Grabette: an open system to record robot-manipulation data

Grabette：一个用于记录机器人操作数据的开源系统

Record your own manipulation tasks in minutes with a handheld gripper, turn them into robot-ready datasets automatically, and help grow an open, collaborative dataset for robot learning. 只需使用手持式夹爪，即可在几分钟内记录下你自己的操作任务，自动将其转化为机器人可用的数据集，并助力构建一个用于机器人学习的开放式协作数据集。

The bottleneck isn’t the model. It’s the data. Robot learning has a supply problem. We have capable policy architectures (transformer-based VLAs, diffusion and flow-matching policies, and even world models) and the GPUs to train them. What we lack is large, diverse, real-world manipulation data. 瓶颈不在于模型，而在于数据。机器人学习正面临供应问题。我们已经拥有了强大的策略架构（基于 Transformer 的 VLA、扩散和流匹配策略，甚至是世界模型）以及用于训练它们的 GPU。我们所缺乏的是大规模、多样化的真实世界操作数据。

Teleoperating a robot to collect it can be expensive and demanding: first of all, it requires a robot. And depending on the teleoperation method, data collection can be tedious for the user if it takes hours and involve significant hardware and logistical challenges. That is difficult to scale with the wide variety of tasks and environments required. 通过远程操作机器人来收集数据既昂贵又费力：首先，它需要一台机器人。此外，根据远程操作方法的不同，如果收集过程耗时数小时，且涉及重大的硬件和物流挑战，对用户来说可能会非常繁琐。面对所需的各种任务和环境，这种方式很难扩展。

But you don’t need a robot to collect robot data. Just a human hand, a gripper, a camera, and a way to recover the 6-DoF trajectory of what the hand did. Capture the demonstration and you have data a robot can learn from. That’s what we’re releasing today: Grabette, an open, low-cost system for recording manipulation data. Pick it up, record a task with your own hand, and get back a clean, robot-ready dataset. No robot, no lab, no teleop rig. 但你并不需要机器人来收集机器人数据。只需要一只人手、一个夹爪、一台摄像机，以及一种恢复手部 6 自由度（6-DoF）轨迹的方法。捕捉演示过程，你就拥有了机器人可以学习的数据。这就是我们今天发布的内容：Grabette，一个用于记录操作数据的低成本开源系统。拿起它，用你的手记录一个任务，然后就能得到一个干净、机器人可用的数据集。无需机器人，无需实验室，也无需远程操作装置。

And that’s the bigger goal: if recording a demonstration is as easy as shooting a video, anyone can contribute. We want Grabette to seed a large, open, collaborative manipulation dataset. One no single lab could ever build alone. 这就是我们更大的目标：如果记录演示像拍视频一样简单，任何人都可以做出贡献。我们希望 Grabette 能成为一个大型、开放、协作式操作数据集的种子，这是任何单一实验室都无法独自完成的。

Standing on the shoulders of UMI

站在 UMI 的肩膀上

Grabette is directly inspired by the Universal Manipulation Interface (UMI) from Stanford: a handheld gripper with a fisheye camera that records demonstrations “in the wild”, recovers camera trajectories with SLAM, and trains visuomotor policies from them. UMI proved the recipe works. Other (closed source) devices exist like Agibot’s MEgo gripper, Genrobot’s DAS gripper and Sunday Robotics skill capture glove. Our goal was to make it effortless to use, to get the barrier from “I have a task” to “I have a trained model” as low as possible. Grabette 直接受到斯坦福大学通用操作接口（UMI）的启发：这是一种带有鱼眼摄像机的手持式夹爪，可以在“野外”记录演示，通过 SLAM 恢复摄像机轨迹，并据此训练视觉运动策略。UMI 证明了这个方案是可行的。目前还存在其他（闭源）设备，如 Agibot 的 MEgo 夹爪、Genrobot 的 DAS 夹爪以及 Sunday Robotics 的技能捕捉手套。我们的目标是使其使用起来毫不费力，尽可能降低从“我有一个任务”到“我有一个训练好的模型”之间的门槛。

Grabette is built into the modern open ecosystem: LeRobot for datasets, the Hugging Face Hub for sharing, and a processing pipeline you run from your browser with nothing to install. Grabette is something anyone can build on a workbench, use in the field, and contribute data from. Grabette 构建于现代开放生态系统之上：使用 LeRobot 处理数据集，使用 Hugging Face Hub 进行共享，并提供了一个无需安装、直接在浏览器中运行的处理流水线。Grabette 是任何人都能在工作台上组装、在现场使用并贡献数据的工具。

Meet Grabette

认识 Grabette

We have been developing Grabette for months, and we feel it has become usable enough to share. We are excited to share it now with you! Grabette is a handheld gripper instrumented with everything needed to reconstruct a manipulation demonstration. It carries two cameras, each with a distinct job. Splitting the two roles is deliberate: the cheap wide fisheye gives the policy the context-rich, wrist-camera-style view it needs, while the RGBD camera does the heavy lifting of robust 6-DoF tracking. 我们开发 Grabette 已经数月，感觉它已经足够成熟，可以分享了。我们很高兴现在能与大家分享！Grabette 是一个手持式夹爪，配备了重建操作演示所需的一切。它携带两台摄像机，每台都有不同的任务。这种分工是刻意为之的：廉价的广角鱼眼镜头为策略提供了所需的上下文丰富的腕部摄像机视角，而 RGBD 摄像机则负责繁重的稳健 6 自由度追踪工作。

And while Grabette records data during tasks performed by a user, it relies on its robotic counterpart to execute the movements it has learned after training. So, meet Gripette, the robotic arm end-effector twin of Grabette. The family shares the same hardware DNA: 虽然 Grabette 在用户执行任务时记录数据，但它依赖其机器人对应设备在训练后执行所学的动作。因此，请认识一下 Gripette，它是 Grabette 的机械臂末端执行器孪生兄弟。这一系列设备共享相同的硬件基因：

Grabette, the handheld demonstration device (camera + IMU + gripper, BOM cost ~490€)
Grabette，手持式演示设备（摄像机 + IMU + 夹爪，物料清单成本约 490 欧元）
Gripette, the motorized gripper (camera + two servomotors, BOM cost ~120€) that closes the loop on a real or simulated robot arm
Gripette，电动夹爪（摄像机 + 两个伺服电机，物料清单成本约 120 欧元），用于在真实或模拟机械臂上实现闭环控制

Built for everyone

为每个人而生

Everything is open source. Go check the repository! 一切皆为开源。快去查看代码仓库吧！

Hardware: CAD and production files for Grabette and Gripette
硬件： Grabette 和 Gripette 的 CAD 及生产文件
Capture service: the on-device Raspberry Pi software
捕捉服务： 设备端的树莓派（Raspberry Pi）软件
Processing pipeline: run locally, or online via our Hugging Face Space
处理流水线： 可在本地运行，或通过我们的 Hugging Face Space 在线运行
Example downstream stack: stock LeRobot training + the OpenArm evaluation, as a reference
下游示例栈： 标准 LeRobot 训练 + OpenArm 评估，作为参考

Components: Standard sensors you can buy, no closed pipeline, no fork lock-in. A Raspberry Pi, a standard Pi camera, an off-the-shelf OAK-D depth camera, magnetic encoders. The whole point is that anyone can build one from parts you can just order. 组件：你可以买到的标准传感器，没有封闭的流水线，没有分支锁定。一个树莓派、一个标准树莓派摄像头、一个现成的 OAK-D 深度摄像头、磁编码器。其核心意义在于，任何人都可以用订购到的零件组装一个。

Robot-agnostic by design. Nothing in the capture or the data format assumes a particular arm. Demonstrations are stored as camera-local 6-DoF cartesian pose plus gripper state, the output is a standard LeRobot dataset on the Hugging Face Hub, so the same data can drive different robots and different learning methods. You will still need the matching Gripette gripper on your arm though. 设计上与机器人无关。捕捉过程或数据格式中没有任何部分预设了特定的机械臂。演示数据存储为摄像机坐标系下的 6 自由度笛卡尔位姿加上夹爪状态，输出为 Hugging Face Hub 上的标准 LeRobot 数据集，因此相同的数据可以驱动不同的机器人和不同的学习方法。不过，你仍然需要在你的机械臂上安装配套的 Gripette 夹爪。

From your hand to a dataset, in two steps

从你的手到数据集，只需两步

This release enables anyone to go from “I want to demonstrate a task” to “I have a training-ready dataset” quickly and without prior expertise. 此版本使任何人都能快速且无需专业知识，从“我想演示一个任务”过渡到“我拥有一个可用于训练的数据集”。

Record: Press the button, and data from the observation camera, the tracking camera (color, depth, and IMU), and the gripper’s encoder joint values are recorded simultaneously, using a single shared clock to ensure proper synchronization. Press the button again to stop the episode, and the data is saved locally on the Raspberry Pi. 记录： 按下按钮，观测摄像机、追踪摄像机（彩色、深度和 IMU）以及夹爪编码器关节值的数据将同时被记录，并使用单一共享时钟以确保正确的同步。再次按下按钮停止片段，数据将保存在树莓派本地。
Process, directly in your browser: Open the Grabette dashboard in your browser. Select the episodes you want to add to the dataset, and with one click, post-processing begins. The episodes are uploaded to the HF Hub. The grabette-slam space performs SLAM using RTAB-MAP library and verifies that the trajectory is correct (with no jumps or loss of tracking). The episodes are converted to LeRobot format. A new dataset is uploaded to your space, where you can view the data for each episode using the LeRobot visualizer. Everything is now ready to start training! 处理（直接在浏览器中）： 在浏览器中打开 Grabette 仪表板。选择你想要添加到数据集中的片段，点击一下，后处理就开始了。片段会被上传到 HF Hub。grabette-slam 空间使用 RTAB-MAP 库执行 SLAM，并验证轨迹是否正确（没有跳变或追踪丢失）。片段被转换为 LeRobot 格式。一个新的数据集会被上传到你的空间，你可以在那里使用 LeRobot 可视化工具查看每个片段的数据。现在一切准备就绪，可以开始训练了！