The report shared today is: Intelligent Driving Annual Report 2023.
Selected Reports**: Industry Report Think Tank Summary: The report shows that with the continuous deepening of automotive intelligence, intelligent driving products are at a historical moment that is about to be popularized. From ADAS advanced driver assistance systems to fully autonomous vehicles, intelligent driving technologies are changing the way we travel at an astonishing rate.
Intelligent driving products can be divided into two categories: driving and parking. Driving functions include active safety functions, high-speed autonomous driving, and city navigation, while parking covers scenarios such as memory parking and AI valet parking. In the realization of intelligent driving, driving and parking are the applications of AI systems in different environments, among which the parking scene can be regarded as the application of AI capabilities in a relatively simple environment, which reduces the demand for AI algorithms and releases more driving pressure.
China's intelligent driving products mainly realize L2 level of intelligent driving, mainly including cameras, millimeter-wave radar, ultrasonic radar and lidar in terms of sensors. These sensors provide critical information to intelligent driving systems, helping the vehicle perceive and react accordingly. Intelligent driving systems also rely on high-precision map data to provide more accurate environmental information and reduce the complexity of algorithms.
At present, intelligent driving product providers can be divided into the leading generation, the next generation and the current generation, as well as the self-developed school, the first business school and the redundant school. The development of intelligent driving is inseparable from three key elements: algorithms, computing power, and data. At present, the intelligent driving algorithm tends to be consistent, and a unified scheme of BEV and Transformer is beginning to be used. In terms of computing power, Tesla has its own FSD chips, and other major players rely on Nvidia and other leading companies. The improvement of algorithms and computing power directly affects the technical level of intelligent driving, which in turn affects the competitiveness and market performance of products.
The key elements of the iteration of intelligent driving capabilities include the level of technology, the number of production vehicles, and the deployment of sensors. Achieving a higher level of intelligent driving products has a leading edge, and more smart car sales mean more data accumulation, which contributes to the continuous progress of technology. Models equipped with sensors as standard are more likely to achieve technological iteration, bringing more possibilities for the development of intelligent driving.
Excerpts from the report are set out below.