Tesla, a pioneer in the electric vehicle sector, draws global attention with its FSD system. On February 25, Tesla officially launched FSD version 13.2.6 for Chinese users. This delayed debut did not crash, but it failed to replicate its North American success due to localization issues. Owners must spend 64,000 yuan (8790$) to purchase FSD. They must also have a new car equipped with hardware 4.0 after October 2023. Many early Tesla owners find themselves excluded. While FSD stands for “Full Self-Driving,” the Chinese version is renamed “FSD Smart Assisted Driving.” Some users found that the Chinese FSD lacks advanced features like “parking to parking.” They also experienced navigation errors, entering bus lanes incorrectly, and making illegal lane changes. Tesla’s once-proud technological edge is narrowing against Chinese competitors. Huawei’s lidar, Xpeng’s end-to-end models, and BYD’s data integration create strong barriers. Local companies have optimized their smart driving systems for domestic conditions. The race for L3 autonomous driving has shifted from technical breakthroughs to ecosystem collaboration. Tesla faces significant challenges in entering the Chinese market. Data, technology, and business models create a “triple siege.”
Tesla’s journey to launch FSD in China faces many challenges. Data security and compliance issues top the list. Tesla’s core strength comes from its “data central kitchen” model. This model relies on real-time data from 7 million vehicles to train algorithms. However, China’s regulations require local data storage. They also mandate security assessments for important data, such as geographic information, facial recognition, and license plates. This policy disrupts Tesla’s global data loop. In January, Musk explained that due to data transfer limitations, the Chinese version of FSD can only “watch videos.” It uses online videos of Chinese roads for training. Since data must be trained locally, Tesla’s FSD technology struggles in China. The “pure vision + end-to-end” approach works better in the U.S., where road conditions differ. China’s complexity arises from diverse traffic participants, intricate traffic signs, and frequent rule changes. Differences in road signs, rules, and driving habits require deep localization of the autonomous driving system. For example, China’s bus lanes, tidal lanes, and mixed electric vehicle scenarios demand higher recognition and decision-making capabilities from the FSD system. Tesla lagged in preparations for entering the Chinese market, causing multiple delays. Chinese automakers chose a different technical path. They now integrate lidar, millimeter-wave radar, and vision. This combination has become standard among Chinese companies. While it raises hardware costs, it creates a safety buffer in China’s traffic environment. Chinese firms also adopt a more localized approach to autonomous driving. They handle various traffic emergencies more effectively than Tesla. Additionally, Tesla’s FSD in China still charges a one-time fee of 64,000 yuan (8790$) or a subscription. However, its current performance does not meet user expectations. BYD offers L2-level smart driving in models priced at 70,000 yuan (9610$). Xpeng and NIO lower usage barriers with “hardware standard + software subscription.” Tesla’s HW4.0 models support advanced features, leaving 2 million HW3.0 users unable to upgrade. This limits Tesla’s user base. Research shows that Chinese consumers trust smart driving only 45%. However, they demand localized features, like “ghost probe” warnings, 37% more than in North America. This contradiction makes it hard for Tesla’s “global” feature package to meet diverse needs. Competition for Ecological Reconstruction and Standard Output in L3 Landing
China’s smart driving market has entered a stage of systematic competition. Huawei builds an ecological loop with “LiDAR + Harmony OS.” Yu Chengdong claims that the “non-LiDAR version is better than FSD.” Xpeng achieves nationwide no-map navigation through XNGP and partners with insurance companies to launch “smart driving insurance” to address liability issues. BYD leverages its annual sales of 5.5 million vehicles to create data scale effects. Its “Tian Shen Zhi Yan” system connects to regional traffic data for real-time alerts on special road conditions. China is shifting from a “data compliance regulator” to a “technology standard exporter.” The 2024 “Cross-Border Data List for Intelligent Connected Vehicles” provides a channel for compliant data flow. However, Tesla struggles to enjoy policy benefits due to geopolitical factors. Local companies accelerate technology implementation through government-enterprise collaboration. BYD connects directly with Shenzhen’s traffic management system for real-time road conditions. Huawei partners with operators to build a 5G vehicle network for efficient interaction between vehicles, traffic lights, and roadside units. In the industrial chain, local computing platforms like Horizon Journey 6 and Hei Zhi Ma A2000 have achieved breakthroughs in 7nm processing. If Tesla cannot integrate into this ecosystem, its localization costs will remain high. Beyond technical challenges, L3 autonomous driving faces complex regulatory and ethical dilemmas. Currently, global regulations for L3 autonomous driving are incomplete. Key issues include defining liability in accidents, regulating testing and road standards for autonomous systems, and ensuring user data security and privacy. For example, when an autonomous vehicle has an accident, it is unclear whether the manufacturer, software developer, or user bears responsibility. This uncertainty complicates judicial practice and increases the risks and uncertainties for companies promoting L3 technology. Tesla’s difficulties in entering China reflect the conflict between global technology and local ecosystems. Hard data constraints, soft technological opposition, and gaps in business models contribute to its struggles. To break through, Tesla may need to accept the reality of a “China-specific FSD” or even restructure its technology. Meanwhile, Chinese automakers aim to leverage their localization capabilities to gain a voice in global smart driving.