1. Tesla will launch a new Robotaxi (self-driving taxi) product named "Cybercab"; 2. Under Baidu, companies such as Luobo Kuai Pao, Didi, Pony.ai, AutoX, Wenyuan, as well as those under GAC like Ruqi Chuxing, and SAIC's Xiangdao Chuxing have all laid out their Robotaxi business. Publicly listed companies like Wanma Technology are also exploring the Chinese model of Robotaxi.
Intelligent driving is attracting attention again.
On October 10th, California time (October 11th, Beijing time), $Tesla (TSLA.US)$ will release a new product, Robotaxi (self-driving taxi), named "Cybercab". Tesla CEO Elon Musk stated that this launch event will go down in history and he believes it is one of the most important moments for Tesla since Model 3.
Domestically, according to sources exclusive to "Science and Technology Innovation Board Daily", Luobo Kuai Pao plans to launch autonomous taxi services in Hong Kong, marking its first business launch outside the mainland.
According to Frost & Sullivan's forecast, China's Robotaxi market size will reach 488.8 billion yuan by 2030. Many industry analysts point out that Robotaxi may be one of the best scenarios for high-level intelligent driving to land. With the decrease in travel costs brought by Robotaxi, the travel market is expected to expand to a trillion-dollar scale.
Currently, $BIDU-SW (09888.HK)$ Luobo Express, Didi, Pony AI, AutoX, Wenyuan, and $GAC GROUP (02238.HK)$ its subsidiary $CHENQI TECH (09680.HK)$ , SAIC's Xiangdao Travel and other companies under SAIC have also laid out Robotaxi business. Wanma Technology and other listed companies are also exploring the China model of Tesla's intelligent driving platform.
However, the precise driving and decision-making capabilities of self-driving cars are attributed to the collection, processing, and utilization of a large amount of data. This is because large-scale, high-quality, and extensive data are key nutrients for training and tuning AI models. The evident "data island effect" among domestic car companies has not yet fully realized its value, impacting the commercialization of Robotaxi and even high-level intelligent driving.
Tesla Robotaxi release, catalyzing the development of the smart driving industry
Currently, the USA is one of the fastest countries in the world to promote the commercialization of autonomous driving. According to IDTechEX data, by the end of 2023, Robotaxi has been tested in 15 cities in the USA, with commercial deployment achieved in 4 cities.
Among all participants, the commercial progress of the autonomous driving leader Waymo is the most advanced, achieving driverless operations in Phoenix, San Francisco, and Los Angeles.
On August 20th this year, Waymo announced that the weekly paid order volume of its Robotaxi service has exceeded 0.1 million, doubling since May. Companies like Cruise, Zoox are also following suit, currently in the testing phase.
Electric vehicle giant Tesla also announced its entry into Robotaxi and will launch a new Robotaxi product on October 10th local time.
As a new entrant in the Robotaxi industry, Tesla adopts a progressive route, upgrading from autonomous driving L2 level to L4 level gradually. CITIC Securities believes that the USA's commercial progress in autonomous driving is leading globally, with an increasingly refined regulatory system. With Tesla's progressive approach to releasing Robotaxi, further industry development is expected. Overseas leading OEMs in smart driving, intelligent industry chain, and Tesla industry chain symbols will all benefit from the rapid development of Robotaxi.
2025 may become the first year of commercialization of Robotaxi globally
Frost & Sullivan predicts that Robotaxi will achieve large-scale commercialization around 2026. By 2030, the market size of Robotaxi in China and globally will reach 488.8 billion / 834.9 billion yuan respectively, with CAGRs of 248% / 239% for 24-30 years, reaching 1.6 trillion/3.2 trillion yuan respectively by 2035, with CAGRs of 27.4%/31% for 30-35 years.
Haitong Securities believes that Tesla's entry will drive a major transformation in the shared travel industry, with 2025 potentially becoming the first year of global Robotaxi commercialization. Open Source Securities also mentioned that the market brought by Robotaxi will not be limited to the existing online car-hailing market. The unit cost of Robotaxi travel is much lower than that of human-driven taxis, attracting a large number of customers who didn't take taxis before, thus stimulating market expansion.
In China, companies like Luobo Kuaipao, Didi, Pony.ai, AutoX, Wenyuan under Baidu, Ruqi Chuxing under GAC, Xiangdao Chuxing under SAIC, and others have already deployed Robotaxi business.
In July of this year, 20 cities including Beijing, Shanghai, Chongqing, Guangzhou, and Shenzhen were designated as pilot cities for self-driving. Apart from defining operational test areas for testing self-driving car-hailing services, various places are also preparing for relevant legislation to support the development of self-driving business. For example, Shenzhen is considering introducing autonomous buses, Guangzhou plans to legislate to support autonomous taxis and buses, Beijing plans to introduce regulations on self-driving cars, and so on.
According to the "Science and Technology Innovation Board Daily", wanma technology is currently cooperating with domestic mainframe manufacturers and related companies to further promote the implementation of Robotaxi in China.
Previously, Wanma Technology's subsidiary, Youka Technology, has officially signed an investment agreement with the Management Committee of the National High-Tech Zone in Changzhou, Jiangsu Province, with a planned total investment of approximately 0.101 billion yuan. In August, Changzhou's first "Edge Cloud-side" self-driving cloud computing platform - the "Youka Self-Driving Cloud Service Intelligence Center" was unveiled. This Intelligence Center adopts $NVIDIA (NVDA.US)$H20 servers, aiming to build a closed-loop toolchain platform for self-driving data and an intelligent driving development platform.
"Wanma Technology has already partnered with relevant companies to create point-to-point Robotaxi routes in Changzhou. Robotaxi is well suited for segmented scenarios of public transportation between points, especially for leisure and entertainment, routes where there are no time constraints." said the above-industry sources.
In addition to conducting scenario-based verification tests for Robotaxi operations, Ukai Technology will collaborate with the Changzhou government in the Changzhou High-Tech Zone to select suitable test sections for innovative research and development testing in directions such as the integration of automated single-car intelligence and V2X (vehicle to everything) cloud.
Data is the key to the implementation of smart driving.
Industry experts in the field of AI analyzed to the 'Star Market Daily' journalist that from the perspective of AI, the intelligentization of the automotive industry is a good entry point and also a very important direction for the future development of AI.
"Taking Tesla as an example, Musk believes that Tesla is not simply an automotive company, but an enterprise focused on AI/robots. Tesla cars are traffic robots, and they are also developing Optimus Prime humanoid robots."
To truly achieve the intelligentization of the automotive industry, data is essential. 'Data, algorithms, and computing power are the three elements of AI. Among them, data is fundamental. Without data, algorithms cannot be developed. Intelligent driving requires extensive data collection for training in the cloud.' A senior professional in the field of vehicle networking shared with the 'Star Market Daily' journalist.
This is currently one of the challenges for the intelligent transformation of domestic automotive enterprises. For instance, in terms of intelligent driving data, as of April 2024, Tesla's FSD has accumulated a total driving distance of 1.25 billion miles (2 billion kilometers), showing a significant advantage over Huawei's 0.22 billion kilometers and Xiaopeng XNGP's 0.07 billion kilometers.
"In terms of overall data accumulation, China's automotive industry still lags behind the United States in total accumulation and data quality. Domestic OEMs do not know how to collect data during the intelligent transformation process. Some car companies have not fully and compliantly collected and utilized a lot of autonomous driving data to reduce storage costs." mentioned the above-mentioned expert.
To address the challenges in training autonomous driving models in the cloud, Tesla has created the computing platform called Dojo. Musk views it as the computing infrastructure of Tesla's AI business system, and a key technology support for achieving autonomous driving.
The 'Star Market Daily' journalist learned that Wanma Technology also aims to help car manufacturers build a Chinese version of Tesla's Dojo, reducing the cost of collecting and storing data, enabling traditional OEMs to quickly grasp intelligent capabilities. Currently, their platform has connected vehicles exceeding 11 million units, providing relevant products and services to mainstream manufacturers like Geely, SAIC Volkswagen, Dongfeng, Zhiji, Lantu, and Ideal as well as emerging forces.
Intelligent driving technologies including Robotaxi require lower network latency and more significant data processing. Compared to 4G networks, 5G networks have advantages such as high speed, low latency, and large connectivity. The low latency feature of 5G enables real-time communication between vehicles and between vehicles and infrastructure, enhancing the response speed and safety of onboard systems. Additionally, the high bandwidth and large capacity characteristics bring a new experience for onboard entertainment systems, vehicle networking applications, and more.
However, several industry insiders told Science and Technology Innovation Board Daily that due to various reasons such as cost and awareness, the penetration rate of 5G in the field of vehicle networking is still low.
"Many car companies spend hundreds of millions to develop smart cabins, but the vehicles still use 4G communication modules. What is the biggest perceptual difference between 5G and 4G? For example, sometimes when a vehicle takes the wrong route, it takes one minute to update the map in a 4G network, while in a 5G network, it only takes three to five seconds." mentioned a practitioner in the field of vehicle networking.
To further enhance the experience of intelligent networked automobiles, 5G is an important development direction. The above-mentioned vehicle networking experts suggest that in the future, the intelligentization of the automotive industry will definitely be based on the high-speed, low-latency 5G network. "Intelligent driving must be based on 5G, and it must be dual-card dual-standby to avoid constant data disconnection."
Editor/Rocky