Editor's note: "He Xiongsong takes you to read the commercialization of autonomous driving scenarios", takes you to understand the commercialization of different scenarios in the field of automatic driving, and objectively and rationally interprets the evolution of investment strategies behind commercialization.
This column is jointly produced by He Xiongsong, executive general manager of Chentao Capital, and the heart of the car, and is updated every Saturday, and the content is exclusively authorized to be released by the heart of the car.
Hello, I am Chentao Capital He Xiongsong, welcome to the second lecture of "He Xiongsong Takes You to Read the Commercialization of Autonomous Driving Scenes".
Let's understand the biggest opportunities and challenges in the autonomous driving industry, and analyze why everyone is so interested in automatic driving and so many investment institutions are firmly pouring into this track. And what are the difficulties behind this?
With these questions in mind, let's explore them together.
In 2021, autonomous driving in commercial scenarios has gained more attention and recognition.
Track financing such as trunk lines, mines, ports, sanitation, and unmanned logistics cars continues to be hot, and autonomous driving veterans and novices have laid out commercial scenarios.
In the trunk logistics scene alone, more than 6 start-ups have emerged: Xiaoma Zhixing has fissioned out 3 companies, namely: Qianhang Technology, Kinte Smart Card and Xingjiao Technology; Tucson ChenMo founded Turing Smart Card; Baidu and Shiqiao jointly established truck brands DeepWay, Geely is also launching remote cars.
The heat of the track is evident.
Why do you like this track so much, to sum up there are two reasons, one is the trend is determined, and the other is that the market space is large. First, let's analyze why the trend is determined.
01
The trend is determined and the market space is large
(1) The development trend of automatic driving is determined
1) Labor shortage situation is grim
According to Prospective Economist, our working-age workforce has continued to decline since 2013, with a cumulative decline of 23 million over seven years.
Combined with the data of the birth population over the years, it can be known that the mainland is gradually entering an aging society, and the trend of decreasing the population of the working-age labor force is not likely to be reversed in the short and medium term.
The shortage of employment caused by the reduction in labor supply first affects industries and jobs with harsh working environments, poor experience and low incomes.
Moreover, with the improvement of living standards, most of the working-age laborers, especially young people, are no longer willing to work as drivers in trunk lines, mines, ports and other scenes.
We once interviewed a company with a head LTL company whose average age of drivers is increasing by 0.7 years per year, and the company can only repeatedly relax the age requirements for drivers, but it is still difficult to meet the recruitment target, and they are worried about the future of driver recruitment.
There is a similar situation in the express delivery and delivery industry, because it is impossible to recruit riders, and many express delivery sites have been forced to withdraw.
Under the dual influence of dwindling labour supply and increasing demand, the shortage of drivers has become a "grey rhinoceros incident".
Although we are also doing program exploration other than automatic driving, such as exploring drones, robotic dogs and other solutions in scenarios such as terminal distribution, in general, the determination to solve the problem of driver shortage through automatic driving solutions is the highest and most worth looking forward to.
The main contradiction of the "driver shortage" is important for commercial analysis in the field of autonomous driving.
Taking unmanned logistics trolleys as an example, unmanned logistics trolleys have declined compared with manual distribution because they cannot go upstairs, and many people are worried that their business logic is not smooth enough.
However, our field research found that there are already a lot of communities and parks that cannot provide delivery services upstairs and home because of insufficient manpower, or prohibit delivery personnel from entering the park or going upstairs because of management needs and epidemic prevention needs.
We believe that even if we do not consider the future of small robots or human-machine collaboration to solve the problem of going upstairs, the business model of unmanned logistics trolleys is reasonable.
Because, in the current situation of shortage and increasing manpower supply, the future needs to weigh not the problem of having and optimizing, but the problem of having and not having.
2) Autonomous driving can achieve cost reduction and efficiency increase
Taking the mine as an example, a mining wide-body vehicle generally requires 2-3 driver shifts, the driver fee for a car matching is about 300,000 yuan per year, and the cost of the automatic driving kit is calculated at 500,000 yuan, assuming 5 years of depreciation.
Comprehensively, after the maturity of automatic driving technology, a car can save nearly 200,000 yuan in costs per year.
There are also two trends that need special attention, one is that the cost of labor is on the rise, and the other is that the costs associated with autonomous driving kits are in a downward trend.
This means that over time, the economic value of autonomous driving kits will become more and more apparent.
The above is an analysis of the trend of the autonomous driving track, in addition to the trend determination, another reason why the track is particularly popular is that the market space is large.
(2) The market space is large
From the perspective of market space, the autonomous driving industry is one of the few top tracks. From the operational point of view, trunk autonomous driving and Robotaxi have trillions of market space, sanitation, mining, terminal distribution and other scenarios are 100 billion tracks, and ports are 10 billion tracks.
The depth of the water can be big, and the autonomous driving industry has the opportunity to give birth to super unicorns.
02
There are many challenges, including technology and supply chain
The above analysis of the reasons why the track is full of charm, summed up that the trend is determined and the market space is large. However, objectively speaking, there are still many challenges in the development of the industry.
We see that the industry leader Waymo has not been smooth in commercialization even after more than 10 years of hard work.
Although a hundred flowers bloom in China, except for some specific scenes, the timetable for batch landing is still not clear enough.
What are the challenges of landing? We believe the biggest challenges lie in the long-tail scenario and supply chain, and right-of-way and business models are not the main obstacles.
(1) The long-tail scene is the most important technical difficulty
The solution of the long-tail scenario is the biggest difficulty of autonomous driving technology.
The difficulty of autonomous driving technology is not in the underlying architecture and algorithm framework, these aspects have mature solutions, each autonomous driving company has no essential difference in the program, even if there is will be gradually smoothed out with the advancement of research and development.
The real obstacle to the landing of automatic driving lies in the solution of the problem of long-tail scenes, which are the key to landing and the decisive player of competition between autonomous driving companies.
Why is the long tail scene a decider?
The autonomous driving function is based on AI algorithms. With current technology, AI algorithms can only complete previously trained tasks, which is fundamentally different from general-purpose artificial intelligence that can reason based on prior knowledge.
This means that if you want to overcome various scenarios for autonomous vehicles, you must find and train these scenarios in advance. Common scenes are particularly easy to find, but there are many special scenes that are very difficult to encounter, which is the so-called "long tail scene".
In theory, we can only approach infinitely and it is impossible to completely find all the scenes, so there are too many "long-tail scenarios" that have not been solved, and the accident rate and failure rate of automatic driving will remain high, which will not only seriously affect safety, but also cannot be commercialized or scaled because of the need for various takeovers.
(2) There is still a gap between the supply chain and meeting demand
On the one hand, the existing assisted driving supply chain is difficult to meet the needs of autonomous driving.
Assisted driving means that the driver is the main driver in the driving, the system is assisted, and the automatic driving is based on the system, the driver is assisted, and even the driver is completely unnecessary.
There are big differences in design concept and function between the two, for example, autonomous driving has higher safety requirements than assisted driving, and requires more redundant backups.
At present, it is difficult to fully reuse the assisted driving supply chain to autonomous driving products, but the real large-scale shipments on the market are assisted driving related products, for most suppliers, the development of components supporting automatic driving is difficult and there are no batch orders for the time being, so the research and development power is insufficient.
Many autonomous driving companies have customized development needs for scenarios, but the degree of cooperation between suppliers is not high enough.
Taking the chassis-by-wire control related components as an example, if you want to achieve automatic driving, there must be redundancy, and some functions need to be specially developed, such as in the mining scene, the chassis redundancy design of mining vehicles is difficult to simply apply the scheme of ordinary collection cards.
Therefore, at present, many autonomous driving companies have been forced to personally go down to do chassis-related research and development.
On the other hand, sensors related to autonomous driving are not yet mature. Some sensors developed for autonomous driving, such as lidar and chips, are still in the iterative process and are still far from the car rules.
Moreover, commercial vehicles are more complex than passenger cars, such as mining scenarios, poor road environment, large temperature difference between day and night, and low outdoor operating temperature in winter, and general lidar cannot meet the requirements.
(c) Right-of-way and business models are not major obstacles
In addition to the long-tail scenario and supply chain challenges, many people also think of right-of-way and business models. But we don't think these two points are major obstacles in the long run.
Let's start with the right of way. The mainland government does have limited access to the right of way for autonomous vehicles, but at present we are still relatively long from the full maturity of the technology, and I believe that when the automatic driving technology matures, the right of way will not be the main obstacle.
There are the following reasons.
On the one hand, the government generally maintains an open and supportive attitude towards autonomous driving technology, and places such as Beijing and Shenzhen have actively promoted relevant legislation.
On the other hand, from the government's point of view, their main concern is the safety of autonomous vehicles, and if they can prove that the self-driving technology is really mature enough, their concerns about safety will not exist.
The second is the problem of the business model. We briefly analyzed the business model of several autonomous driving scenarios above, and from the calculation results, the cost reduction effect achieved by the automatic driving suite after the implementation of L4 is very obvious.
There are two other factors that need to be emphasized again. First of all, labor costs are rising; secondly, the cost of autonomous driving kits will continue to decline with the maturity of the industrial chain and the amount of products.
On the whole, once the commercial model of autonomous driving has passed the inflection point, the economic benefits will become higher and higher. So, in the long run, the business model is not a major obstacle.
Well, the above is an analysis of the challenges related to the autonomous driving industry, and this is basically the end of the talk.
In this speech, I mainly shared the biggest opportunities and challenges in the field of autonomous driving.
I think the opportunity for the autonomous driving industry is that the trends are very certain and the market space is huge. At the same time, the main challenges are the long-tail scenario and the solution of supply chain problems, etc., but I don't think the right-of-way and business models will be the main obstacles.
Finally, I would like to invite you to talk about what you think about the drivers of the autonomous driving industry? What challenges are you worried about landing?
We look forward to your sharing in the message area.
That's all for this talk, and we'll see you next.
Draw the point
Autonomous driving does not solve the problem of having and being excellent, but the problem of having and not having;
The long-tail scenario is the most important technical difficulty; right-of-way and business models will not be the main obstacles.
Produced in this issue
Speaker: He Xiongsong Producer: Zhu Shan
Editor: Ye Fang Later: Zhu Shan
Design: Chen Xiyang Operation: Lighthouse
Audio platform
Himalayas | microcosm
Podcasts | Little Goose Pass
Search for the heart of the car on the above platform
to listen to all the shows
"The Heart of the Car · The connoisseur said" trailer
Thursday, March 10, 20:00-21:00 pm, E-Cafe CTO Bai Junbo will be a guest car heart · The connoisseur said, "What rhythm will the hot skateboard site land?"