XPeng Inc. has officially refuted recent online claims regarding the development progress of driver assistance systems for certain vehicle models, labeling the information as inaccurate. The company’s official Weibo account, @XPeng辟谣, issued a statement today clarifying that the second-generation VLA (Vision-Language-Action) distillation integration work is proceeding on schedule as planned.
According to the statement, the rumors circulating online about the development progress of driver assistance for specific models are untrue. The company emphasized that the second-generation VLA distillation integration work is advancing in an orderly manner, and any future updates will be promptly communicated through official channels.
Earlier reports from IT Home noted that XPeng Inc. officially unveiled its second-generation VLA smart driving technology on March 2. This new system boasts a 30% improvement in traffic efficiency compared to traditional Level 2 driver assistance and Robotaxi solutions. It also includes advanced features such as recognizing traffic police hand gestures and automatically slowing down during drunk-driving checkpoints.
XPeng Inc. Chairman and CEO He Xiaopeng stated that end-to-end small models have hit their capability ceiling, marking a "watershed moment" in smart driving research and development. The company’s official position is that autonomous driving fundamentally revolves around solving physical AI challenges. Understanding the operational rules of the real physical world serves as the starting point for XPeng Inc.'s Turing second-generation VLA model.
The second-generation VLA model has already achieved a unified system that operates across both China and Europe. XPeng Inc. is currently accelerating mass production adaptation for the European market, with a target to gradually begin delivering to overseas customers by 2027.
Comments