China has 14 million freight trucks and 30 million truck drivers, who contribute three-quarters of the freight traffic of China's transportation industry.
Putting these trucks into the Internet, replacing drivers with algorithms, and achieving driverless driving opens up a new market worth trillions. This market is called robotruck.
Self-driving trucks, an industry that has taken a lot of financing and endured a lot of controversy, is a way for autonomous driving technology to land. The counterpart is the Self-Driving Taxi (Robotaxi), except that Robotaxi replaces taxi drivers and Robotruck replaces truck drivers. One is to carry people, the other is to pull goods, essentially to make the vehicle autonomous.
Further subdivided, self-driving trucks are divided into trunk logistics and closed scenarios (including ports, mines, airports, etc.), of which the most promising and the highest market share is trunk logistics. This article mainly explores the trunk logistics scenario of self-driving trucks.
The autonomous driving application scenario of trunk logistics freight trucks is relatively simple, there are no pedestrians and non-motorized vehicles on the highway, there are no traffic lights, and the road conditions are simple, so it is considered to land faster than passenger cars.
The huge development prospects have attracted a large number of entrepreneurs to enter the market, forming a complete ecology. But the challenges facing the industry are also unprecedented, and no player can take it lightly until large-scale commercialization arrives.
True and false trillion market
When it comes to self-driving trucks, entrepreneurs and investors in the industry will excitedly give a judgment: this is a trillion-dollar market.
The blueprint they draw to the outside world is that China's long-distance truck freight market is more than 3 trillion yuan, and the market space and business prospects are extremely attractive.
Feng Yanjiao, partner of CIC Insight Consulting, analyzed the deep way, the market space of trunk logistics is particularly large, and it is an operating market, with about 4 trillion yuan in revenue every year, plus 2 trillion yuan of revenue from the same city freight, and the total scale of the entire freight market is about 6 trillion yuan.
In the past, the transportation of these goods was mainly done by truck drivers, but in the future, new trucks with autonomous driving capabilities can completely or partially replace drivers.
A seemingly simple alternative can create huge economic benefits.
CICC believes that the economy of self-driving trucks is significantly better than that of traditional fuel heavy trucks, which is mainly reflected in three aspects: one is to increase the running time of trucks, the other is to reduce driver costs, and the third is to improve fuel economy. According to its estimates, self-driving trucks can eventually reduce costs by 45%.
Source / CICC
He Yuhua, managing partner of Hegao Capital, told Shentu: "The biggest driving force behind the unmanned commercial vehicle is the economic value behind it. In essence, commercial vehicles are a production tool, is an economical thing, what changes is not comfort, but how to help enterprises make more money, so the industry has a great incentive to promote the automatic driving of commercial vehicles, and many companies are willing to try, so that it lands faster than passenger cars. ”
Seeing the huge business prospects, many talented entrepreneurs and star investors flocked to the game, representative enterprises are Tucson Future, Zhijia Technology, Yingche Technology, Xiaoma Zhika, Qianhang Technology, Kinte Smart Card and so on.
Tucson future listed in the U.S. stock market in April last year, becoming the "first stock of autonomous driving". It is aimed at L4-level autonomous driving, the main market in the United States. Back in 2018, Tucson was conducting small-scale commercial operations of self-driving trucks on Highway 10 in Arizona, USA. In 2019, it took over UPS's self-driving truck operations service between Phoenix and Tucson.
Zhijia Technology has a deep layout in the Chinese market. It obtained the China autonomous truck test license in 2018, established an autonomous driving joint venture with FAW Group in 2019, and obtained an inter-provincial test license in the Yangtze River Delta region, and began to cooperate with FAW Jiefang to build cars in 2020, and the self-driving truck "J7 L3" began mass production last year.
Another strength player is Wincher Technology. It obtained the Changsha Autonomous Driving Logistics Heavy Truck Test License in 2019, and then began to build vehicles jointly with Dongfeng Commercial Vehicles, and released the truck automatic driving system "Xuanyuan" last year.
The above-mentioned companies have strong research and development capabilities, are good at technology, understand products, and are an important force to promote the development of the self-driving truck industry.
There are also companies that come from other tracks to switch lanes or cover multiple tracks at the same time.
For example, the autonomous driving company that does the closed scene also includes trunk logistics into the business scope. Taking mainline technology as an example, this company started from the port, the advantage has been in the port, and later also did trunk logistics.
Technology companies that make unmanned taxis have also begun to cannibalize the self-driving truck market. Pony Chi Heng announced its entry into the truck field in 2019 and obtained a self-driving truck test license in 2020. Wenyuan Zhixing acquired self-driving freight company Muyue Technology last year to lay out self-driving trucks. Yuanrong Qixing also said last year that it would lay out autonomous driving of freight in the same city.
A general-purpose autonomous driving technology company that will cover both commercial vehicles and passenger cars. The application scenarios of the automatic driving system developed by Hongjing Intelligent Driving also cover L3 heavy trucks, passenger car high-end ADAS, buses, unmanned mining cards, Robotaxi and other fields. Intelligent Driving Technology took commercial vehicle ADAS as a key development direction in the early days, and later developed the passenger car ADAS business, and now it has increased the number of commercial vehicles.
Yang Tengfei, deputy general manager and operation director of MAXIEYE of Intelligent Driving Technology, told Shentu: "From the perspective of the lowest core technology chain, algorithm model, and most data scenarios, commercial vehicles and passenger cars are universal, with an overall overlap of more than 70%, but there will be some differences in the combination of the underlying technology and engineering landing. ”
This is the premise that some algorithm companies can lay out passenger cars at the same time.
However, there is no self-driving truck company in China that has achieved truly unmanned commercial operation (excluding closed scenarios), and only a few are testing with modified trucks. Because the premise of large-scale commercialization is that self-driving trucks can achieve L4 level automatic driving. At present, L4 in the true sense of the word has not yet been implemented.
Technology is only one aspect. "The biggest problem with trunk logistics is that it involves legal moral and ethical issues." He Yuhua said. According to the current policies and regulations, it is not allowed to allow real driverless trucks to be commercialized on a large scale in trunk logistics.
Technology and policy are two mountains, which have become the biggest obstacles in the commercialization of self-driving trucks.
This means that this heart-warming trillion blueprint is only a fantasy in the short term.
Want cash or a dream?
Most companies see 2025 as a turning point, when breakthroughs in technology and policy will be made. But that means these companies may not make any money until 2025.
This discouraged many investors. "A company can't always tell a story, and it has been telling it for 10 years without revenue." He Yuhua said.
There are two ways in front of the self-driving truck company, one is to continue to adhere to the L4 route, rely on burning money to maintain, and then in 2025 one step in place; the other is to choose a gradual route, first let some of the technology land, in practice progress iteration.
The representative company that chose the first route is Tucson Future. According to Tucson's 2021 annual report, it currently operates about 100 L4 self-driving trucks, with revenue of $6.261 million in 2021, and there are currently about 6975 orders for self-driving trucks.
A special point is that Tucson's future main position is in the United States, where 75 of these 100 trucks are in the United States, and the United States is relatively open to regulation of autonomous driving. "In the Chinese market, Tucson's commercialization process is lagging behind, for example, it used to lease port terminals for automatic driving tests, but there is no real sense of integration into port operation scenarios and systematic docking port operating systems, which will lead to commercial operations can not really land." He Yuhua said to Shentu.
In the past two years, many domestic self-driving truck companies have chosen the second route, reducing the technical level of autonomous driving from L4 to L3 or L2.5, doing advanced driver assistance systems for trunk logistics, and gradually realizing L4 level automatic driving.
Typical players are Wincher Technology and Zhijia Technology.
Both companies have chosen to mass-produce SL4/L3 self-driving trucks first, which can be simply understood as L4 trucks with safety officers, which belongs to the transition state between L3 and L4. They do not expect to achieve unmanned people above L4 level, but "manned" below L4 level.
Source / pexels
Cleverly, even this can replace the current operation mode of two drivers taking turns driving the truck, which has the effect of saving costs.
Ma Zheren, founder and CEO of Yingche Technology, has publicly pointed out that the biggest feature of trunk logistics is "people do not rest", one-way usually reaches 800 kilometers or more, need to be equipped with two truck drivers, with an average monthly salary of 15,000 yuan per driver. Self-driving technology can generate economic benefits if it can replace one of the drivers and make the truck more fuel-efficient during the journey.
Fuel saving is a major point for many self-driving truck companies to prove that self-driving trucks are better than traditional trucks.
Rong Li, general manager of Zhijia Technology China, said that the operation of the Shandong-Yangtze River Delta route they cooperated with Rongqing can save 10% of fuel and more oil under the mountain road. Because autonomous driving has all the data on road judgment, it is more accurate than human operation and can better control the throttle.
According to the calculation of Zhijia Technology, with the increased cost of the current system of assisted driving, if it is an efficient operation of the fleet to operate it, one year according to the number of kilometers of normal trunk logistics to run, about two years can recover the cost.
Therefore, even if the current self-driving trucks cannot be unmanned, they already have the ability to commercialize to a certain extent, especially those companies that choose the progressive route, or will be the first to find a change scene.
And once the boundaries between L3 and L4 are crossed, the profit path for self-driving trucks is very clear.
Feng Yanjiao told Shentu that Robotruck has three ways to make a profit, one is to set up a fleet as a logistics platform and provide capacity, collect freight or service fees, Tucson is currently taking this way in the future; the second is to transform trucks for automatic driving, install software and hardware systems, and then sell cars; the third is to provide autopilot solutions including perception + algorithms for automakers as a supplier of technology and system solutions.
In these three modes, logistics contractors, operators, and OEMs pay for them. At every key link in the trunk logistics industry chain, there are opportunities to monetize autonomous driving technology.
A criss-crossing web
Behind the variety of ways to make money is the extreme complexity of the ecology of this industry. The companies mentioned above are just the tip of the iceberg of the entire self-driving truck industry.
By nature, there are about four categories of heavy players in the jianghu, one is a self-driving truck company, such as Tucson Future; the other is a truck company, which is also known as an OEM; the third is a logistics company or logistics platform, which is a carrier or shipper; and the fourth is a supply chain partner, involving chips, batteries and so on.
Due to the large number of stakeholders involved, no company can fully penetrate the entire industrial chain, and self-driving truck companies must form a certain degree of bundling with the upstream and downstream of the industrial chain.
The conclusion of cooperation is the most common means. The OEMs that Tucson will cooperate with in the future are Navistar, Traton, Foton Motor, NVIDIA to Tucson Future supply autonomous driving chips; Zhijia Technology binds FAW Jiefang and SAIC Group, and at the same time reaches in-depth cooperation with manbang, SF, amazon and other logistics platforms; the OEM manufacturers bound by Wincher Technology are Dongfeng Commercial Vehicle, China National Heavy Duty Truck, and logistics platforms such as Debon and JD.com.
However, cooperation at the pure business level is highly uncertain and often difficult to exclude.
In order to form a closer binding, they will choose to use equity investment.
Tucson Future began using driverless trucks to transport goods in the United States in 2019, when the US express delivery company it cooperated with was UPS, and soon after Tucson Future announced that it had received an investment from UPS. Tucson Future announced in January this year that it will develop driverless domain controllers based on NVIDIA's latest vehicle-grade AI chips, after Nvidia participated in Tucson's future Series B financing.
Zhijia Technology and FAW Jiefang established a joint venture company, the depth and breadth of cooperation far beyond the scope of general business, it also received investment from FAW Group. In addition, it also received investment from Manbang Group, China's largest digital freight platform, and in 2019, the two sides reached an exclusive strategic cooperation, and Zhijia became manbang's exclusive L4 solution partner in the field of autonomous driving at that time.
The identity of Wincher Technology is more special, it is a combination of three companies. The three companies are G7, an IoT technology company that provides fleet operation solutions, GLP, a logistics real estate giant, and NIO Capital. In terms of logistics resources, in addition to GLP, there are Debon, JD.com, and Meituan, all of which are shareholders of Yingche Technology.
Combing through these cooperations, it can be found that self-driving truck companies are mainly bound to commercial vehicle OEMs and logistics platforms.
Feng Yanjiao said that the autonomous driving of trunk logistics should be landed, and it is necessary to operate the fleet at the beginning, so it is impossible to avoid cooperation with logistics companies. "Logistics companies generally pick a relatively simple point-to-point route and let the fleet of self-driving trucks run."
Therefore, the industry has formed a "network freight platform + automatic driving" architecture, such as "Manbang Group + Zhijia Technology", "G7 Internet of Things + Winche Technology". Under this architecture, the network freight platform provides supply and manpower, and the autonomous driving enterprise provides cost reduction and efficiency-enhancing technology, and the two are in agreement.
At the end of the vehicle, self-driving trucks can't avoid the main engine factory that controls the car-making resources.
He Yuhua believes that the autonomous driving companies that do trunk logistics want to do their own cars, because only their own cars, product definition can do the best, but it is not a full-stack self-developed route, some automatic driving companies will consider themselves to do the core domain control, in the chassis we still have to find an OEM.
So these self-driving truck companies are crazy together. They are weaving a vast network of industries, expanding their boundaries through cooperation and financing, forming complex collaborative relationships. They are both operators and autonomous driving service providers, and there will be more roles.
Yang Tengfei believes that the future industry pattern is bound to be ecologically coordinated. "Robotruck will be different from the traditional truck in the past, from OEM to user (logistics company or individual driver) one-way sales, technological breakthrough innovation and subversive value of landing, will drive the needs of all parties to converge and collide, forming a pattern of collaborative co-creation."
The industry has not yet entered the second half
In the years of self-driving truck development, the entire industry is still in the first half of the competition, and the second half has not yet begun. Now it is still in the stage of fighting technology, and there is still a long way to go from the stage of fighting the market and fighting for scale.
Even in terms of technology, there are still many bottlenecks that the industry has to break through.
There is a consensus in the industry that electric vehicles are easier to achieve autonomous driving than fuel vehicles. However, most of the current self-driving trucks are still based on fuel vehicles, and only a few vehicles are based on pure electric architecture.
An insider of an autonomous truck company told Shentu that this is because there was no pure electric wire control chassis in China in the past few years, and even the wire control chassis of fuel commercial vehicles is rare, so this generation of products has been mass-produced or demo commercial vehicles Automatic driving are mainly fuel vehicles.
"Today and tomorrow, there must be rice in the pot first." He concluded.
This limits the speed of development of self-driving trucks to some extent. He Yuhua analysis: "Robotruck is very core is the ability of wire control, to achieve high-quality wire control, first of all, on the basis of electronic control, there is no cooperation between electronic control and wire control, to do automatic driving in many aspects will not be in place." ”
Now, more and more self-driving truck companies have put the development of pure electric trucks on the agenda.
In addition to the vehicle architecture, the industry also has a lot of room for improvement in the accuracy of autonomous driving technology.
The outside world usually believes that carrying objects is safer than carrying people, the scene is relatively simple, and the self-driving truck technology will land earlier. But in fact, in some aspects, the perception and prediction ability of the truck self-driving system is very demanding.
Trucks are much larger in size and weight than cars, and a large truck with a full load weighs more than 40 tons, which causes the truck to require longer reaction times and braking distances. Passenger cars only need 40 meters to brake, and trucks need to be 250 meters, which requires lidar to have a longer range of detection capabilities. The driving scenario of the truck is generally a highway, and the speed is very high, which also puts forward higher requirements for the control ability of the algorithm.
Source / Visual China
If these problems are not solved, it will be difficult to land on a large scale.
But on the other hand, from the perspective of industry competition, this also means that the overall situation of the industry is undecided, there is still a possibility of reshuffling, and new players still have the opportunity to emerge.
In the second half of last year, the self-driving truck track suddenly established three new companies - Qianhang Technology, Xingjiao Technology, Kintex Smart Card, of which Qianhang Technology and Kinte Smart Card got financing in a very short period of time, and SF, Baidu, and Xiaopeng Automobile are all investors in Qianhang Technology. In March this year, it was reported that Tao Ji, former head of Baidu's autonomous driving, intelligent transportation technology and products, was appointed CEO of Qianhang Technology.
In the view of Pan Zhenhao, CEO of Kinte Smart Card, the current self-driving truck industry is still in the initial stage of development, and the industry pattern and business model have not really formed a fixed trend. In other words, this is also an opportunity for startups such as DynaSky Card.
He Yuhua believes that from a technical point of view, the difference between the head self-driving truck companies will not be too big, and the real gap lies in customers, mileage, and commercial revenue. The biggest disadvantage of startups such as Qianhang Technology is that it debuts too late, and it is relatively backward in terms of test mileage, data and the number of landing customers, and once the policy of trunk logistics is released, if it cannot grab the first wave of high-quality head customers, it will be more difficult in the later stage.
"The reason why they dare to come out and start a business is because the fully autonomous driving of domestic trunk logistics is still a long time cycle from the commercial landing, and they believe that before that, their technology can be iterated to the same level as competitors." He Yuhua said to Shentu.
The turning point is around 2025, when the industry enters the second half.
Until then, self-driving companies need to continue to raise funds, but fortunately, some investors are willing to continue to pay for their dreams. Last year, Both Yinche Technology and Zhijia Technology received financing, and in February this year, Yingche Technology announced that it had obtained a new round of financing, and at this rate, Zhijia Technology should also carry out at least one round of financing this year.
There is still a lot more to be done in making trucks driverless. Letting dreams shine into reality is never easy.
Reproduced from Zhiche Technology, the views in the text are only for sharing and exchange, and do not represent the position of this number, such as copyright and other issues, please inform, we will deal with it in a timely manner.