy vehicles and intelligent networked vehicles.
The proposal of the above-mentioned technology industry leaders is bound to accelerate the development of the mainland's autonomous driving and intelligent car industry.
The final sprint is racing against the clock
As a core component of the new round of industrial transformation in the future, the importance of autonomous driving is self-evident, and domestic and foreign countries are also seizing this new highland.
The ranking of the 2021 autonomous driving road test report released by the California Transportation Authority (DMV) is regarded as a reflection of the strength of autonomous driving players.
Take driverless testing, for example. As of November 19, 2021, a total of 8 companies have obtained California DMV full driverless road test licenses, with half of them from Chinese players.
However, only four of the eight companies submitted data in the California all-unmanned autonomous driving test, namely Baidu Apollo, Xiaoma Zhixing, Cruise and Nuro. Among them, Baidu ran the highest score in the total test mileage (nearly 10,000 miles) with only 3 test cars, and there was no disengagement, and the performance was very good.
More importantly, China's urban roads are more complex than those in the United States, such as congested and narrow roads, pedestrians crossing the road, and a variety of special-shaped cars, which are all challenges that cannot be ignored for self-driving cars.
Tempered by Chinese-style road conditions, the technical strength of domestic players has been at the forefront of the world together with the United States.
Especially at the critical node of the opening of the second half of automatic driving, the industry has moved from technical verification to large-scale commercial landing, and it is necessary to further break through the policy bottlenecks that are incompatible with the development of technology and industry, and stimulate the innovation ability in the field of automatic driving.
In the first half of the competition for autonomous driving, the United States, with its first-mover advantage, carried out the road test of autonomous driving that allows enterprises to be equipped with safety personnel in 2014, and in the following years, it continued to support the development of the industry through relevant policies, and further issued a fully driverless road test and commercialization license in September 2021.
At the same time, due to the practical experience and test data accumulated in the field of autonomous driving in the past decade, California's regulatory mechanism for autonomous driving companies has gradually improved, and it has greater confidence in the technical level of these companies and is willing to give them greater freedom, such as allowing licensed companies to conduct tests without safety officers and carrying out unmanned commercial operations.
However, the mainland's policy will be relatively conservative. According to the provisions of the current relevant laws, domestic enterprises still need to retain the co-pilot safety officer when conducting unmanned road tests; if they need to carry out commercial operations, they must retain the main driving safety officer.
In addition, domestic policy support for the autonomous driving industry is no different from that in the United States.
Do a good job of "small me" and achieve "big me"
Robin Li's proposal is not a castle in the air. Baidu's long-term accumulation of operations, data, and technology in the field of autonomous driving is precisely where its strength lies.
Today, Baidu Apollo Day was officially held in Yizhuang, Beijing.
In order to highlight the strength of the unmanned vehicle test team, Xu Baoqiang, director of Baidu's automotive robot department, broadcast the daily life of the test team at the event, including easy and comfortable passage through the intersection with the most complex traffic environment and the highest density of people and traffic in Yizhuang.
Such road scenes, which are very common in China, are very rare in the United States.
Generally speaking, urban planning in Europe and the United States is presented as a small area of urban core area, residential and living areas are scattered and sparse, the overall road scene is more standardized, and the traffic participants are relatively small; while the typical urban scene in China is presented as a large-area urban core area, the high-density mixed distribution of living areas, commercial areas and residential areas, and congestion during the peak travel period is a normal phenomenon.
In addition to different urban planning, in the road scenes of Chinese cities, the problem of mixed traffic between people and vehicles is serious, especially the large number of express delivery cars and takeaway cars, which puts forward higher requirements for the recognition ability and response ability of the automatic driving system.
In Beijing Yizhuang, for example, the density of urban road traffic participants is 15 times that of California. This is also why Xu Baoqiang pointed out that it will be more challenging for autonomous driving to achieve large-scale unmanned deployment in China.
In fact, the reason why Baidu's test team has been able to be so comfortable in the face of challenges is inseparable from the accumulation of bit by bit in the previous years.
According to Wang Yunpeng, vice president of Baidu and general manager of the autonomous driving technology department, Baidu Apollo's autonomous driving test mileage has exceeded 25 million kilometers, and the unmanned test has driven from the Shougang closed park to the social road in the core area of Yizhuang, and the commercial toll operation has also been applied from the park to the public's daily travel services.
Specifically, Radish Run has served eight cities so far, and the order volume is growing rapidly, with 213,000 orders completed in the fourth quarter of last year alone, doubling compared with the third quarter, and receiving more than 300,000 orders in half a year. In just half a year, the number of vehicles and stations in Yizhuang has increased by more than 50%, and the peak number of daily bicycles has exceeded 28.
It can be seen that as a service platform focusing on short-distance travel, Radish Run has achieved a bicycle efficiency close to the existing level of special cars, and its operational capabilities far exceed Waymo, ranking among the world's largest autonomous driving travel service providers.
At the same time as the scale of landing continues to expand, the commercialization of Radish Fast Running is also advancing in an orderly manner, and commercialized charging operation services have begun in Beijing, Chongqing and Yangquan, continuously bringing cash flow to Baidu Apollo and strengthening self-hematopoietic ability.
The data collected by Radish Run from different cities, different locations, and different scenarios can further feed back the iteration of Robotaxi.
Baidu opened its unmanned vehicle service to the public in Shougang Park in May last year, taking the lead in entering a new stage of unmanned driving. Subsequently, the journey of unmanned testing was launched on the open road of Yizhuang.
Such unmanned progress is not only the leading level in China, but also the first echelon even in the world.
In order to better provide driverless services to the outside world, Baidu Apollo Moon came into being, in addition to the most basic driverless capabilities, it can also bring passengers a personalized and comfortable ride and interactive experience. Moreover, compared to the cost price of millions of autonomous vehicles in the industry, Apollo Moon only needs 480,000 costs.
It is worth noting that the increasing test mileage, the continuous expansion of the landing range and the gradual withdrawal of safety personnel are behind its thinking on "safety first" - after all, if it cannot complete the guarantee of safety, the development of the autonomous driving industry will become unsustainable.
Xu Baoqiang introduced that the bicycle intelligent main system, redundant monitoring system and parallel driving system are the three core methods for Baidu's autonomous vehicles to achieve safe operation.
Among them, the bicycle intelligent system can turn more unsafe scenarios into safe scenarios, while the redundant monitoring system can fully identify the system status and risks, turn unknown scenarios into known scenarios, and intervene in advance when necessary.
The parallel driving system is the so-called 5G cloud driving. As a remote driving solution, 5G cloud driving is an important supporting service for autonomous driving, based on new infrastructure facilities such as 5G, intelligent transportation, and V2X, which can provide the bottom for the unmanned operation of the Apollo fleet.
In addition to continuously improving its own technology and polishing the autonomous driving fleet, of course, Baidu, as the leader of the domestic automatic driving industry, has also been personally involved in the formulation of relevant autonomous driving road test policies, autonomous driving technology capability evaluation and automatic driving operation service management system research, supporting the introduction of the first domestic automatic driving policy and building the first set of automatic driving evaluation standard system and management system.
In November last year, Baidu also released the "White Paper on Traffic Safety for Autonomous Vehicles", actively exporting safety concepts and safety methods, hoping to promote common progress in the industry.
On the other hand, Baidu understands that the development of autonomous driving is not a single battle, and has worked with partners to build the world's largest open ecosystem for autonomous driving.
It is conceivable that not only for its own progress, But also for every step of Baidu's automatic driving forward can play a certain degree of industry value. Robin Li has a full say in the development status of autonomous driving, and his prediction of future trends is of great reference significance to the direction of the industry and even the introduction of policies.
The Whampoa Military Academy
As the most disruptive technology of the 21st century, autonomous driving requires not only pathfinders and navigators, but also evangelists.
At the same time, in the process of moving towards the end of automatic driving, it is also necessary to be in the right place at all times, that is, the development of the times, the support of policies, and the accumulation of talents are indispensable.
Baidu, on the other hand, is recognized by the industry as the Whampoa Military Academy in the autonomous driving industry.
In 2013, Baidu was the first to start autonomous driving research and development in China, and by 2015, it had achieved automatic driving testing and verification in beijing's fifth ring road. This is the earliest road test of autonomous driving in China, and it can also be regarded as the starting point for Chinese technology companies to develop autonomous driving.
Since 2015, many technology giants have left Baidu and re-contributed in the field of automatic driving. Looking back, Baidu people in that year have occupied most of the country's autonomous driving field, promoting the development of the industry in their respective positions, and increasing the possibility of the mainland overtaking the United States in the automatic driving industry.
In fact, the "big bulls" who have fled are only a very small number, and more people have stayed in Baidu or are about to join Baidu to contribute to the full driverless.
It is understood that the high-T personnel of Baidu's automatic driving team are not small, there are currently at least 80-100 people, "these small partners are like "sweeping monks" to do things steadily, "sweeping monk spirit" is also Baidu Group's top-down work style. ”
Soon, Baidu will also have a large number of students recruited by the school, which is twice as many as the total number of school admissions in the past three years.
Baidu's school recruitment this year is very hot, but it should be noted that the current shortage of talents in the industry is still relatively serious.
According to the survey data of a number of job search software, there are only 30,000 technical talents in China with the labels of "automatic driving" and "new energy vehicles"; by 2025, the stock of R&D personnel in the mainland's intelligent and connected car industry will be about 72,000, and the net talent gap will reach 10,000-40,000.
In addition to internal talent training, Baidu is also actively responding to the national call. For example, we will jointly cultivate talents with institutions, universities and universities to send high-quality talents to the industry.
In addition, with the inclusion of intelligent networked vehicles in the "14th Five-Year Plan" development plan, the Ministry of Education has added a "smart vehicle engineering" major in the "Catalogue of Undergraduate Majors in Ordinary Colleges and Universities", and many colleges and universities, including Harbin Institute of Technology and Chongqing University of Technology, have opened this major.
In the foreseeable future, more people with lofty ideals will join this track with huge imagination space, and Baidu will become a good place for these talents to jointly contribute to the overtaking of China's autonomous driving industry.
END
Autonomous driving, may not have entered the "second half"
Who is the performance king of China's intelligent driving middle and upper reaches of the industry chain?