Written | Wang Pan Leng Zelin
Edit the | Wu Xianzhi
This year, the global autonomous driving industry has accelerated its landing, and foreign regulations have been greatly liberalized.
On April 28, the Beijing Intelligent Connected Vehicle Policy Pilot Zone officially issued the "Unmanned Manned Test License", and Baidu became the first batch of approved enterprises in China.
It is reported that Baidu's self-driving travel service platform "Radish Run" will carry out unmanned autonomous driving manned demonstration applications on open roads within 60 square kilometers of Beijing's Economic Development Zone. Photon Planet learned that the operation time of unmanned autonomous driving travel services is 10:00-16:00, Baidu first invested 10 vehicles, and follow-up plans to add 30 more unmanned vehicles.
Beijing took the lead in loosening the policy of unmanned driving, which is of extraordinary significance to the industry. The Robotaxi platform, represented by Radish Run, has been operating normally throughout the country for a long time, and related enterprises have already accumulated mature technology and operational experience.
At present, the accumulation of technology and operational experience has been quite sufficient, everything is ready, only owe Dongfeng, this opening may mean that relevant policies and supporting regulations will be accelerated.
People are the biggest security risk
The "Unmanned Manned Test Permit" does not completely invite the safety officer out of the car, but "there is no one behind the steering wheel", and the safety officer will still sit in the co-driver.
As we all know, any iteration of the automotive industry must comply with safety, based on this, it can be understood why after Baidu got the test license on April 28, Robin Li mentioned security issues in the circle of friends. As can be seen from the screenshot of the circle of friends, he reprinted a report on open licensing and wrote: Unmanned driving will definitely have accidents, and even fatal accidents, but this probability is much lower than that of someone driving.
Driverless cars are certainly not 100% safe, but they may not be as dangerous as the public perception suggests. The "White Paper on Traffic Safety for Autonomous Vehicles" jointly released by China Automotive Technology and Research Center Co., Ltd. on December 16 last year pointed out that talent is the most uncertain factor in traffic safety.
More detailed data comes from the CIDAS (China Traffic Accident In-Depth Survey) database.
The data shows that in the past 10 years, 5664 accidents involving passenger cars, the driver's human factor accounted for 81.5%. Among them, because the driver cannot identify and perceive the danger in advance, the subjective error of the driver causes 79.9% of the accidents. Accidents due to failure to give way according to regulations accounted for 43.4%, followed by excessive speed, illegal use of lanes, drunk driving, violation of traffic lights and fatigue driving.
People can only rely on existing experience to judge, but everyone's driving time, the scene encountered are not the same, driving technology is uneven, which makes people the biggest variable in traffic safety.
Zhang Yaqin, academician of the Chinese Academy of Engineering, chair professor of intelligent science at Tsinghua University, and dean of the Intelligent Industry Research Institute, once said in an interview: "Based on the general consensus of the industry, the safety requirements for automatic driving are at least one order of magnitude higher than that of human driving, reaching the level of 99.99999%. To achieve such a safety goal, it is necessary to deep learning and vehicle-road collaboration, supported by these two major directional technologies. ”
Through the continuous "feeding" of the database and the evolution of AI itself, coupled with the continuous iteration of hardware such as vehicle-road collaboration, automatic driving will be a weapon to solve the uncertainty of "people". Under The circle of friends of Robin Li, Zhang Yaqin also reiterated that the probability of unmanned accidents is "at least an order of magnitude lower."
Technology is ahead, policy is lagging behind
Domestic autonomous driving started late, even the oldest Baidu, it has only been 9 years.
In 2015, Baidu set up a driverless division, at this time in the United States, a blind man sat in a car without human intervention control and safety officers, and passed through downtown Austin - this is a new "three noes" (no steering wheel, no brakes, no throttle) pure electric fully autonomous vehicle launched by Google.
Two years later, Waymo's test mileage on the actual road has reached 1368 kilometers, and what is even more terrifying is that the team will import the actual road test results into the simulated environment and let the driverless system continue to practice in software. On the other side of the ocean, Robin Li sat on the five rings of the unmanned car and ate the industry's first ticket, and Beijing opened the manned cargo test until the end of 2019.
Today, with Baidu, Didi, Xiaoma Zhixing, Wenyuan Zhixing and other companies in the past few years continue to invest heavily in the field, there are companies from start-ups to maturity, China in many ways has even caught up with the United States, one of the characteristics is "increase the amount without price increase".
Taking the fifth-generation Robotaxi as an example, through further integrated design and greatly improved operational stability, Baidu has compressed the cost to 480,000 yuan, about 1/3 of Waymo. Photon Planet understands that the next generation of Robotaxi will also be released this year.
Moreover, as we all know, due to the degree of population density, differences in urban planning and construction, the challenges faced by domestic driverless companies are much more stringent than those of foreign counterparts, if Waymo's test environment is ordinary difficulty, then Apollo has not only tested under the difficulty of hell for a long time, but also maintains a zero accident rate so far. Under the tempering of the complex environment, domestic enterprises have a more solid accumulation of technology.
Chinese and American enterprises are the two poles of driverless technology, and the competition between domestic enterprises and foreign enterprises is not just a simple technical competition, but a multi-dimensional competition covering capital investment, fleet operation scale, and even macro-level laws and regulations.
At present, the United States has not yet allowed Waymo, Cruise and other companies to charge driverless orders, and the country has started commercial operations. From a commercial perspective alone, China has also completed overtaking in curves.
Some giants have taken the lead in launching driverless travel services in Beijing, such as Turnip Run, which has launched autonomous travel services in nine cities, including Beijing, Changsha, Guangzhou, Shanghai, Chongqing, Shenzhen, Yangquan and Wuzhen. Among them, in Beijing, Chongqing and Yangquan, commercial toll operation services have been operating for a long time. According to Baidu's 2021 annual report, as of December 31 last year, the number of manned orders has reached 213,000.
In fact, Waymo allowed testing of self-driving cars on public roads as early as 2012, but today's manned operation areas are only California and Arizona, with a single quarter order peak of 52,000 orders, only 1/4 of Baidu.
If technology is the first half of autonomous driving, then commercialization and scale operation is the second half of the industry, and the premise of driving into the second half is that it must give the "green light" in the formulation of policies and laws.
Although many American companies have temporarily lagged behind domestic enterprises in terms of cost control, commercial orders, and technology investment, they have always been ahead of technology in terms of laws and regulations.
On March 10, the National Highway Traffic Safety Administration (NHTSA) issued a 155-page "final rule" that removes several routine control requirements for vehicles equipped with autonomous driving systems and driverless cars.
48 days later, 12 U.S. lawmakers wrote to U.S. Secretary of Transportation Pete Buttigieg urging him to develop a comprehensive federal framework for self-driving cars.
In addition to the United States, Germany became the first country in the world this year to allow driverless vehicles to participate in daily traffic and be applied nationwide; the United Kingdom amended the Road Law to allow owners to watch movies in the car in self-driving mode.
Although domestic driverless technology is running in small steps, the policy relaxation is not obvious and extremely prudent. Professor Zhu Xichan of the Automotive College of Tongji University said that laws and regulations and standards have not kept up with the development of technology.
Deng Zhidong, professor and doctoral supervisor of the Department of Computer Science of Tsinghua University and an artificial intelligence expert, recently called for: "It is expected that our policy planning and deployment will not only be able to keep up with the rapid development of autonomous driving in the United States and Europe, but also be more conducive to the commercialization of all unmanned vehicles on the road and the autonomous driving industry as soon as possible." ”
Driverless driving will drive industry and consumption
Industrial development relies on technology, but the progress and scale of development are closely related to policies.
In fact, cooperation and integration have long been the norm in the new energy industry. The reason for the rapid development of domestic new energy vehicles is largely due to policy support, so we have reason to believe that intelligence can reproduce the path of electrification and bring new kinetic energy to the automobile and even other industrial chains.
Through the above figure, we can see that the complete industrial chain of the Robotaxi industry, involving IT services, cloud computing capabilities, road infrastructure companies, to vehicle manufacturers, sensing devices, chips, network suppliers, and then to technology companies that provide algorithms, autonomous driving capabilities have initially formed a long high-tech industrial chain.
The liberalization of autonomous driving policies may have a "butterfly effect". Under the wave of new infrastructure, smart transportation can not only play a role in stimulating domestic demand, but also help to improve traffic problems in large cities and reduce traffic accident rates.
Robin Li has previously said that if intelligent transportation is realized, within 10 years, the problem of congestion will be basically solved by relying on the improvement of traffic efficiency; the intelligent transportation system built by new technologies, new concepts and new models may reduce traffic safety accidents by 90%.
Every time a new technology is born, a new business model is created and a new job is created, and autonomous driving is no exception. New forms like Robotaxi will create new jobs such as security officers, cloud drivers, technicians, and testers, contributing to the transformation of the workforce structure.
If we use our imagination to continue to expand, the complete display of autonomous driving capabilities will not only liberate the driver, but also release the vitality of the entire vehicle ecology, especially the pull on social consumption.
According to the data released by the Development Research Center of the State Council, "Effective expansion of automobile consumption needs to focus on three growth points", the market size of automobile commodities exceeds 4 trillion yuan, accounting for 10% of the total retail sales of social consumer goods. The intelligence represented by unmanned driving not only stimulates the demand for new cars, but also may also give birth to a scene that has never been seen on the PC and mobile ends in the limited space of the car.
NIO's latest delivered model ET7 will support passengers to wear AR/VR glasses in the car to enjoy audio and video entertainment functions, the entertainment system of the next model of the ideal car, L9, will support Switch and other game hosts to connect to the screen, and then at the model S Plaid conference, Musk also used the car machine to demonstrate a "cyberpunk 2077". In addition, Musk also revealed on social media that Tesla hopes to move the entire Steam library to Tesla, rather than traditionally porting one game after another.
Obviously, when the driver is "killed", not only will a Focus media be born in the interior space of the car, but it may be the best undertaking and extension of the mobile and PC ends.
Driverless driving not only frees the driver's hands, but also brings iterations of consumption scenarios – such as delivery to the car, cockpit entertainment, and so on. There is no doubt that in the era of AI, the car will still be an important engine to drive consumption.
And now, everything is ready, only the policy east wind is owed.