On October 22, 2026, the World Physical AI Conference & Unitree Partner Ecosystem Conference took place in Hangzhou. At this significant event, Wang Xingxing, Chairman of Unitree Technology, unveiled the company’s three primary strategies for future growth: scaling up mass production capabilities for robots to meet market demands projected to reach tens or even hundreds of millions of units; pursuing a dual strategy of in-house research and external collaboration to develop customized physical AI small models specifically designed for the company’s hardware; and deploying ultra-large-scale computing clusters to strengthen the technological foundation for continuous innovation. Simultaneously, Unitree Technology is in the process of constructing a self-evolving closed-loop system for robotics. This system aims to enhance the efficiency of algorithm development through a structured process: AI-driven self-research → simulation-based training → real-world testing → feedback-based iteration. Furthermore, Wang Xingxing announced the launch of the World Physical AI Competition, featuring an annual prize pool worth tens of millions of yuan and encompassing eight distinct challenge tracks. The competition’s overarching goal is to unite global talent and resources to collectively push the boundaries of physical AI technology.