Apple Discloses HALO Autonomous Control System for Humanoid Robots with 91% Average Single-Task Success Rate
苹果披露人形机器人自主控制系统HALO 单项任务测试平均成功率91%丨盘前线索
On November 4, Apple's research team submitted a paper titled 'Perceptive Humanoid Motion from Simulation' to arXiv, disclosing the HALO autonomous control system for humanoid robots. The system covers action generation, tracking, task orchestration, and execution in both simulation and real robots. Tests showed a 91% average success rate on single tasks across six unscanned environments and a 70% average success rate on combined tasks in one environment. The team noted reliance on visual perception with tactile sensing as a future improvement direction.
Source: Chinese robotics — Hardware and sensing · Read original article ↗
English translation unavailable. Choose the original article to read it.
What the source reports
Publisher-reported claims, with original evidence. These results have not been independently verified by RoboSignal.
- Control
- Reported autonomousOpen source S2
Reported numbers
photo captures
360
Reported trials
View original evidence
据论文披露，研究团队在6个未预先扫描或建图的真实环境中开展360次单项任务测试，平均成功率为91%；另在一个真实环境中，对3种组合任务分别开展10次测试，平均成功率为70%。
Open source S3photo captures
10
Reported trials
View original evidence
据论文披露，研究团队在6个未预先扫描或建图的真实环境中开展360次单项任务测试，平均成功率为91%；另在一个真实环境中，对3种组合任务分别开展10次测试，平均成功率为70%。
Open source S3Reported success rate
91%
View original evidence
据论文披露，研究团队在6个未预先扫描或建图的真实环境中开展360次单项任务测试，平均成功率为91%；另在一个真实环境中，对3种组合任务分别开展10次测试，平均成功率为70%。
Open source S3Reported success rate
70%
View original evidence
据论文披露，研究团队在6个未预先扫描或建图的真实环境中开展360次单项任务测试，平均成功率为91%；另在一个真实环境中，对3种组合任务分别开展10次测试，平均成功率为70%。
Open source S3
- code: Reported pending Open source S2
- paper: not_reported
What remains unknown
Not established in the collected evidence: Environment, Data origin.
Reported performance applies to the described task. It does not establish general autonomy or deployment readiness.
Source excerpts and review record
Automatically extracted; no manual editorial approval recorded.
近日，据澎湃新闻报道，当地时间11月4日，苹果研究团队在预印本平台arXiv提交论文《Perceptive Humanoid Motion from Simulation》，披露了人形机器人自主控制系统HALO，并公开了数据生成、模型训练和部署代码，涵盖动作生成、动作跟踪、任务编排以及仿真和真实机器人运行等环节。
Open source S2
据论文披露，研究团队在6个未预先扫描或建图的真实环境中开展360次单项任务测试，平均成功率为91%；另在一个真实环境中，对3种组合任务分别开展10次测试，平均成功率为70%。
Open source S3
2025年全球人形机器人出货量约1.8万台，2026年上半年国内产量已超4万台。
Open source S6
进入2026年，人形机器人产业化进程显著提速，工信部数据显示2026年上半年国内产量已超4万台，全年整机产量有望突破10万台。
Open source S11
Source：Chinese robotics — Hardware and sensing · sfccn.com