Tesla's FSD Hits Roadblocks in China as Vision-Only Approach Faces Regulatory and Commercial Hurdles

Tesla's Full Self-Driving (FSD) technology has entered a critical phase of its rollout in China, drawing significant industry attention. Domestic new energy vehicle manufacturers hold mixed views on FSD—some hail it as groundbreaking, while others argue that local companies are not lagging behind in related technologies. Although FSD has been in development for nearly a decade and benefits from data collected by millions of vehicles equipped with the system, its progress in the robotaxi sector remains limited. Tesla’s current robotaxi test fleet comprises only 30 to 50 vehicles, and its testing scope is even shrinking. In contrast, Waymo has already launched commercial operations with a fleet of approximately 4,000 vehicles in the United States. A key reason for Tesla’s slow robotaxi progress is its steadfast reliance on a pure vision-based perception approach. This strategy fails to meet regulatory requirements in certain regions; for instance, Tesla has not yet received permission to test robotaxis in California, whereas Waymo has already been authorized to operate large-scale commercial services there. FSD’s business model hinges on software subscriptions and revenue from robotaxi ride-hailing services, but the latter’s sluggish advancement constrains its monetization potential. Meanwhile, jurisdictions worldwide are progressively introducing regulations that restrict the use of pure vision systems for advanced autonomous driving. As LiDAR costs decline and multi-sensor fusion technologies mature, Tesla may reconsider its technical approach. Industry observers anticipate that after mass production of the AI5 chip begins, Tesla could gradually move away from its pure vision strategy—though such a transition would require substantial time to retrain and accumulate new data.