Behind The Expansion Of Robotaxi'S Operation: The Problem Of Commercialization Still Remains To Be Solved

The driverless vehicle on the air outlet is ushering in a new development opportunity.

On May 27, a meeting of the Standing Committee of the seventh Shenzhen Municipal People's Congress heard and deliberated the explanation on the "Regulations on the administration of intelligent Internet connected vehicles in Shenzhen Special Economic Zone (Draft)" (hereinafter referred to as the "draft"), which clearly stipulates that intelligent Internet connected vehicles can drive on the roads of the special zone after obtaining the registration certificate, license plate and driving license.

What is noteworthy is that the draft clearly points out that the highly automatic and fully automatic intelligent connected vehicles can not be equipped with drivers for road test or demonstration application after being examined and approved by the relevant municipal authorities; At the same time, the authorized municipal government can choose the district wide open road test, demonstration application and commercialization pilot with relatively perfect vehicle road collaborative infrastructure.

This may mean that the new field of driverless is about to have a law to follow. More importantly, the demonstration application from the test road to the open road with perfect infrastructure may also mean the faster realization of the commercial landing scenario of robotaxi.

"As one of the important landing applications of intelligent Internet connection technology in shared travel scenarios, automatic taxi has the potential to realize commercial promotion and landing at the first time, and has become the most likely breakthrough point for automatic driving to achieve ice breaking promotion. In recent years, it has been a hot area for the layout of major car enterprises and technology companies. Especially in the past 2020, more and more professionals and more cities are joining the robotaxi commercial development track, from traditional car companies, autonomous driving start-ups to travel companies and technology enterprises. "

On May 26, Ma Jian, chairman of the China Society of automotive engineering and director of the working committee of automobile intelligent shared travel, said at the eighth annual meeting of the international intelligent network Union (cicv 2021).

Since this year, many domestic car companies have announced plans for mass production of lidar. Visual China

In 2020, autox and didi successively launched driverless taxi service in Jiading, Shanghai. The pilot of auto driving travel service of Volkswagen Group officially landed in Hefei, and obtained the first automatic driving vehicle test license in Hefei; Since 2021, the commercialization of the field of automatic driving has been rapidly promoted, especially in the robotaxi scene close to C-end users, and start-ups are competing fiercely.

On January 28 this year, autox officially built China's first fully unmanned operation center in Pingshan District, carried out the demonstration application of L5 level fully unmanned robotaxi fleet, and explored the commercial operation mode.

According to the incomplete statistics of the 21st century economic report, so far, 27 cities in China, including Beijing, Shanghai, Guangzhou, Shenzhen, Suzhou, Wuhan, Changsha, Cangzhou, Hangzhou and Hefei, have issued automatic driving test policies, and the open test road mileage exceeds 3000 km.

It is worth noting that on April 19, Beijing Municipal Bureau of economy and information technology issued the "overall implementation plan for Beijing intelligent connected vehicle policy pilot area", which formed a 225 square kilometer planning range of Yizhuang new city, covering multiple scenes such as urban roads, expressways and expressways.

With the gradual expansion of the scope of automatic driving test, the rapid implementation of policies and the large-scale launch of robotaxi, which is close to C-end users, the development conditions of automatic driving are becoming more and more mature.

However, there is also a danger hidden in the broad prospect. As a long-term and heavy investment industry, when to realize the commercialization is still far away.

Multiple contradictions to be solved

At present, there are three operation modes of robotaxi: the first mode is independent operation, represented by waymo and Uber; The second is the crowdsourcing model proposed by Tesla, which allows owners to add idle vehicles to the taxi network; The third is the joint venture or cooperation mode between the automatic driving company and the main engine factory.

From the perspective of long-term market-oriented operation, commercialization and profitability are the most concerned issues of enterprises. From the perspective of the industry, the commercial landing of robotaxi is indeed facing many difficulties and problems, at least three contradictions.

The first contradiction is the contradiction between scale operation and cost.

"Robotaxi is a typical two-sided platform economy model. If there are not enough autonomous vehicles and the waiting time of users is too long, the whole platform is difficult to attract users. On the contrary, if there are not enough users, the cost of enterprise layout is very high. " On May 26, he Xia, chief engineer of the Institute of policy and economics, China Academy of information and communications, pointed out that robotaxi will become a benchmark for the commercialization of autopilot landing, which is a view that everyone agrees with. But the cost is a very important obstacle in the promotion of robotaxi.

Enough autonomous vehicles mean scale, and scale means high operating costs.

It is estimated that the refitting cost of a test vehicle is more than 1 million yuan. If the operation and maintenance cost and the labor cost of safety personnel are included, the number will only be higher; At the same time, the training of automatic driving technology and algorithm needs to rely on a large number of test vehicles to collect data, which makes the research and development cost more heavy.

Data show that waymo's operating loss has exceeded $20 billion from 2015 to 2019. As of April this year, waymo's driverless fleet was about 700, far from its target of 20000.

"At present, most of the robotaxi are concentrated in the first tier cities. Both the cost of manpower and site rental is very high. The cost of vehicle procurement, insurance and license application exceeds one million, so it is difficult to expand the scale of this fleet. Enterprises do not have the qualification of operation service, and there is no way to carry out trial operation to achieve revenue. " JieFei, deputy director of the science and technology innovation Bureau of Beijing Economic and Technological Development Zone, pointed out.

However, the operating costs of enterprises are very high, the market environment lacks the support of professional insurance, and the test premium is relatively high. The willingness of payers to pay still needs a long time of market cultivation, and they do not have the conditions for profit.

Due to the lack of real application scenarios, high cost of commercialization process and limited scale of industrialization, automatic driving has only achieved a technical breakthrough from 0 to 1, and the commercialization landing is still in the mode of "one car, one road and one platform".

"There are also two contradictions. One is the contradiction between technology development and product access. If products are not admitted, how can technology be verified on a large scale? The second contradiction is the management contradiction between technology maturity and commercial operation. If there is no large-scale demonstration application and technology is not tested and evaluated in a large scale, commercial landing may also have certain risks. " Bi Xin, executive director of Shenzhen future intelligent network transportation system industry innovation center, pointed out.

The existence of these two contradictions also directly affects the commercialization of robotaxi.

It is understood that at present, the local governments have different policies and route standards for robotaxi. In addition, the open area and time for testing are limited, and the commercial operation has not been fully opened up. There are many long tail challenges that have not been solved. It is necessary to conduct long-term verification in the real scene.

"The challenges we are facing are also common problems faced by the industry, such as the construction of technical system, the development and promotion of standard protocols, which require the industry to jointly develop iteration. At the same time, it also faces the problems of large investment amount of roadside equipment, unclear business model, and the number of vehicles entering the network needs to be further improved Jaffe points out.

In addition, there are some problems in the whole process of robotaxi testing.

On the one hand, automatic driving scenarios are infinitely rich, but also face some very uncertain difficulties. How to effectively model the real scene in the automatic driving scene, and how to ensure the functional safety, network security and OTA security of the automatic driving system are very important issues.

On the other hand, the diversity of automatic driving scene and the irregularity of road signs are also very important problems in the evaluation process.

Policy first

"There are three clear directions for the commercialization of automatic driving: the first is robotaxi, the second is long-distance freight trucks, and the third is unmanned distribution vehicles. The last two are now being promoted, including policies, including rules and standards." He Xia pointed out that if robotaxi wants to truly commercialize, it needs a strong ecosystem composed of platforms, autonomous driving system integrators and vehicle manufacturers, and through this ecology to carry out relevant large-scale promotion.

However, according to He Xia, the most important thing in the whole business operation of robotaxi is the implementation of the government framework. At present, the development of the whole industry needs strong support from policies, regulations, standards, finance, insurance, etc.

"Smart cars and the ecosystem construction of self driving travel services are inseparable from urban infrastructure. As far as the road test area of the pilot project is concerned, within 16 square kilometers, from policies and regulations to lane markings, without the support of the government, it is impossible to complete the task of talking about smart travel by enterprises alone. Without the participation of large-scale enterprises, a single enterprise can not form a development climate. " Previously, Su Weiming, former director of Volkswagen Group (China) and executive vice president of Volkswagen Group (China), said in an interview.

Of course, all regions pay more and more attention to the development of automatic driving technology.

At present, Beijing has built the world's first high-level automatic driving demonstration zone with cloud controlled network connection. In April this year, the first domestic policy leading area for intelligent Internet connected vehicles was set up. A total of 112 road test plates were issued for Xiaoma Zhixing, Baidu and didi. On May 25, Jingdong, meituan and Neolithic enterprises issued the road code of unmanned distribution vehicles, which realized the first employment of unmanned distribution vehicles with certificates.

After Guangzhou issued the first batch of 24 Road Test licenses in 2019, it is expected to reach 200 by the end of this year.

In Shenzhen, in addition to taking the lead in realizing the unmanned robotaxi demonstration in China, on March 23, the Shenzhen Municipal People's Congress promulgated the "Regulations on the administration of intelligent Internet connected vehicles in Shenzhen Special Economic Zone" (Draft for Soliciting Opinions), which not only set up the exclusive official number plate of Shenzhen special zone for intelligent connected vehicles, but also explicitly granted the legal driving status of automatic driving system.

This move also means that the intelligent connected vehicles will really move from the test products of scientific research to the sales products on a large scale, which will enable the relevant enterprises to maintain their operation from only relying on capital investment to the normal commodity economic cycle through commercial profits.

"There are two problems to be solved in the implementation of business model, one is how to charge and the other is to truly realize nobody. The key to these two problems is policy and technology."

According to Li Lin, deputy chief engineer of Shanghai International Automobile City, only when the business model is implemented can we talk about profitability. With the gradual maturity of various sensors, lidar and on-board computing platform technology and the expansion of scale, the cost of single vehicle will gradually decrease. In the future, with the spread of large-scale, commercial landing will be realized.

"We hope that we can have a better business model in terms of business model. At present, what we have put forward is smart car and road, one-time investment in infrastructure, and sustainable operation ecology, including smart cars, smart roads and cloud network maps, to form an integrated unified planning." He Peng, a senior researcher at Baidu intelligent driving business group, said.

Momenta's goal is to make a profit by 2024.

"The first is the cost of research and development, to solve technical problems; The second part is the cost of data, to accumulate test mileage. " Ma Chen, senior director of government and enterprise affairs of momenta, said that the marginal cost of research and development was reduced by data-driven algorithm, and the cost of data was reduced by crowdsourcing.

"If bicycle profits are realized in 2024, the commercialization of robotaxi and autonomous driving will enter a large-scale outbreak period." Ma Chen finally said.

?