Apple Unveils HALO Humanoid Robot Research Using Unitree G1 Body with Mac mini M4 Pro
苹果罕见公布人形机器人进展，宇树 G1 背着一台跑千问的 Mac mini
Apple's research team has published the HALO project, a humanoid robot system that combines a Unitree G1 body with a Mac mini M4 Pro running Alibaba's Qwen models. Trained entirely on synthetic data generated via simulation rather than human teleoperation, the system achieved an average success rate of 91% across 360 trials in six unseen real-world homes without prior scanning. The robot autonomously performs tasks such as sitting, navigating under tables, and pushing boxes.
Source: Chinese robotics — Research news discovery · 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 S6
Reported numbers
elapsed_hours
300 hours
elapsed_hours
9 hours
View original evidence
论文里还提到，仅生成 SIT（坐下）一个技能的 5 万条样本，就用了 250 张 H100 GPU 跑了 9 个小时；全部 1178 个房间的生成用了约 300 张 GPU、26 个小时。
Open source S22elapsed_hours
26 hours
View original evidence
论文里还提到，仅生成 SIT（坐下）一个技能的 5 万条样本，就用了 250 张 H100 GPU 跑了 9 个小时；全部 1178 个房间的生成用了约 300 张 GPU、26 个小时。
Open source S22Reported success rate
91%
Reported trials
Reported success rate
43%
Reported trials
- 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.
这套看起来有点混搭的系统，已经能够让机器人自主走进陌生家庭，找到椅子坐下、钻过桌底、推走箱子。
Open source S6
研究团队用几百张 GPU，在上千个虚拟家庭中批量生成机器人的动作和视觉数据，再把学到的技能直接搬进现实世界。
Open source S9
团队在约 1000 个程序生成的房间里做出了 16.3 万条、近 300 小时 的配对样本，数据规模只受算力限制。
Open source S21
论文里还提到，仅生成 SIT（坐下）一个技能的 5 万条样本，就用了 250 张 H100 GPU 跑了 9 个小时；全部 1178 个房间的生成用了约 300 张 GPU、26 个小时。
Open source S22
在 6 个从未见过、没有提前扫描的真实环境里，360 次测试的平均成功率达到 91%。
Open source S7
在 100 个没见过的模拟房间里，坐椅子的成功率从只用 1 个房间训练时的 43%，涨到用 100 个房间时的 81%。
Open source S24
Source：Chinese robotics — Research news discovery · ifanr.com