The trap of autonomous trucks lies deep at the bottom of the IPO river, and TuSimple's future lesson Pony.ai needs to be vigilant.
@科技新知 Original
Author丨Fushen Editor丨Saike
From the first year of 2016 to the cold winter of 2022, the autonomous driving industry has spent six years and completed a bull-to-bear cycle.
According to statistics, from 2015 to the end of 2017, there were 193 investment and financing cases in the field of autonomous driving in China, with an amount of up to 143.8 billion US dollars. Also in 2016, Peng Jun, the chief architect of Baidu's autonomous driving, and Lou Tiancheng, who graduated from the first Yao class of Tsinghua University and is known as "the first person in China's programming", hit it off and founded Pony.ai, pointing to the endgame L4 of autonomous driving, and becoming the darling of capital for a while.
In the cold winter of 2022, the number of financing and financing amount in the domestic autonomous driving field fell by 32% and 61% year-on-year. Pony.ai has also inevitably experienced a nerve-wracking downsizing, with a personnel optimization ratio of 50%, and was once rumored to be a classic lesson of "genius boys lose to reality".
Fortunately, unmanned driving is not without life, and the first rays of hope began to pierce the dark clouds last year. The flare gun was first fired by LiDAR companies, Hesai Technology and Suteng Juchuang went public one after another, and WeRide Zhixing, Black Sesame, Horizon and other solution providers and chip manufacturers also sounded the horn.
Pony.ai seems to be a little late in this wave of IPOs. It was not until April this year that the China Securities Regulatory Commission issued a filing notice on Pony.ai's overseas issuance and listing, which stated that "it intends to issue no more than 98149500 ordinary shares and list them on the United States NASDAQ stock exchange or the New York Stock Exchange".
Pony.ai, which has reined in its youthful edge, once regarded survival as its first goal after a heavy fall, and not only lost the L4 fantasy, but also put the autonomous truck business with a clearer commercial scenario at the center, which is difficult not to say that it is a reference to TuSimple Future, which successfully listed in the United States with this business.
But now the latter, which has long been mired in performance losses and compliance disputes, has officially announced its voluntary delisting in January this year. Pony.ai, who copied the wrong answer, how should he stage a good show of changing his life against the sky?
Part.1
From dreams to reality
The real protagonist in the story of autonomous driving was originally a new energy vehicle. Because the power system of electric vehicles is simpler than that of traditional fuel vehicles, the electronic system of autonomous driving no longer needs to go through a gas-to-electricity loss, which is more conducive to controlling the implementation of vehicle behavior.
However, with the rise of new energy vehicles, the market penetration rate has been more than half, and autonomous driving has experienced the grinding from heaven to hell.
The bubble inflation began with the superstition of L4 in the capital market. Unmanned driving in the real sense replaces the jobs of hundreds of millions of passenger car and freight truck drivers, and greatly reduces labor costs in labor-intensive industries, which is tantamount to a myth in the eyes of investors from all walks of life.
In the industry carnival around 2016, autonomous driving companies that are not absolutely loyal to L4 are even regarded as absolutely disloyal to L4, and they will be laughed at by their peers and disdained by capital if they want to earn L2 money that is not decent enough at all.
Pony.ai was also born with an L4 birthmark, coupled with a star-level founding team lineup in the industry, which quickly attracted market scramble. According to public information statistics, it has received a total of 9 rounds of financing in the eight years since its establishment, involving an amount of more than 1.1 billion US dollars, including Sequoia China, IDG Capital, Wuyuan Capital, Toyota Motor, Kunlun Wanwei, Saudi Arabia Fund and other powerful investors, with a valuation of more than 8.5 billion US dollars.
The bursting of the bubble stems from the limited patience of capital. L4, which is in place in one step, faces objective factors that cannot be ignored such as policy supervision and complex scenarios in reality, and has never been able to bear commercial fruits of gold. From Robotaxi to Robotruck, from auxiliary safety officers to vehicle-road collaboration, the retreat of autonomous driving in the face of commercialization dilemma has instantly made capital lose interest. The driver becomes a safety officer, the open scene becomes a closed scene, and the story immediately becomes less sexy.
At the 2020 Pony.ai Internal All Hands Meeting, Lou Tiancheng, CTO, who is respected by the team as the "leader of the building", compared unmanned driving technology to a towering tree. He believes that if he does a good job of the trunk of the tree, he will be able to do other branch businesses, while Pony.ai's branch does not include L2 and only focuses on L4.
Dreams deserve encouragement, but they also have to be questioned by reality. On the eve of the layoffs at the end of 2022, Pony.ai CEO Peng Jun sent an email to all employees, saying that Pony.ai's primary goal is to survive.
To live is to earn money and do something. Pony.ai actually landed the autonomous truck business in 2020, but when it was laying off employees and transforming, it did not achieve self-hematopoiesis. At the beginning of 2023, Pony.ai officially announced the establishment of an independent business unit for the intelligent driving business of passenger cars, which is the L2 business that was once frowned upon. Now the former has become a pillar with a revenue of more than 100 million yuan, and the latter has also reached a cooperation with car companies in software service outsourcing. According to a person familiar with the matter, Pony.ai currently has abundant cash flow on its books.
Being able to bend and stretch is a necessary quality for a successful business, and although Pony.ai's broken arm survival is embarrassing, it is also worth emulating by peers who die on the beach when the tide recedes. Now that a new round of IPOs is sweeping, Pony.ai, which holds the script of the autonomous truck business, wants to go public in the United States, and has to be compared with TuSimple, which lost to the NASDAQ.
Part.2
Follow the guillotine road
Pony.ai's transformation of Robotruck is not a desperate idea, but an industry consensus, and with TuSimple's future jewel in front, it is indeed a choice with a brighter future. It's just that I didn't expect that the brilliance of the pearl and jade dissipated too quickly.
Founded in 2015, TuSimple has raised more than $800 million in 10 rounds of financing, and landed on the NASDAQ in April 2021 at a valuation of over $8 billion. The share price of the "world's first autonomous driving stock" also climbed to $79.8, but in the end, it shrank and fell by 99% in less than three years and was delisted, in which subjective and objective factors have played a significant role.
Robotruck is seen as a way back for Robotaxi, and the key reason is that the landing scenarios of autonomous freight are relatively simple, such as closed environments such as mines and ports. However, a few leading customers in a single industry obviously cannot support the high valuation of technology stocks, and realizing unmanned trunk logistics on a more open highway is the ultimate pursuit of Robotruck companies such as TuSimple Future.
Truck drivers in the United States belong to the high-class blue-collar class, and labor costs are expensive. TuSimple focused on North America early on, with the Chinese team only providing simple autonomous transportation services at the port. However, as a result, TuSimple's future North American business has never been able to operate on a large scale, and a large amount of investment in technology research and development can only be carried out on a small experimental basis. Its financial report shows that the revenue in the third quarter of 2023 is $26.53 million, but the net loss is as high as $113 million.
One of the barriers that cannot break through the market is that the truck freight market in both China and the United States is in a state of low threshold and high competition.
The United States Department of Transportation has estimated that there are more than 500,000 trucking companies operating more than 3.6 million Class 8 semi-trucks in the United States; The vast majority of them are small carriers, with 95% operating fewer than 20 semi-trucks and more than 84% operating fewer than 6 semi-trucks. On the other side of the ocean, the situation in China is not much different or even more scattered, and the "husband and wife car" driver bloggers commonly found on various content platforms are the most real epitome in the industry.
According to TuSimple's prospectus, Robotruck's profit model is generally divided into two types. One is to develop self-driving trucks, which are sold to a few powerful freight companies in the first few with hardware and software; The other is to sell self-driving transportation services based on mileage, which is equivalent to a freight version of the Didi platform.
The first model requires autonomous driving companies to have the ability to build cars, and the threshold is extremely high. TuSimple took the initiative to invite United States truck manufacturer Navistar to take a stake and pay $10 million to the latter for the joint development and mass production of new energy self-driving trucks, but finally announced the termination of the cooperation at the end of 2022, without any updated progress.
The second model requires autonomous driving companies to face competition from a large number of "retail" freight drivers, just as Didi challenged the traditional taxi industry in the past. The tragedy of online car-hailing subsidies is vivid, and TuSimple, which has already been listed, cannot afford the sky-high cost of market education in the future.
In the true sense of the unmanned trunk logistics, it also has to face a series of uncertain cost tests such as weather and vehicle speed, and TuSimple will almost enter a dead end in the future at both the business model and technology implementation level.
Part.3
The final window?
It's not just China's Tucson Future that has fallen in the United States market. In March last year, Embark Trucks, a domestic United States robotruck company with a market capitalization of more than $5 billion, announced its collapse, just 16 months after its listing.
The story of the autonomous truck has taught United States investors a lesson, and Pony.ai, which is now heading for an IPO in the United States, is more like a rush to start its own trial.
From lidar and smart chips on the hardware side to solutions on the software side, the autonomous driving industry chain ushered in a wave of small climaxes of Hong Kong stock listings last year. Not long after the news of Pony.ai's choice of overseas routes was released, some media reported that Momenta, which is also a leader in the industry, has secretly started the process of IPO in the United States, and negotiated with CICC, Goldman Sachs Group and UBS Group, and is expected to complete the listing as early as this year.
The same capital market has limited attention to the asset target in a field, and who can win the first opportunity in the secondary market of the same type of competitor will not affect the survival of life and death, but it is also related to the potential for subsequent development. Even if Pony.ai is not weak in survivability, it must seize the window to race against peers such as Momenta.
What's more, it's not so much the warming up of autonomous driving out of the cold winter, but the dark night just before dawn. According to statistics from industry research institutions, the total public financing in the field of autonomous driving in China will only exceed 20 billion yuan in 2023, which is even worse than nearly 30 billion yuan in 2022. Rushing to the market may be a scramble for one of the few lifesavers.
In fact, the focus of the capital level on autonomous driving has not shifted for a long time, but has focused on the "end-to-end" concept based on AI large model technology.
In May this year, Wayve.AI, a United Kingdom startup focusing on the development of end-to-end autonomous driving and autonomous driving models, officially announced that it had received a $1.05 billion financing from SoftBank, Nvidia and Microsoft, the last round of investors.
This new wave was set off by Tesla's FSD V12 released in August last year, and a number of domestic automakers and autonomous driving solution providers quickly followed. Huawei's latest release of Qiankun ADS3.0 is a promotional selling point with end-to-end technology, and He Xiaopeng also went to San Francisco, North America, to experience the differences between Tesla's FSD V12 and Google's waymo, and praised it.
Pony.ai's competitors cannot be limited to peers such as Momenta, who survived the cold winter.
The head of the new car-making forces, such as Wei Xiaoli, have their own intelligent driving teams, and even get involved in self-developed chips; Huawei and Baidu, which do not build their own cars, have been deeply engaged in the field of autonomous driving and AI for many years. Huawei's HarmonyOS Zhixing belongs to the standard L2 business, and Baidu's Robotaxi business Radish Express has also been reported frequently recently, and it is bound to face off against Pony.ai with a straight ball.
Nowadays, from the hot large-scale model track, a number of AI companies such as SenseTime, iFLYTEK, etc., have also been inserted, all of which regard the automotive scene as the best choice for the commercialization of large-scale model products.
In the field of autonomous driving, it seems that the spring water has finally melted away and the ice is rippling again. However, the Pony.ai people who have survived the cold winter are facing a new chaotic pattern. It's full of pitfalls and opportunities, but like a pony crossing a river, you still need to step into it to see it.