Ping An Securities: Tesla, Inc. (TSLA.US) brings autonomous driving assistance to China, accelerating the pace of intelligent driving development.

Ping An Securities released a research report saying that considering that the training of computing power and data collection require a certain period of time, it is expected that the release pace of Tesla, Inc.'s (TSLA.US) urban road assisted driving function is controllable, and the optimization iteration is progressive. However, based on the good generalization of end-to-end intelligent driving solution, Tesla, Inc.'s high vehicle ownership base in China, and Tesla, Inc.'s long-term accumulation of data closed-loop capability will make Tesla, Inc.'s urban assisted driving function experience in China's road conditions worthy of high expectations, making it a heavyweight in the intelligent driving industry. This update of Tesla, Inc.'s urban Advanced Driving Assistance Software has far-reaching significance in further driving the development pace of the intelligent driving industry in China and forcing local car companies to accelerate the adoption of various intelligent driving solutions. Event: Tesla, Inc. officially pushed a software update, stating that version 2024.45.32.12 has started to be rolled out in batches, with main upgrade content including: 1) Urban road Autopilot automatic assisted driving (optimizing the existing NOA automatic assisted navigation driving function); 2) Driver's cabin camera; 3) Map package version update. The main points of Ping An Securities are as follows: The introduction of Tesla, Inc.'s urban assisted driving in China will significantly accelerate the pace of development in the intelligent driving field in China. This software update from Tesla, Inc. did not use the term FSD, but the updated features such as "urban road Autopilot automatic assisted driving" and "on and off ramps, left turn, right turn, U-turn, lane change" generally refer to urban intelligent driving functions. As early as September 2024, the Tesla AI official account had shown that FSD would be launched in China in the first quarter of 2025 (subject to regulatory approval). The firm believes that this software update is of great significance for both Tesla, Inc. and the development of intelligent driving in China. Tesla, Inc.'s FSD transition to end-to-end improves generalization capabilities (across multiple scenarios and regions), enhances interactions, and makes driving more human-like. Although this software update is limited to some vehicles and rolled out in batches, Tesla, Inc. has a large ownership base in China, and its end-to-end intelligent driving algorithm has strong generalization capabilities, so Tesla, Inc.'s intelligent driving experience in China has the potential to achieve ideal results in a relatively short period of time. In 2025, economical car intelligent driving equal rights and full-scene intelligent driving will advance together, significantly accelerating the pace of intelligent driving development. In February 2025, BYD Company Limited held a strategic conference on intelligence, with high-speed assisted driving functions covering the market segment below 200,000 units. Given the strong industry influence of BYD Company Limited, other car companies will quickly follow suit with equal rights in intelligent driving strategy, and companies in the intelligent driving related supply chain will significantly benefit. In the report "Mileage 2.0, AI Intelligent Driving 1.0", the firm pointed out that in the era of AI intelligent driving 1.0, there is a battle of experience and scale. The release of the parking-to-parking intelligent driving function based on end-to-end intelligent driving solutions is the first touchstone of the intelligent driving experience in 2025. Adaptability to scenarios, take-over mileage, and human-likeness are the three dimensions for evaluating the AI intelligent driving experience. It is necessary for new vehicles above 200,000 units to have high-order intelligent driving as a standard feature, and it was pointed out that the specific evolution steps of high-level intelligent driving in 2025 are: leading car companies for high-level intelligent driving completing the rollout of parking-to-parking functionality by the end of 2024 or the first half of 2025. In the second half of 2025, the average take-over mileage of full-scene intelligent driving will significantly increase, the smoothness and human-likeness of the full-scene intelligent driving experience will continue to improve, and some scenes will achieve unmanned driving. The release of Tesla, Inc.'s urban assisted driving features will boost owners' understanding and confidence in high-level intelligent driving, and explore the reasonable pricing boundaries for intelligent driving packages. Currently, domestic car companies offer high-level intelligent driving as a high-end option for consumers. Due to the varying experiences of city navigation intelligent driving among different brands, the pricing of intelligent driving software packages has not yet been explored. However, Tesla, Inc.'s FSD intelligent driving package already has a stable price in the United States, and after the release of Tesla, Inc.'s urban road assisted driving in China, it may explore a pricing level for intelligent driving packages suitable for the domestic market. As a global leader in smart cars, Tesla, Inc. has a high user base in China. If its high-level assisted driving functions are widely experienced by local consumers, it will help boost the awareness and confidence of the general public in high-level intelligent driving, making intelligent driving a more critical factor affecting consumer purchasing behavior. Forcing local leading intelligent driving companies to accelerate the reduction of intelligent driving hardware costs. Tesla, Inc.'s FSD vehicle-side perception adopts a pure vision solution, and the intelligent driving SOC chip is independently developed to achieve high vertical integration. Due to the uniform configuration of intelligent driving hardware on vehicles, the global leader in vehicle ownership with high-level intelligent driving capabilities (incorporating HW3.0 or HW4.0) further reduces the intelligent driving BOM cost. Under the influence of Tesla, Inc., the firm believes that there is a stronger demand for reducing the cost of intelligent driving hardware in local high-level intelligent driving solutions. High-quality training data will become one of the protective barriers for intelligent driving, and integrated intelligent car solution providers and some leading manufacturers of high-level intelligent driving will be the biggest beneficiaries of the intelligent driving industry chain, forming a competitive landscape of concentrated top players and the strong getting stronger.