Real2Sim2Real Closed-Loop automates the entire process, minimizing manual intervention and bridging the gap between simulation and reality. The same network can be applied to various robots and scenarios, ensuring versatility and adaptability. It efficiently produces and collects valid data to train high-performance policies. As the first biped conquers the wild using reinforcement learning, P1 delivers stable gait walking in complex environments. With high-performing joint modules, high-frequency real-time communications, and simple software development interfaces, P1 provides a reliable, efficient, and user-friendly test platform for developing bipedal locomotion control algorithms.
For more information, please check the reference URL above!
Welcome!