A single smart vehicle can produce massive amounts of data in a single day.
Gartner estimates that an autonomous vehicle generates at least 4TB of data daily with the usage of LiDAR (Light Detection and Ranging), cameras, and other sensors that allow the vehicle to observe the road. Furthermore, every increase in the level of autonomous driving drives demand for computing power at least ten-fold, thus further propelling the demand for cloud services.
The reason for this demand is that data mining and processing cannot be done without the support of the cloud and computing power. As such, vehicle cloud technology is crucial to the next phase of competition in the field of smart vehicles. According to Chinese publication 21CBR, vehicle cloud technology witnessed rapid development in 2022, with projections for revenue being 2.5 times higher than in 2020.
As vehicle manufacturers shift from testing to full commercial implementation, the need for vehicle cloud technology becomes evident. In the testing phase, data processing for smart vehicles does not present a major obstacle for manufacturers as this phase requires fewer vehicles. Some autonomous driving companies will even use hard drives to copy data and carry out data mining. However, as the number of vehicles soars, companies require data centers to support storage and computing requirements.
Private Cloud Solutions and Their Issues
Some companies will handle these requirements by constructing their own super-computing centers. In August 2022, XPeng, for example, announced the construction of what was to be the largest autonomous driving intelligent computing center in China to handle the data processing for autonomous driving model training.
Vehicle manufacturers that build private clouds, however, must continuously expand their private cloud capacity in the face of increasing demand for data storage and high computing power as their smart vehicles hit the market.
A smart vehicle generates about 1PB (1,000TB) of data per year and requires some 28 36TB portable hard drives for storage. The average price of a 36TB portable hard drive is RMB 6,000 (USD 872), making the total cost RMB 168,000 (USD 24,420). Therefore, an enterprise with an annual sales volume of 100,000 smart cars could have data storage costs of up to RMB 16.8 billion (USD 2.44 billion).
Public Cloud Solutions as an Alternative
Rather than construct private clouds, vehicle manufacturers are increasingly turning to public cloud vendors to handle front-end data collection, data annotation, and back-end data services. Public clouds offer more cost-effective solutions, strong computing power, and data management abilities, making them a popular choice for many new energy vehicle (NEV) manufacturers.
These cost-effective options could prove crucial in the future as the usage of smart cars increases. According to the National Development and Reform Commission, a Chinese macroeconomic management agency, the penetration rate of smart cars in China is expected to reach 82% in 2025, with up to 28 million units, and will reach 95% in 2030 with about 38 million vehicles. Demand for vehicle cloud services will grow as the smart car penetration rate continues to rise, meaning that the vehicle cloud sector will also grow rapidly in the future.
Cloud services can ensure that the relationship between vehicle manufacturers and users does not stop after the sale, but continues to create value and revenue down the road. For example, vehicle cloud technology can support remote diagnosis, roadside assistance, fault monitoring, status detection, used car evaluation, and other applications, thus increasing revenue for vehicle manufacturers.
Data from business consulting firm Frost & Sullivan shows that in 2021, the annual budget of Chinese small vehicle manufacturers for vehicle cloud and other sectors was approximately RMB 10 million (USD 1.4 million), while group-level large vehicle enterprises had budgets of around RMB 50 million and showed more rapid growth. Data showed that the overall market size of the vehicle cloud industry was as high as RMB 33.52 billion (USD 4.87 billion) in 2021; by 2026, this figure is projected to increase to RMB 80 billion, almost tripling within three years.
Current Market Players
The entry of cloud vendors, such as Alibaba, Tencent, Huawei, and Baidu, into the vehicle cloud sector highlights its importance as a key battleground for many cloud enterprises.
According to the same data from Frost & Sullivan, Huawei Cloud, Alibaba Cloud, Baidu Cloud, and Tencent Cloud account for nearly 70% of the market. In the field of autonomous driving, Huawei Cloud, Alibaba Cloud, and Tencent Cloud account for 26.9%, 16%, and 15%, respectively. In the Internet of Vehicles sector, Huawei Cloud, Alibaba Cloud, and Tencent Cloud account for 20.2%, 18.9%, and 18.3%, respectively. Finally, in the area of vehicle-road collaboration, Huawei Cloud, Baidu Cloud, and Alibaba Cloud account for 24%, 22%, and 18%, respectively.
Baidu, Alibaba, Tencent, ByteDance, Huawei, and other leading cloud service providers all made their move into the vehicle cloud sector between 2021 and 2022.
One of the earliest to do so was ByteDance, which formed its vehicle cloud team in June 2021. In November 2021, Tencent Cloud officially announced the launch of its vehicle cloud, and in June 2022, it announced a strategic upgrade, proposing a vehicle-cloud integration strategy. It introduced a whole set of independently developed products that can be installed in vehicles, establishing a data-driven closed loop to integrate the cloud with vehicles and mobile devices and reduce migration costs for vehicle manufacturers.
In September 2022, Baidu officially launched three clouds for the vehicle industry. Baidu’s three clouds are currently dedicated to establishing digital bases for vehicle manufacturers, connecting them with upstream and downstream supporting enterprises, and providing intelligent cloud solutions.
Alibaba Cloud and XPeng jointly completed the Fuyao autonomous driving intelligent computing center in August 2022. Alibaba already has experience in autonomous driving and computer-aided engineering simulation and has landed products such as autonomous logistics robots and autonomous driving trucks. The smart vehicle jointly developed by Alibaba and SAIC Motor is also one of Alibaba Cloud’s key products.
Huawei is also betting on the vehicle industry and has built a stack of autonomous driving cloud services, creating an integrated solution for manufacturing, production, supply, marketing, after-sales, and applications. In fact, Huawei Cloud established an Internet of Vehicles Laboratory as early as 2014, giving it an early advantage in building up technology reserves for applications such as intelligent connected electric vehicles.
Cloud vendors are providing a growing range of robust services that cater to the needs of NEV manufacturers in terms of computer processing and data storage for smart vehicles. However, manufacturers have strict requirements for protecting the confidentiality of their core technologies, which presents challenges for vehicle manufacturers that seek to leverage public cloud services. Data protection is therefore the next area of concern that needs to be addressed — especially in a world where user privacy is paramount.
This article was adapted based on a feature originally written by Chebai Think Tank. 36Kr Global is authorized to translate, adapt, and publish its contents.
