Beach-rushing autopilot

Technology, capital, regulations and scenes, where will the autonomous driving industry go with many uncertainties?

Wen Wei Song Doudou?

10 year 10 On October 22nd, Tesla CEO elon musk said on his Twitter that in view of Tesla's fully automatic driving test version (FSD? Beta) has been sold in limited quantities, and its price (new version or upgraded version) will increase by about 2000 dollars next Monday. 10 year1On the evening of October 20th, Tesla introduced the fully automatic driving test software to some users. In addition, Musk said in a conference call after the release of the earnings report that it is expected that the autopilot function will be fully launched by the end of this year.

This behavior of Tesla once again attracted people's attention to autonomous driving.

At the key node of the automobile industry's transition to the era of stock game, the changes in the field of travel are ready to go, and OEMs, travel platforms and technology companies have rushed to the beach track. As an industry in the early stage of development, autonomous driving is still exploring from technical route to landing scene, from cooperation mode to industry ecology. There are many entrants, who can lead? How far is it from commercial application? Technology, capital, regulations, scenarios, where will the autonomous driving industry go under many factors?

From ideal to reality

With the rise of a new round of scientific and technological revolution, autonomous driving, as an advanced form of intelligent and networked development, has become a competitive position in the automobile industry. In recent years, this gorgeous track has poured into countless top technologies and talents, and also attracted countless capitals and customers.

The emergence of automotive intelligent networking will provide unlimited possibilities for future travel. With the gradual application of primary assisted driving function, autonomous driving is no longer out of reach and is becoming a real experience at your fingertips. According to the iResearch report, by 2022, the global autopilot penetration rate from L/KLOC-0 to L5 will reach more than 50%, and in 2030, the global autopilot penetration rate will increase to 70%.

It is worth noting that because the technology is not yet fully mature, there are many restrictions in the fields of policy, ethics and scenarios. Whether it is a traditional car company, an autonomous driving company or a technology giant, the business related to autonomous driving can't be profitable, and it depends on capital transfusion to survive.

Before the large-scale commercial application of autonomous driving, Robotaxi has become one of the most likely first commercial scenarios, exploring commercial models and collecting a large amount of data, thus promoting the improvement of technology. At present, the battle for the head of Robotaxi industry is being staged around the world. Waymo, Zoox, Cruise and Tesla are scrambling abroad, while Apollo, AutoX, Ma Xiao Zhixing and Wen Yuan Zhixing make efforts in turn.

UBS Group AG Group AG predicts that by 2030, the global self-driving taxi market may exceed $2 trillion. Among them, China market is an important force. According to McKinsey's research report, China will be the largest autonomous driving market in the world. By 2030, the total sales of self-driving cars will reach 230 billion US dollars, and the travel service orders based on self-driving will reach 260 billion US dollars.

Robotaxi's huge and broad commercialization prospects have attracted many enterprises to enter the market one after another, and the domestic pilot of Robotaxi has blossomed in an all-round way. In the first half of the year, AutoX and Didi successively launched driverless taxi services in Jiading, Shanghai; Following the pilot opening experience in Changsha and Chenzhou, Baidu Robotaxi service was fully opened in Beijing this month. Since its opening a week ago, its single-day order volume has peaked at more than 2,600 orders, and that of a single site has peaked at 600 orders.

Previously, start-ups such as Ma Xiao Zhixing and Wenyuan Zhixing had already been laid out. Since the end of 20 18, Ma Xiao Zhixing has successively launched Robotaxi service in Guangzhou and California, and this year began to lay out the truck market. Wenyuan Zhixing provides the first fully open Robotaxi service in China in Guangzhou, covering the open road in Huangpu District and the core area of Guangzhou Development Zone, with about 200 boarding and alighting points, which has been in operation for more than 300 days.

1010.21day, Zhikujun went to a specific site in Huangpu District to experience Wen Yuan Zhixing Robotaxi travel service. About 8 minutes after placing an order through its APP, an overhead laser Leila and a new species shuttling on the road slowly stopped. After the vehicle arrives at a specific boarding point, Wen Yuan Star will send a short message to remind you. If the passenger still doesn't get on the bus after waiting for 5 minutes, the order will be cancelled automatically. It is worth noting that the security officer will not call to remind passengers to get on the bus or confirm the location.

"This is to create an autonomous or even unmanned experience and try to weaken the existence of security personnel." A staff member of Wen Yuan Zhixing told Zhikujun.

Take a seat in the back of the car. After the safety officer carries out the routine temperature test and fastens the seat belt, click "Start" on the front screen, and the vehicle will start slowly and enter the automatic driving state. Real-time traffic interface will be displayed on the screen: the current speed, lane, surrounding vehicles, zebra crossings, pedestrians and bicycles parked on the side of the road can be seen at a glance.

In the nearly 20-minute drive, the security officer kept an eye on the road ahead, put his palm up on his leg, did not take over, and did not communicate with the passengers. During driving, a group of cyclists rode across the road and the car stopped to make way for them. When driving near the rightmost lane and parking social vehicles on the roadside, the vehicle will briefly identify the vehicles parked in front and then turn to the middle lane; In addition, with the help of high-precision maps, vehicles know the speed limit value of each road, constantly change the driving speed, and slow down several times when the vehicles behind them overtake.

In June of this year, Zhixing cooperated with taxis, and think tank Jun experienced the use of taxis for the second time, without submitting an application or qualification review, which was roughly the same as the ordinary network car rental process. However, it is worth noting that Gaode taxi will check multiple models at the same time. If you want to experience Robotaxi, you have to cancel other inspections. The reporter compared the time of the two orders, and the difference was not big.

"We adopt a mixed mode of operation+testing. Vehicles collect data on the road to provide travel services for users. This is convenient for us to promote operations through research and development, and then feed back technology through operations. This is a positive cycle. " Wen Yuan knows what to do. Zhang Li, chief operating officer, said in an exclusive interview with think tank Jun.

"After the epidemic, orders are also growing slowly. During the National Day, the average daily order volume of bicycles reached 18, and the average service mileage of each order increased from 4.9 km to 6.9 km. This achievement is very close to the operational efficiency of an ordinary taxi, which is a very good verification for our self-driving taxi company. "

From excitement to calm

In the field of autonomous driving, Waymo, the leader of American autonomous driving, has become the goal and direction of the players who rushed into the autonomous driving track. From the perspective of valuation alone, there are currently only two companies in autonomous driving-Waymo and other companies. In 20 19, Morgan Stanley's valuation of Waymo was lowered from the previous 175 billion US dollars to 105 billion US dollars, because the commercialization of driverless cars took longer than expected. Despite the downward adjustment, the myth of $100 billion is still difficult to replicate, and the technological accumulation of domestic leading enterprises and the financial strength of parent company Google are also difficult to match.

65438+1On October 9, Waymo announced that it would provide a completely unmanned taxi service in Phoenix. This is the first time that an autonomous driving company has opened a driverless taxi without a security officer to the public, which means that Robotaxi has taken an important step in technology and commercialization.

Driven by capital, China's autonomous driving industry has also experienced a period of barbaric growth. China electric vehicle committee of 100 statistics show that from 20 15 to 20 17, 1 1, there were *** 193 investment and financing events related to self-driving vehicles, amounting to143.8 billion US dollars.

Fortunately, in recent years, China's autonomous driving industry is becoming more and more rational in terms of technology and capital. The data shows that in 20 18, the total financing of auto-driving parts and solutions suppliers in China reached162.3 billion yuan and 20107/kloc-0.0 billion yuan, down 34% year-on-year.

Essence Securities research report pointed out that self-driving cars will be the definer of the rules of the game in the future automobile industry. Players on the autopilot track have accelerated the pace of autopilot landing. Compared with traditional OEMs, autonomous driving companies have advantages in algorithms, data and other technologies, but car companies are better in hardware such as vehicles, sensors and chassis. Technology companies need to make up for the shortcomings in hardware, while car companies are aiming at the huge travel service space derived from autonomous driving and trying to complete strategic transformation.

Different from foreign autonomous driving companies, such as GM and Cruise, Ford and ArgoAI, Amazon and Zoox, domestic OEMs, travel service platforms and technology companies cooperate strategically to form an "iron triangle" model.

"The most difficult aspects of autonomous driving technology are three aspects: one is perception, the other is decision-making and planning, and the third is hardware related to the body itself. Among them, the main engine factory is good at automobile-related things, but the main engine factory lacks the relevant talents and mechanisms for the perception, decision-making and planning of autonomous driving, while the core competitiveness of technology companies is technology and talents, and travel companies can provide real passenger landing and operation. Platform. "

Zhang Li told think tank Jun that the combination of autonomous driving companies and original equipment manufacturers formed the supply side, while orders from travel companies and taxi companies formed the demand side. Multi-party cooperation can connect orders and user requirements with multiple data types. The superposition and accumulation of such data will produce more valuable things.

It should not be overlooked that, despite the attractive prospects, Robotaxi faces many challenges from trial operation to large-scale commercialization: there are unknown variables in technology, funds, regulations and scenarios. Can it really be unmanned and operated on a large scale? Can you overcome the high cost and profit? The eve of commercialization of autonomous driving is still long.

At present, most Robotaxi in the testing stage or in operation are equipped with security officers to take over vehicles in case of emergency to ensure safety, but this runs counter to the vision of unmanned driving to free hands. Tension shows that in the taxi cost composition, the driver cost accounts for about 60%. If the driver cannot be changed, the business model of unmanned Robotaxi will not be established.

For all companies with Robotaxi as their main business goal or business model, the sooner they replace drivers and realize complete driverless driving, the key point is to decide their fate. "If unmanned driving wants to make a profit, it must achieve bicycle profit. The premise of achieving bicycle profitability is to be able to achieve driverless driving. At present, it is impossible to make money in the mode that vehicles are equipped with security officers, and even lose money than the ordinary mode. "

In addition, at present, the automatic driving level of smart cars generally stays at L3/L4 level, and technological breakthrough has become a restrictive factor for its further development. Driverless driving is more than simple code. In the real operation process of Robotaxi, there are more complicated factors than the simulated experimental environment: industry standards, policies and regulations, and responsibility identification after accidents.

According to Zheng Yun, senior global partner of roland berger and vice president of Greater China, the autopilot scenario needs at least four aspects of support, namely, the blessing of policies and international environment, the maturity of technology, the improvement of infrastructure, and the construction of new business models (including vehicle ownership, travel scenarios, vehicle architecture and systems, and user data). ).

"Autonomous driving is the most subversive trend in the field of travel, and it will be divided into different application scenarios in the process of implementation. Although the concept of unmanned driving has landed, Robotaxi is still equipped with security officers, but it has only reached L3/L4 level. Autopilot. " Zheng Yun previously said in an interview with think tank Jun, "It is a false proposition to talk about profits at this stage, because the cost control, regulations and the profit distribution model of the industrial chain are unpredictable at this stage."

This article comes from car home, the author of the car manufacturer, and does not represent car home's position.