Skip to content
Chinese robotics — Research news discovery·· 4 days agoSignalEditorial score75

Apple Research Team Puts Mac mini on Unitree G1 Robot for HALO System

Apple 把 Mac mini 装进了机器人,也得感谢英伟达

Summary

Apple's research team has developed a system called HALO that mounts an M4 Pro Mac mini on the Unitree G1 humanoid robot to handle perception, instruction understanding, and action planning. The system uses the Mac mini for high-level task scheduling and motion generation while an NVIDIA Jetson Orin NX handles low-latency joint control. In real-world tests, the robot achieved an 85% success rate in sitting on chairs. Training data was largely generated in simulation, producing approximately 300 hours of robot-view footage across 1,000 indoor environments.

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.

Reported numbers

  • Reported success rate

    85%

    Reported trials

    View original evidence
    苹果搞的这一套,坐椅子的真实测试成功率是 85%,这就说明:人类觉得最稀松平常的动作,机器依然需要反复练习,还不一定呢做好。
    Open source S21
  • times per second

    50

    View original evidence
    相对来说,还是有点慢慢腾腾 😂 下面的运动跟踪器就快了,运行在 G1 自带的英伟达 Jetson Orin NX 上,以每秒 50 次的频率,结合深度图像和机器人自身状态,输出关节控制目标,负责行动这一块。
    Open source S14
  • paper: not_reported
  • code: not_reported
  • weights: not_reported
  • dataset: not_reported

What remains unknown

Not established in the collected evidence: Environment, Control, 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.

苹果搞的这一套,坐椅子的真实测试成功率是 85%,这就说明:人类觉得最稀松平常的动作,机器依然需要反复练习,还不一定呢做好。

Open source S21

相对来说,还是有点慢慢腾腾 😂 下面的运动跟踪器就快了,运行在 G1 自带的英伟达 Jetson Orin NX 上,以每秒 50 次的频率,结合深度图像和机器人自身状态,输出关节控制目标,负责行动这一块。

Open source S14

背着它的是宇树 G1 人形机器人。

Open source S7

Apple 把 Mac mini 装进了机器人,也得感谢英伟达

Open source S1

前两天写 5 万块的 MacBook Pro 时大家还在聊,Mac 到底适不适合跑本地大模型。

Open source S2

有人说,视频模型得用英伟达显卡,Mac 再贵也没用啊。

Open source S3

到底用英伟达还是用 Mac 的 GPU,这个得根据场景具体分析。

Open source S4

不过我今天看了个视频,苹果研究团队最近给 Mac 找了个新岗位:背在机器人儿身上,干嘛呢?

Open source S5

Source:Chinese robotics — Research news discovery · sohu.com

Timezone · UTC

Article dates follow your selected timezone. Briefing editions use Hong Kong time (UTC+8).