According to CoinMeta, Apple's research team has disclosed progress in their research on robot “brains.” On November 4th local time, the Apple research team submitted a paper titled “Perceptive Humanoid Motion from Simulation” to the preprint platform arXiv, revealing an autonomous control system for humanoid robots HALO. The research team deployed this system on Yushu Technology’s G1 humanoid robot for verification, enabling the robot to perform actions learned in a simulated environment in a real-world setting. They also made the code for data generation, model training, and deployment public, covering aspects such as action generation, action tracking, task scheduling, and the operation of both simulated and real robots. As revealed in the paper, the team conducted 360 single-task tests in 6 real environments that were not pre-scanned or mapped, with an average success rate of 91%. In another real environment, they tested three different combination tasks ten times each, with an average success rate of 70%. Among these, the combination of “bending down to pass over obstacles—moving towards a target—pushing a box” had a success rate of 50%.