On July 17, 2026, at the prestigious World Artificial Intelligence Conference (WAIC), WeRide proudly introduced its revolutionary fundamental physical AI cognition model, WIIT. This innovative model introduces the pioneering concept of the 'minimal unit of physical facts.' It ingeniously dissects the ever-evolving real-world scenarios into discrete, identifiable, and verifiable factual units. In doing so, it lays a solid foundation for an AI understanding framework tailored to the physical world.

WIIT boasts four essential core capabilities: fact extraction, which allows it to sift through vast amounts of information and pinpoint relevant facts; fact reasoning, enabling it to draw logical conclusions based on the extracted facts; fact verification, ensuring the accuracy and reliability of the information by cross-checking and validating it; and fact orchestration, which coordinates and integrates these facts to form a comprehensive understanding.

Its primary objective is to significantly enhance AI's capacity for environmental understanding and decision-making in embodied intelligence scenarios. This includes areas such as autonomous driving, where split-second decisions are crucial, and robotics, where precise interaction with the physical environment is essential. By doing so, WIIT is driving a paradigm shift in artificial intelligence, propelling it from mere data comprehension to in-depth real-world cognition.