Text: Zheng Driving @ Tan Qing said AI editor-in-chief
Can autonomous driving still be supported in recent years?
It seems that unlike the ADAS market that has often pondered L2++ or L2 Plus in recent times, L4 has always been the most ruthless existence in intelligent driving, silently carrying road test cars through the future sea of stars.
However, according to many media exposures recently, Waymo, which occupies the high-cold C position in the track, will also begin to provide a completely driverless taxi service in San Francisco, officially showing the internal strength of the retreat cultivation for many years to the market.
When we talk about commercialization of autonomous driving, what exactly are we talking about?
Today, when the wave finally photographed Google, the "OG" of the industry, it may be time to think about this problem.
Climb Mount Everest and lay eggs along the way
In the autonomous driving industry, there is often a saying that "climb Mount Everest, lay eggs along the way", in short, Mount Everest is an ultimate goal, it can also be a corporate mission, and egg laying is to maintain awe of the long-term value of Mount Everest and avoid being overlooked by the cold in the high place.
Yang Wenli, founder and CEO of Lingjun Technology, said to Tan Qing that AI said, "If the autonomous driving industry considers from the two dimensions of time and market maturity, it is divided into four stages, before 2021 is the core supply chain maturity period, 2021-2025 is small-scale commercialization, 2026-2030 is the reshaping of the business pattern, and free market competition after 2030." ”
So today, the industry has entered a new stage of commercialization from 1 to 10.
He Wei (pseudonym), a technician of an autonomous driving enterprise, believes that "startups are different from giants, the industry has developed to the present, if you want to survive, you must create commercial value as soon as possible, and the financing environment of the track is also cooling down, so we need to lay eggs." ”
One of the benefits of laying eggs is that it is better to help companies integrate into the commercialization tide that has already been filmed on the ankles of the industry. He Wei said that there are roughly two clear "egg laying" models in the commercialization stage, but the common point is that compared with climbing Everest, it is necessary to do dimensionality reduction to a certain extent compared to climbing Everest.
One is the scene dimensionality reduction, such as sinking to L4 autonomous driving logistics, unmanned minecarts and other relatively easy to land scenes; the other is the technical dimensionality reduction, do ADAS (assisted driving system) or subdivide a little, to AVP (automatic valet parking) such a specific energy production.
From the perspective of the current situation in the industry, He Wei's team currently prefers the mass production route, he believes that first of all, "although there are giants such as Baidu and Huawei in the (mass production) market, many new forces are also doing it themselves, but the market size of ADAS trillion is huge." ”
Another point is that under the market demand, the current certainty of ADAS mass production will be stronger, and the overall financing environment is still good vitality.
From the perspective of the capital market, an investor also told Tan Qing that AI said that the mass production route is indeed hotter in terms of investment now, and many second-tier projects currently have a good financing environment, and the head is like a foreign Mobileye has now started to list.
"Mainly (new energy) passenger car market up, many OEMs (assisted driving) demand in this regard is very high, Mobileye and Extreme Krypton cooperation, domestic such as Zongmu (technology), Momenta, these mass production startups, are also cooperating with OEMs to carry out more."
However, for the "laying of eggs", the industry is actually a status quo of seeking common ground while reserving differences. For example, multi-legged walking players such as Momenta and Xidi Zhixing, or lingjun technology, Tage Zhixing, etc. have begun to focus on the sinking of the scene, serving the mass production of car manufacturers is not the only choice in the commercialization stage.
On this point, He Wei believes that climbing Mount Everest is a long-term value, can not lay eggs, what eggs to lay, how many eggs to lay, see the wisdom, but in essence, they are all dormant, for climbing Everest to do more technical and industrial synergy experience reserves.
In fact, there is a certain degree of consensus in the industry, Xiaoma Zhixing CTO Lou Tiancheng once said in an interview, "I don't object to laying eggs along the way, provided that you are really Everest." If you judge that Everest is really Everest, I don't think laying eggs will affect it. ”
"We have made a lot of commercial attempts in recent years while adhering to the L4's passenger car technology stack to maintain the industry's leading position." Yang Wenli told Tan Qing that AI said, "In this way, we can have a clearer understanding of the pattern and future of the industry, have a deeper planning and thinking about how to reach Mount Everest in the future, and practice the truth." That's good. ”
"If it is good to lay eggs along the way without affecting the peak climbing before 2021, if you do too many eggs that are not related to the company's core business after 2021, the impact will be negative, the ability and energy of the company and the core team are limited, and it is for autonomous driving companies to focus and focus on the core business to have a longer-term future."
Admittedly, laying eggs is precious, and climbing peaks is more expensive.
So as more and more players gradually enter the new stage of the industry from 1 to 10, in terms of the integration of technology and industry, what subtle changes are taking place in the industry? Let's take a closer look.
Technology is king to global collaboration
In previous years, technology is the most popular in the macro track of automatic driving, after all, in the process of engaging in from 0 to 1, whether it is Tesla in the main engine factory, or the first tucson future of automatic driving, the head enterprise of the domestic track, Xiaoma Zhixing, Wenyuan Zhixing, at that time, the weight of software was greater, and technology occupied the vanguard position to a greater extent.
However, with the realization of autonomous driving technology from 0 to 1 in part in recent years, many start-ups have also had a premonition that it is likely to become more and more difficult to continue to rely on pure technology to subvert the traditional automobile industry.
If you want to cite the most dazzling unicorn in the self-driving track, then Xiaoma Zhixing is definitely one, but with the news of the shelving of the listing plan, the adjustment of the truck business, and the departure of key technicians last year, the aura of Xiaoma Zhixing seems to be fading.
Did Ma Zhi walk down the altar? We can't make up our minds for the time being, but compared to the $12 billion valuation of Xiaoma Zhixing when it sought to go public last year, Tianyancha App shows that after completing the latest Series D financing this month, its valuation has been temporarily locked in the less than ideal $8.5 billion.
He Wei believes that Xiaoma Zhixing was relatively conservative in commercialization before, but from the early commercialization stage of 1 to 10, it is difficult to maintain the vitality of closure. A new test for many start-ups today is whether the supply chain, partnerships, customers and other synergies can be combined with technology to make comprehensive competitiveness.
In fact, this also points out two directions for the technical collaboration of start-ups, one is to take care of both soft and hard, and the other is to give priority to software and seek cooperation on hardware.
In terms of soft and hard integration, according to the "Late Post" report, Xiaoma Zhixing began to consider car manufacturing last year, planning to build electric vehicles based on Geely's "SEA Haohan" platform, and recruited people from major OEMs, and got the joining of Xiaopeng design director. However, Xiaoma Zhixing said that he would not comment on this news at the time, and did not announce the relevant progress in the follow-up.
He Wei's company has also had relevant planning and attempts to personally do hardware, in the past experience, He Wei summed up two important trade-off points for doing hardware, one is the efficiency of technology, and the other is the input-output ratio of enterprises.
He Wei believes that "their own software and hardware integration, in the integration can indeed give technical personnel an efficiency improvement, this promotion can be a moat, but it will also give the capital chain, to the company's energy to bring great challenges, so the choice of this route is the need for multi-dimensional consideration, it is best not to easily enter the game, but the entry must be firm to go on to have the possibility of seeing more returns." ”
Although doing hardware has visible long-term value, it is undeniable that software is still the strongest barrier for most autonomous driving technology companies, and under the comprehensive trade-off, in-depth cooperation is actually a more favored choice for many startups.
For example, Wenyuan Zhixing and Momenta, which have a focus on Robotaxi's business, coincidentally, last year successively carried out the "iron triangle" in-depth cooperation model of technology companies, OEMs, and travel platforms.
Further, as a corner of the iron triangle, Ruqi Travel received strategic financing from Wenyuan Zhixing at the end of last year, and Lingjun Technology also received the Pre-A round invested by AI chip company Horizon last year, so in fact, it also allows us to see the trend of more and more deep collisions in various fields within the macro autonomous driving track.
Under such changes, Yang Wenli mentioned to us the cooperation model between Lingjun Technology and Horizon, "We provide help for Horizon in decision-making and planning, and Horizon supports us in chips." ”
It is not difficult to find that the cooperation model of the industry has begun to change and innovate in nature, no longer the same as the pure one-way supply and demand relationship of Robotaxi companies placing orders from OEMs or intelligent driving Tier1 manufacturers supplying technology companies.
So in today's world where everything is ready to go, how will startups get a tide in the same sea as giants such as Baidu and Google?
The New "Controversy"
Survival is a long-term thing for startups, and commercial attempts in the short term are actually accumulating strength for climbing the peak.
Regarding the peak of the industry, since last year, there has gradually emerged a problem that is difficult to ignore, that is, if you want to achieve a truly 100% safe driverless, it is almost impossible to achieve it by relying on the ability of the car end alone, which will not be much ambiguity in the industry.
Yang Wenli also agrees with us on this point, but he believes that "the proportion of road mileage that has been deployed in China has not yet started." ”
Therefore, for the choice of route, Yang Wenli still prefers to continue the strategy of maintaining the core and deepening cooperation. Yang Wenli said, "We have core advantages in bicycle intelligence, vehicle-road collaboration we will find advantages in technical cooperation, Nebulas Interconnection is the company created by our Tsinghua Automation Department teachers, Huali Zhixing is the company of Tsinghua Automation Department's brothers, and cooperation has a very good foundation." ”
However, since last year, in the autonomous driving industry, especially in some Chinese companies, the voice about vehicle-road collaboration has begun to boil, which seems to make people have such an intuitive feeling, that is, bicycle intelligence and vehicle collaboration are incompatible opposing camp relationships.
Just like the giant, Baidu and Google are in the two program routes of car road coordination and bicycle intelligence, winning and losing, and gradually making the sense of opposition become the basic cognition of many people on the two routes.
When referring to this layer of hidden worries, He Wei stressed to us, "Vehicle-road collaboration is a problem that the industry needs to consider from the 10th to 100th stage, and the bicycle intelligence is the main force of the 1st to 10th stage, and there is no direct contradiction." ”
In fact, this also has a commonality with Tan Qing's understanding that AI has always been, tan Qing said that AI believes that the two schemes of vehicle-road collaboration and bicycle intelligence seem to be opposites, but in fact, they are a certain degree of inclusion and inclusion relationship. Creating a car that is smart enough is the essence of the bicycle intelligent route, but it is actually one of the basic needs to ensure the smooth implementation of the vehicle-road coordination plan.
"At this stage, autonomous driving is actually a bit like the electrification reform of automakers, Wei Xiaoli hugged the group to warm up, and the public also looked for Tesla to learn from it." He Wei said, "Our industry is the same, in fact, at this stage, there is no need to have so many confrontations and be aggressive. In the early days of commercialization, I think there is still a friend and foe, and an open and win-win atmosphere is more conducive to everyone's development. ”
As he said, imagine if Tesla had not opened up a large amount of patented code, but used technical advantages to kill a number of new players outside the door, then what it needs to face today will only be the inertia of the era when fuel vehicles are more powerful.
In the matter of car building, today's Tesla can coexist and prosper with Wei Xiaoli, but in the sea of automatic driving, will the future be giants and start-ups swimming together, or the weak and the strong, winners take all?
Perhaps it is too early today, it is difficult for us to get a unified consensus in the industry. But what is certain is that in the sea ecology of autonomous driving, the drama that surges is being released.
Those who want to stand on the waves, they must begin to evolve a keen sense of smell in addition to technology.