Shanghai (Gasgoo)- On July 23, Chinese autonomous driving company Pony.ai announced that public road testing of the ARCFOX αT5 powered by its Gen-7 autonomous driving system (namely, Pony.ai's seventh-generation autonomous vehicle model) has officially begun in Beijing.
This milestone comes just half a month after the same model was approved for testing in Shenzhen, following earlier authorization in Guangzhou.
The simultaneous greenlight in China's three leading megacities marks a strong start for Pony.ai's plan to deploy a Robotaxi fleet of 1,000 units by the end of this year.
The rollout of multiple seventh-generation models across Beijing, Guangzhou, and Shenzhen underscores the robustness and versatility of Pony.ai's L4 autonomous driving solution. The vehicles are being put to the test in diverse urban environments that vary widely in traffic behavior, road layouts, and climate conditions—highlighting the system's adaptability, platform stability, and scalability.
In Beijing alone, the seventh-generation autonomous vehicle fleet is already conducting routine operations within a 225-square-kilometer autonomous car demonstration zone. These efforts are expected to enrich mobility options for local residents while contributing to the city's broader push for smarter, more efficient urban transportation systems.
The licensed ARCFOX αT5 vehicles are co-developed by Pony.ai and BAIC BJEV, and first debuted at the AUTO SHANGHAI 2025 this April. Built natively on BAIC BJEV's BEV (battery electric vehicle)-dedicated platform, the model supports full-stack redundancy required for L4 autonomy. It is equipped with Pony.ai's Gen-7 integrated hardware-software system.
The vehicle houses 34 sensors across six categories, including LiDAR, cameras, millimeter-wave radar, microphones, water sensors, and collision detectors. This comprehensive sensor suite enables full-scenario, all-weather, and round-the-clock autonomous driving capabilities that span urban roads, highways, and ring roads.
At the heart of the technology is Pony.ai's "world model + virtual driver" framework, which allows the system to safely navigate variable lighting, complex road conditions, and rare long-tail edge cases. Notably, with a 70% reduction in autonomous driving hardware costs compared to earlier generations, the new vehicle delivers smoother, safer performance in challenging scenarios—such as narrow street stops, emergency vehicle yielding, sudden pedestrian intrusions, low-light environments, and toll booth passages.
