Image source @ Visual China
Text | Talking about AI, the author | Zheng drove
On December 21, wenyuan zhixing, an autonomous driving unicorn, held a cooperation and investment signing ceremony with GAC Group and Ruqi Chuxing, and the three parties will play their respective advantages in the future to promote robotaxi's model design and development, front-loading mass production and commercial operation.
At the conference, Wenyuan Zhixing also announced a strategic investment in Ruqi Travel, which set an industry precedent for the first autonomous driving start-up to invest in a travel platform.
Coincidentally, on December 8, SAIC's Xiangdao Travel also officially opened its Robotaxi operation, which was carried by SAIC's Feifan NARVEL R model, and Momenta provided autonomous driving solutions.
The Robotaxi industry has changed several times, with the virtual fire of Waymo fading, the aura is no longer there, and the new industry cooperation model of Wenyuan Zhixing has begun to appear, so in this round of new exploration and attempts, can it help Robotaxi players achieve the synergy effect of "1+1+1>3"?
Robotaxi, the virtual fire has faded, and the time has come
The autonomous driving track is classified from the technical level, mainly divided into two factions, one is tesla and other OEMs-based L2 starting, progressive climbing mode; the other is the L4 starting peak climbing mode based on technical unicorns, of which Robotaxi market prospects are the broadest, and thus gather a number of head players within the faction.
At present, the wave of automobile intelligence has struck, and assisted driving players led by Tesla have conducted practical demonstrations of their business models and gained market recognition.
With the support of most countries that already have policies to allow L2-level assisted driving right-of-way, Tesla can not only take the lead in tasting the fruits of large-scale profitability, directly sell cars equipped with L2 technology to the C-end, and in terms of very important data assets for autonomous driving, the huge user fleet also poses a dimensionality reduction blow to L4 starters.
Therefore, with the blessing of "dual assets", the development of this group of players is very rapid, and many new forces of smart cars have already equipped with L2+ auxiliary driving functions in the mass production models. Mercedes-Benz even obtained a road permit from the German Federal Traffic Authority for L3 autonomous driving this month.
Watching the assisted driving players begin to "feast on guests", the advanced autonomous driving players such as Robotaxi on the other side are still fighting over the question of whether the business model can run through. Not only is there a large number of Corner Cases (long-tail scenarios) to be overcome in technology, but in terms of policies and regulations, they are also circled in "one acre and three points of land" to a certain extent.
Tan Qing said that AI believes that for the autonomous driving track, safety, data, and policies are a mutually beneficial ecosystem, data can accelerate the breakthrough of security problems, security can accelerate the landing of policies and regulations, and policies are linked to the openness of road tests, which directly determines how big the potential data pool of technology companies is.
Starting from the demands of robotaxi players themselves, the policy is undoubtedly like a trigger, but now the trigger has been pulled.
At the beginning of this year, led by Baidu, jointly drafted by the Highway Research Institute of the Ministry of Transport and a number of enterprises, the Robotaxi technical requirements group standard was officially released, and on the basis of rules to follow, in November, Beijing officially opened the first Robotaxi commercialization pilot in China, allowing commercialized toll operations in some areas.
It is not difficult to find that as the entire Robotaxi track gradually enters the deep water area, with the participation of more and more players, the policies and regulations begin to be gradually and perfectly guided, and the amount of data in the future industry is also expected to usher in growth, assisting in breaking through more security technology bottlenecks, and the entire cycle will also tend to develop benignly.
At this time when it is expected to "reinvigorate the mighty wind", how can the enterprises that develop robotaxi seize the "time" of the policy? Reconstructing an industry model that was once restrictive is the first step.
Why is the "Iron Triangle" model starting to gain popularity?
19 years can be said to be the beginning of the chaos of Robotaxi's industry model exploration, Robotaxi is a B-end business in the short term, but in the long run, the essence is a C-end business, and therefore, in 19 years, we can often see such a situation, that is, the travel companies that can directly connect with downstream consumers, the enthusiasm is even beyond the technology unicorn in some aspects.
Uber, Lyft, Didi, and the three giants of online ride-hailing have begun to exert Their power on Robotaxi in 1919. First, Uber began placing orders from 2919, successively buying 24,000 self-driving customized XC90 SUV cars worth more than $1.9 billion from Volvo; in June 2019, Lyft also announced that Robotaxi, which cooperated with technology company Aptiv, had completed 50,000 services; Didi's self-driving team was officially upgraded to the company in 2019.
However, the travel company has reduced the technology company to a "hit worker" and has a "rough" card slot market play, and the good times are not long. Since the end of 2020, ride-hailing companies have withdrawn from the C position on the Robotaxi track.
In December 2020, Uber announced that it would sell its autonomous driving division, Advanced Technologies Group (ATG), to self-driving car startup Aurora Innovation, and in April 2021, Lyft would sell its self-driving division, Level 5, to Toyota subsidiary Woven Planet Holdings for $550 million in cash.
At this point, only didi autonomous driving company, the only seedling of the three giants, is still insisting, and the sound volume in the industry has quietly changed.
Before 20 years, the sound of technology companies on the Robotaxi track has been suppressed by cost, especially in the important perception device of lidar, which was once sold for as much as 80,000 US dollars for The Velodyne 64-line lidar once mounted by Google, and more importantly, there are very few production enterprises, and they may not be able to buy it.
Google, Baidu and other giants have been lidar card neck, not to mention a number of start-ups, Tesla seems to be in this specific period of industry phenomenon, forced to Liangshan to take the pure visual route, and then crazy criticism of lidar to "retaliate".
The turning point came 20 years later, at the CES (International Consumer Electronics Show) at the time, the thousand-dollar lidar debut, technology companies felt that they were doing it again, no longer subject to the high cost of perception components, and self-driving technology companies began to gradually move to the C position of Robotaxi.
However, compared with the "warm hospitality" of technology companies in previous years due to the bright prospects of the Robotaxi industry, in the past few years, although the cost of 20 years is spring, start-up players have ushered in the capital winter in the technical bottleneck, and they have to rethink the industry development model.
Until now, in the new round of thinking, players have begun to hand in their papers.
Players such as Wenyuan Zhixing and Momenta in China have given similar "iron triangle" answers, and on the other side of the ocean, in September this year, Mobileye also announced a cooperation with the travel platform SITT, planning to launch Robotaxi business in Munich, Germany, in 2022.
As a result, autonomous driving technology companies, ride-hailing companies, and OEMs are no longer competing for the real C position, and the trend of turning to cooperative and collaborative layout is more obvious.
How to understand the "Iron Triangle"? As the earliest autonomous driving enterprise to propose the "iron triangle" model, Wenyuan Zhixing disassembles this model into two parts, on the one hand, Wenyuan Zhixing and other technical enterprises cooperate with the main engine factory to form an autonomous driving capacity supplier; on the other hand, technology companies cooperate with platforms such as travel or freight, thus forming a capacity demand side in specific scenarios such as Robotaxi or unmanned freight.
Tan Qing said that AI believes that in such an industry model, the entire "iron triangle" has formed a supply and demand cycle of capacity that can be ecologically consistent.
The C-position change of the main team of Robotaxi is actually linked to the phased demand of the macro track, which is understandable, but now a more balanced cooperation model has emerged, and from the perspective of the entire track, it actually means that Robotaxi has reached a new stage of maturity.
After all, no one will bet on the patience of capital, and for the current track players, even waymo, no company can come up with its own technology to convince the market, convince capital, I can achieve 100% compliance and security on the basis of 100% no one.
Therefore, for players in the track, the next top priority is undoubtedly to collect more long-tail scene data from large-scale landing operations, further optimize algorithms, improve technical security, and then accelerate the landing process.
Technology, OEMs, travel platforms, how to divide the cake of Robotaxi?
On the Robotaxi track, technology, carrier, and operation cannot be separated from anyone, which is a consensus of the industry. However, the current increasingly strong landing demands of the industry have directly accelerated the emergence of the "iron triangle" cooperation model.
Back to the hottest years of financing for autonomous driving start-ups, considerable market prospects and some star team lineups are often the main reasons for leading capital to enter the market, but now, on the one hand, some star players in the track have appeared, and the overall financing has begun to be difficult; on the other hand, for some star players who have raised a lot of funds, money is not the most lacking now.
Tan Qing said that AI believes that if "financing" is the primary keyword of the 1.0 stage on the autonomous driving track, "landing" indicates that the entire track has entered the 2.0 stage, and pure financial investors can no longer become the thickest leg that technology companies need to climb the peak, and more importantly, let some collaborative industries participate.
In the automotive circle, there are many cases of joint ventures or cooperative car companies failing, and in the era of intelligence, the Zebra Zhixing cooperated by SAIC and Ali has actually given us corresponding thinking from highlight to fall.
Cooperation between enterprises often has basic elements such as learning from each other's strong points, jointly making the cake bigger, and long-term non-conflict of interest demands. In the new phase of Robotaxi's cooperation, how to ensure that the cake can be enjoyed by all parties? This is also a question that the "Iron Triangle" cooperation model must answer.
First of all, specifically looking at the demand to learn from each other's strengths, for self-driving startups, there is natural pressure to do Robotaxi in terms of the huge cost of fleet investment and the lack of consumer-facing operating experience.
If you want to speed up the progress of technology landing, you need not only the OEMs to provide them with vehicles that are more suitable for technology, but also a port like an online ride-hailing platform that is better at directly connecting consumers.
For the online ride-hailing platform, Robotaxi and today's ride-hailing business are actually fundamentally different, online car-hailing is a typical asset-light industry, and Robotaxi requires the release of a large number of driverless cars equipped with expensive equipment, which is a typical heavy asset.
At present, independent to do, not only the advantages are not obvious, Robotaxi technology, regulatory development is still not perfect, is undoubtedly to enter a bottomless pit to game, Uber, Lyft's successive folding, all reflect the travel company's single-handed drawbacks.
We can also see from the cooperation between Wenyuan Zhixing and Ruqi Travel, Momenta and Xiangdao Travel that whether it is a regional or national travel platform, it is also suitable for the Robotaxi "Iron Triangle" model, which actually reflects the thinking on the phased value of automatic driving landing.
Tan Qing said that AI believes that in addition to the difference between light and heavy assets between Robotaxi and online ride-hailing, there is also an essential difference in stages, that is, Robotaxi can only test and operate in a specific area at present, but online ride-hailing vehicles have no such restriction at all.
Taking WenyuanZhi as an example, although the cooperation of Ruqi Travel is not as large as that of an online ride-hailing giant like Didi, with more coverage of cities and richer operational experience, it is a key travel platform in the Guangdong-Hong Kong-Macao Greater Bay Area, and Wenyuan Zhixing cooperates with it, not only giving full play to the headquarters advantages of both sides, but also preemptively seizing the Robotaxi market in the Bay Area.
At present, when Robotaxi is landed, most players are actually still in their infancy, and the regional advantages of the two cooperative enterprises are consistent, and the advantages and disadvantages of the capabilities are complementary, which can often reflect a more stable trade-off of the elements of cooperation. At the same time, such a regional cooperation model in the later stage also has certain replicable value.
OEMs that do not seem to have much benefit in cooperation are actually not.
In the iron triangle cooperation model, it is not simply for the main engine factory to give the car to the automatic driving technology company, which then installs the automatic driving technology and hands it over to the online ride-hailing platform for operation, which also has a dark line of data value behind it.
This dark line can be a valuable asset of autonomous driving technology companies, for the main engine factory, the B-end scale landing of the new generation of models will also play a certain positive feedback role on the main business, not only can reduce the mass production cost of smart cars under the scale effect, launch more competitive intelligent products, but also is expected to promote technological upgrading through C-end data, and feed back the C-end market by enhancing product strength.
In general, the "Iron Triangle" cooperation model is valuable in complementary appeals, perhaps in the local process of cooperation, such as technology research and development, autonomous driving companies will dominate, but from the perspective of the overall cooperation process chain, from the test and development of technology companies to the new platform architecture of the main engine factory, and then to the final landing operation, there is no absolute macro dominance.
In addition, in the "iron triangle" cooperation model of Momenta and Wenyuan Zhixing, another point of reflection lies in the balance of control, which can actually reduce the risk of long-term consistency of interest claims to a certain extent.
In terms of control, SAIC actually has a lot of say, not only in the industry's well-known "body and soul theory" clearly stated that it will not cooperate with Huawei, in the previous cooperation with Ali Zebra Zhixing, has also been "injured" because of the needs of the two subjects on a special path.
In today's Robotaxi, we can actually see SAIC's quite cautious attitude, although it bluntly says that it does not cooperate with Huawei, but traditional car companies have natural shortcomings in automatic driving, saic still chose to cooperate with start-ups such as Momenta and invest in it.
On the other hand, Wenyuan Zhixing is also the same, Tianyancha information shows that on December 14, Wenyuan Zhixing received a new round of strategic financing from GAC Group, in this cooperation, Wenyuan Zhixing also invested in Ruqi Travel, and the entire "Iron Triangle" took a clearer step in control.
In the face of the current stage of Robotaxi landing appeal, perhaps Baidu, Google and other Internet giants can be in the operation, technology, car building three habitat C position layout.
But for self-driving technology start-ups, in the short term, embracing cooperation will become a more explosive way to play; in the long run, only by trying to balance control and interests can we more steadily control the unknown risks of cooperation.
At the end of the year, whether the "Iron Triangle" cooperation model will become a new milestone on the Robotaxi track will be left to 2022 to verify for us.
[Talk about Qing said AI, left turn new energy vehicles, right turn intelligent driving, there is depth, there is temperature, the author Zheng driving, reprint please retain copyright information. 】