跳到主要内容Skip to main content

遥操作Teleoperation

Teleoperation

  • 图像语音视频Image, speech & video
  • 第二站 · 上手Station 2 · Hands On
  • 引用核验 2026-11-07refs checked 2026-11-07
解决什么问题What problem it solves

让人在远处实时控制机器人,也顺手把人的操作录成训练数据。Let a person control a remote robot in real time, and record the operation as training data.

这是什么What it is

遥操作指人通过操纵杆、主从臂或视频反馈,实时控制远处的机器人完成动作。

在机器人学习里它还是数据采集手段:ALOHA 与 Mobile ALOHA 用低成本主从臂让人演示双臂与全身操作,把轨迹录下来供模仿学习使用。

看机器人演示要先分清:动作是人实时遥操作出来的,还是模型自主执行的——两者展示的能力完全不同。

Teleoperation means a person controls a distant robot in real time through a joystick, leader-follower arms or video feedback.

In robot learning it is also a data-collection method: ALOHA and Mobile ALOHA use low-cost leader-follower arms for human demonstrations of bimanual and whole-body operation, recording trajectories for imitation learning.

When watching robot demos, first distinguish whether actions are produced by real-time human teleoperation or autonomous model execution — the capabilities shown are completely different.

为什么需要知道Why it matters

很多惊艳的机器人视频背后是人在遥操作。知道这一点,你就不会把「人控演示」误读成「自主能力」;也能理解遥操作数据集为什么是具身智能训练的原料。Many impressive robot videos have a human teleoperating behind them. Knowing this prevents reading human-controlled demos as autonomous capability, and explains why teleoperation datasets are raw material for embodied AI training.

出处Sources
  • Zhao 等《Learning Fine-Grained Bimanual Manipulation with Low-Cost Hardware》(ALOHA)Zhao et al., ALOHA: low-cost bimanual teleoperation ↗
  • Fu 等《Mobile ALOHA: Learning Bimanual Mobile Manipulation with Low-Cost Whole-Body Teleoperation》Fu et al., Mobile ALOHA: whole-body teleoperation ↗