Grabette: an open system to record robot

Grabette Slam
Convert Grabette recordings into LeRobot SLAM dataset
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.
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 environnements required.
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.
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.
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 exists 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 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.
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.
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:
Grabette hand held recording device (left) and Gripette robot gripper (right)
Everything is open source
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.
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.
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.
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.
Open the Grabette dashboard in your browser. Select the episodes you want to add to the dataset, and with one click, post-processing begins.
Everything is now ready to start training!
A dataset is only interesting if it trains something. So, to show the loop end-to-end, we ship a complete example. But to be clear, this is an example of what is enables, not the product: the release is the recording system. Your Grabette data works with any method that consumes LeRobot datasets.
Our example takes 200 recorded demonstrations such as :
Dataset available on this HF Dataset
and:
Policy available on this HF Model
The data bottleneck doesn’t get solved by one lab, but by a community recording demonstrations everywhere. You can now help build the dataset: build a Grabette, record tasks you are interested in, and share them on the Hub. Every episode makes the open dataset bigger and more diverse, and every contributor makes robot learning a little less gated behind expensive hardware.
This release is only the start of the project! Grabette will keep evolving. We already have more coming, including Casquette, a head-mounted POV device to complement Grabette for egocentric capture (still a work in progress). But the most important next step isn’t only ours, it’s also yours: start recording, and let’s build the dataset together.
👉 Build a Grabette (GitHub) · Process your data (HF Space) · Contribute to the dataset
Built with ❤️ by Pollen Robotics.
Convert Grabette recordings into LeRobot SLAM dataset
@SteveNguyen @glannuzel Congrats on the release, really excited to build one of these.
One small snag while working through the BoM: the electronics repo it links to isn't publicly accessible.
Could elec_grabette
be made public? Thanks!
Hi Arita, good catch! It is now public : https://github.com/pollen-robotics/elec_grabette

Facts Only

* Grabette is an open, low-cost system for recording manipulation data.
* The system involves a handheld gripper with two cameras: a wide fisheye camera and an RGBD camera.
* It records data from the observation camera, tracking camera (color, depth, IMU), and the gripper’s encoder joint values simultaneously using a shared clock.
* Grabette is paired with Gripette, a robotic arm end-effector twin.
* The hardware components are open source: Raspberry Pi, standard cameras, OAK-D depth camera, and magnetic encoders.
* Demonstrations are stored as camera-local 6-DoF Cartesian poses plus gripper state.
* The output is a standard LeRobot dataset shared on the Hugging Face Hub.
* The release includes an example of 200 recorded demonstrations and corresponding policies available on the Hugging Face Hub.
* An electronics repository for Grabette was made public.

Executive Summary

A new open, low-cost system named Grabette is released for recording robot manipulation data, aiming to address the supply problem of large, diverse real-world manipulation datasets for training robot learning policies. The system allows a human user to record a task using their hand, capturing demonstrations that are subsequently converted into a robot-ready dataset. Grabette functions as a handheld gripper instrumented with two cameras: a wide fisheye camera providing context and an RGBD camera for robust 6-DoF tracking. It is paired with Gripette, a robotic arm end-effector twin, sharing the same open-source hardware DNA. The system is designed to be robot-agnostic, storing demonstrations as camera-local poses and gripper states, which are formatted into standard LeRobot datasets on the Hugging Face Hub. The project aims to seed a large, collaborative manipulation dataset by enabling anyone to contribute data without requiring specialized robotics labs or teleoperation rigs.

Full Take

The narrative frames data collection as a democratizing force, shifting the bottleneck from expensive, specialized lab work to widespread, accessible human contribution. The core tension lies between the high demand for large datasets in advanced machine learning (vision-based policies) and the practical barriers of physical interaction and setup traditionally associated with robotics research. Grabette seeks to resolve this by proposing a "low-cost, no-robot" data acquisition method inspired by UMI. The pattern involves leveraging open hardware standards and community contribution—specifically tapping into the collaborative nature of the Hugging Face ecosystem—to achieve scale where centralized, proprietary efforts have failed. A significant implication is decentralizing the creation of robot learning artifacts, potentially reducing reliance on expensive proprietary infrastructure for foundational data. The skepticism lies in whether the ease of capture truly translates to high-fidelity, perfectly labeled demonstrations necessary for complex policy training when compared to highly controlled laboratory setups.
What are the mechanisms by which user-collected, field-based data inherently mitigates the noise or supervision issues often inherent in lab-controlled experiments? How does the community-driven approach ensure the diversity and quality required for robust representation across varied environments? What governance structures are necessary to maintain the integrity of a massive, open dataset seeded through decentralized contribution?

Sentinel — Human

Confidence

This text reads like a detailed, enthusiastic technical announcement from a research or engineering team aiming to open-source a specific hardware/software solution, exhibiting high internal consistency typical of expert communication.

Signals Detected

Moderate sentence length variance; conversational but technically dense.

Strong, focused argument with clear calls to action; exhibits genuine enthusiasm for the technical goal.

Uses specific terminology (SLAM, VLA, LeRobot, Hugging Face) coherently; attribution is direct.

Self-correction and addressing community feedback suggest an authentic development process.

Human Indicators

The flow transitions smoothly from identifying a problem (data scarcity) to proposing a specific, actionable solution (Grabette), then detailing the architecture, inspiration (UMI), and community goal. The tone is promotional yet grounded in technical feasibility.

The inclusion of a direct request for collaboration ('start recording, and let’s build the dataset together') feels characteristic of a project leader engaging the community.

The postscript acknowledging an issue with external links and immediately correcting it via a direct reply strongly suggests human interaction within a development context.