Author 丨 Song Doudou, Zuo Maoxuan
Editor 丨 Zhang Mingyan
Figure Source 丨 Figure worm
The commercialization of autonomous driving has taken a key step.
On April 28, the "Implementation Rules for the Management of Unmanned Road Testing and Demonstration Applications for Passenger Cars in the Beijing Intelligent Connected Vehicle Policy Pilot Zone" (hereinafter referred to as the "Implementation Rules") was officially released, and Baidu and Xiaoma Zhixing were approved as "Unmanned Manned Demonstration Application Notice" and were allowed to provide the public with autonomous driving services (Robotaxi) with "no safety officer in the main driver's seat and safety officer in the co-driver's seat".
Compared with the previous "self-driving taxi" equipped with safety personnel in the main driver's seat, this time it is a pilot to open unmanned manned people with safety officers in the co-driver. This means that the "no one behind the wheel" Robotaxi service is being liberalized for the first time in China's megacities.
"We expect that before the end of this year, Beijing will introduce a new policy, not only allowing the co-driver to be unmanned, but also to move the co-driver safety officer to the back row, and even further introduce the policy of no one in the whole vehicle." It can be seen that the entire policy development direction is very positive for the industry. On April 28, Zhang Ning, vice president of Xiaoma Zhixing and head of the Beijing R&D Center, said in an interview with the 21st Century Business Herald reporter.
On the same day, Wei Dong, vice president of Baidu's intelligent driving business group, also said that according to the policy regulations, the co-pilot seat of the vehicle for unmanned autonomous driving travel services needs to have a safety officer. In the future, with the introduction of policies, the abolition of security officers will be gradually realized. "Yizhuang Management Association is studying the management method of all-driverless driving, and it is conservatively estimated that all-unmanned autonomous driving may be opened at the end of this year."
On the same day, the 21st Century Business Herald reporter also actually experienced the "main driving no one" radish fast running robotaxi in Beijing Yizhuang. During the experience, the reporter found that the vehicle was relatively smooth in driving, turning and stopping at traffic lights. However, in the case of a two-lane road, the road sweeping vehicle next to it occupies two lanes, and for safety reasons, the vehicle does not choose to overtake through the reverse lane, and can only follow at ultra-low speed. In the face of some special road scenes, there is still some room for improvement in automatic driving.
Robotaxi has entered the "master driver unmanned" stage
It is understood that the demonstration application subjects that have obtained the notice can carry out unmanned autonomous driving manned demonstration applications on open roads within 60 square kilometers of the Beijing High-level Autonomous Driving Demonstration Zone.
The Detailed Implementing Rules put forward higher requirements for the security management, network and data security, and risk response of test subjects. At the same time, according to the unified requirements of the policy, in the early stage of unmanned operation, in order to fully ensure the safety of unmanned vehicles, the co-driver of the vehicle will temporarily configure the braking device.
Beijing citizens have the opportunity to call on a self-driving vehicle with no one through Baidu's Radish Run or Xiaoma Zhixing's PonyPilot+ App, to yizhuang's multiple subway stations, public facilities such as parks and sports centers, as well as key business districts and residential areas. Robotaxi, which is currently uninhabited, is free to the public and will move towards paid operation in the future.
According to Wei Dong, 10 Baidu autonomous vehicles built based on Hongqi vehicles have obtained the first batch of unmanned license qualifications, while 30 self-driving cars based on Jihu have also entered the unmanned license qualification application process.
Since the main driver is unmanned, Wei Dong said that the government strictly defines the safety boundary, and all tests are designed for "unmanned" on the traditional test system with many more subdivided and high-standard parameters.
At the same time, the scheme of supervision and control is different. Unmanned licensing examinations not only include information security, data security and vehicle safety, but also assessments of managers, co-pilot safety officers, and cloud drivers.
In terms of technology, Baidu has built a three-layer security system. The first layer is the ability of bicycle intelligence, the second layer is monitoring and security redundancy; the third layer is cloud driving. In addition, together with the team's support mechanism, information security and other peripheral systems, a complete security system is formed.
The shortlisted Xiaoma Zhixing has opened the unmanned test of the main driver "de-safety officer" in Beijing, Guangzhou and other cities, and gradually expanded the scope, time and vehicle scale of the test. Among them, Xiaoma Zhixing has opened its experience to specially invited users in Guangzhou to experience Robotaxi, which has no one in the main driving position, and has completed nearly 1,000 travel orders, with the purpose of collecting ride feedback, so as to quickly feed back technology research and development and service upgrades.
It is worth noting that Pony Zhixing has been moving continuously recently. On April 24, Xiaoma Zhixing announced that it won the bid for the 2022 taxi capacity index in Nansha District, Guangzhou, and was allowed to invest 100 self-driving cars in Nansha, Guangzhou to provide taxi services, which is the first taxi operation license issued to an autonomous driving enterprise in China; on April 26, Xiaoma Zhixing, which participated in the A round of financing of Ruqi Travel, announced that it would build a 100-scale autonomous driving fleet with it, and opened a Robotaxi demonstration operation in Guangzhou during the year.
"With the accumulation of technology in the past few years, autonomous driving has begun to gradually enter commercialization. There are several important links in the commercialization process: one is policy support, the other is partners, and the third is the use of users. Recently, Xiaoma Zhixing's actions have revolved around how to commercialize - accumulate partners, obtain more qualifications and policy support, and provide users with a quality experience. Peng Jun, co-founder and CEO of Xiaoma Zhixing, told the 21st Century Business Herald reporter.
Compared with the previous Robotaxi must be equipped with safety personnel in the main driver's seat, this time the main driver is unmanned, the co-driver has a safety officer to take another step towards unmanned manned, and how to ensure the safety of "unmanned behind the steering wheel" automatic driving is crucial.
"Whether the driver is in the main vehicle or not, our version of autonomous driving is consistent, which means we can verify safety with the same standards." Zhang Ning told reporters, "The main responsibility of the co-driver safety officer is to serve as the final safety pass, and the second is to give passengers a sense of psychological security, gradually cultivate public awareness, and improve the acceptance of the whole society." ”
In addition, in addition to some safety takeovers during the driving process, the takeover of obstacle avoidance for situations such as traffic jams or detour obstacles is mainly completed by remote assistance systems. However, in very extreme cases, if the remote assistance system does not successfully get out of trouble, there are still co-pilot safety officers to get out of trouble.
Large-scale commercialization faces challenges
According to IHS Markit forecast, by 2030, the total market size of shared travel in China will reach 2.25 trillion yuan, of which Robotaxi will account for 60%, that is, 1.3 trillion, and the shared travel market will gradually shift to the unmanned car travel market in the future.
Wei Dong said that in the third quarter of last year, radish fast running completed 110,000 orders, and in the fourth quarter, more than 210,000 orders were completed. The average daily single volume and monthly total single volume of radish fast running continued to rise, entering a steady growth process.
In the face of this blue ocean market, achieving large-scale mass production and improving commercial profitability, and feeding technology with massive data, has become the goal of the second half of autonomous driving.
Wei Dong believes that at this stage, Baidu does not take profitability as a goal, but more importantly, the commercialization of large-scale automatic driving.
He believes that the commercialization of autonomous driving needs to meet multiple conditions.
"First, the geographical scope of coverage should be wide enough. As a physical displacement service, the wider the area of service, the larger the audience that can be served and the larger the market. The size of the market depends on the size of the road area that the government will open to unmanned vehicles in the future. Second, the cost structure should be relatively reasonable. If the cost structure is not reasonable, the larger the commercialization, the greater the financial pressure on the enterprise. Wei Dong said.
If you want to achieve a reasonable cost structure, the first dimension is hardware, Baidu will continue to integrate hardware capabilities through software technology, reduce hardware consumption, and continuously reduce hardware costs. Baidu's goal is to bring the hardware close enough to the current price range of private cars. The second dimension is driver cost. In scenarios such as taxis, ride-hailing cars, and buses, drivers account for a high proportion of costs. However, in terms of commercial value, only by unmanned can we achieve a subversive change in the cost structure of autonomous vehicles.
"Unmanned and large-scale are two important issues that need to be overcome in autonomous driving technology, and we will rely on these two points to promote Robotaxi to achieve true unmanned on a larger scale, in a more batch way or a real front-loading mass production method." In Peng Jun's view, with the improvement of the car end, the chip computing power has been greatly improved, as well as the decline in sensor cost and performance improvement, the inflection point of automatic driving development is coming.
Lou Tiancheng, co-founder and COO of Xiaoma Zhixing, said that Xiaoma Zhixing has helped improve mass production capabilities through in-depth cooperation with OEMs. Not long ago, Xiaoma Zhixing released the latest generation of automatic driving software and hardware platform - the sixth generation of automatic driving system, the hardware on the car uses sensors and hardware that can be mass-produced by the car, laying the foundation for future mass production. In addition, cooperation with OEMs can further optimize and reduce costs, so that commercialization can arrive earlier.
IHS Markit pointed out that driverless taxis from the demonstration operation to the opening of the charge, will be the real scene + real demand to verify the rationality of the driverless taxi business model, and further optimize the product design, with the unmanned vehicle operation section of the mature conditions, the cost of the hardware of the car terminal gradually reduced, the car safety officer from the main driver to the co-driver, and then further to remote control, can further save labor costs, then the driverless taxi is expected to usher in the breakeven opportunity.
The McKinsey report predicts that in 2025, the cost of Robotaxi will be equal to that of online ride-hailing and taxis, and the cost of Robotaxi will be reduced year by year after that, and gradually replace manually driven taxis.
"Last year, the cost of hardware vehicles has reached 480,000 yuan, according to the 5-year depreciation apportionment, the cost of a month is 8,000 yuan, and there is an opportunity to start doing business in the car market." Wei Dong judged, "In the next 2-3 years or so, we believe that there is an opportunity to achieve unmanned commercial profitability at a single point." ”
Despite the attractive prospects, autonomous driving still has a long way to go before it can be commercialized on a large scale.
In Lou Tiancheng's view, large-scale commercialization needs to experience three challenges, one is that the technical level of safety should be high enough, to achieve mass production and can be reduced to a low enough cost, while establishing overall operation, monitoring, scheduling, remote assistance and maintenance, etc.; the second is to continuously promote completely unmanned commercial applications, unmanned large-scale applications and other laws and regulations; third, how to let the public accept a truly unmanned Robotaxi from experience, price and other aspects, and use it as a means of transportation for daily travel.
"In addition to technology, regulations and users, the development of autonomous driving technology also needs the maturity of the entire ecosystem, because the autonomous driving industry is certainly not an OEMs, an autonomous driving company, a sensor manufacturer can do it, and there must be more capital and talent to promote the development of the industry." Lou Tiancheng finally said.
This issue is edited by Li Yutong Intern Huang Jingshan
