OPENARM: A RESEARCH-GRADE FORCE-FEEDBACK ROBOT ARM
An open-hardware arm built for physical AI — backdrivable QDD actuators and true bilateral teleoperation, the kind of rig imitation-learning labs actually use.
by Enactic
RoboticsAI
- difficulty
- ●●●●●
- time
- a serious project
- license
- CERN-OHL-S-2.0
- repo
- repo ACTIVE700 stars
●●●●● · a serious project · CERN-OHL-S-2.0 · 700 stars · repo ACTIVE
WHAT YOU’LL NEED
- 3D printer + filament — printable parts — files are in the repo
- Electronic parts — full list with part numbers in the repo BOM
- Dev board / microcontroller — runs the project firmware
- Soldering kit + basic tools — one-time buy, serves every future build
Partner
COMPAREE VERDICT
This is the serious one on our list — the arm you build when you've outgrown the cheap printed ones. Enactic open-sourced a research-grade platform for physical AI: backdrivable quasi-direct-drive actuators, true bilateral (leader-follower) teleoperation with force feedback, and a full stack — CAD, URDF, MuJoCo and Isaac Lab simulation, a CAN control library, ROS2, and teleop — spread across nine repos. The point is imitation learning where the robot can feel contact, not just replay positions. Two things to respect before you start: it's a big-ticket, expert-level build (band $$$), and the hardware is licensed CERN-OHL-S-2.0, not Apache — so read the license if you plan derivatives. For a lab, a serious hobbyist, or a research-minded team, it's outstanding.
IN THE REPO
NOT IN THE REPO
- —CAD (STL/STEP/Fusion) plus URDF, MuJoCo and Isaac Lab sim, CAN control, ROS2, teleop across 9 repos
- —Hardware license is CERN-OHL-S-2.0 (not Apache) — software repos are Apache-2.0; matters for derivatives
- —BOM is on Google Drive alongside the CAD
- —This is research-grade, not a weekend build
Can I build this?
Build at your own risk. Projects involve tools, electronics and sometimes mains voltage — follow the creator’s safety notes.
Partner · KickstarterHeyGears G1: 10M+ colors and transparent parts in one print, plus UV printing on flat objects. Figures, parts and labels — no painting.
Videos
OpenArm 01 | Open-source humanoid arm for physical AI1:22
Short official intro from Enactic — the docs are the real reference.
Gallery
Start here
Navigation into the creator’s own docs — we don’t rewrite the guide, we route you to the source.
- 1.Read the docs (start here — the full reference)
- 2.Watch the intro (90 sec overview)
- 3.Explore the repos (hardware, description, can, ros2, teleop, sim)
- 4.Read the purchase/BOM guide (actuators + parts)
- 5.Join the Discord (the OpenArm community)
Resources
Documentation, files and community threads for this build — we link straight to the original sources and never rehost the creator’s files.
Partner · KickstarterCORE POWER BANK S: 0.2" thin, 2.8 oz, 5000 mAh with 22.5W fast charging. Semi-solid battery, carbon fiber, IPX7 waterproof.
KNOWN ISSUES
- Hardware is CERN-OHL-S-2.0, not Apache — read it before building derivatives
- It's spread across nine repos plus a Google Drive BOM — expect to piece the picture together
- DAMIAO actuators dominate the cost and the difficulty — this is expert-level
What makes it different from a cheap printed arm?
Force feedback. The backdrivable QDD actuators and bilateral teleoperation let the arm feel contact, which is what serious imitation-learning setups need — not just position replay.
Can a hobbyist build it?
An experienced one, yes — but it's expert-level and expensive. If you're starting out, build SO-ARM100 first and come back to this.
What sim tools does it support?
MuJoCo and Isaac Lab, with URDF/xacro descriptions and a CAN control library plus ROS2.
What does it cost?
Band $$$ for a bimanual system. The BOM on the creator's drive has current prices.
Community builds
No community builds yet — be the first, we feature the best ones.
Discussion1
FROM THE COMPAREE TEAM
OpenArm's whole argument is force feedback — backdrivable actuators that feel contact instead of blindly replaying positions. How much does that actually matter for imitation learning, in your experience?
SO-ARM100: The Low-Cost Robot Arm Behind HuggingFace LeRobot
●●●○○ · start here — the cheap, printable arm before this one
ALOHA: The Stanford Robot That Learns From You
●●●●○ · another bimanual teleoperation platform, built on commercial arms
Amazing Hand: A Fully 3D-Printed Robotic Hand
●●●○○ · a dexterous printed hand to pair with an arm
Enactic
A robotics company open-sourcing a research-grade arm for physical AI — force-feedback teleoperation and a full sim-to-real stack — to make state-of-the-art results reproducible.
Keep building — related projects
Same category, picked by how close they are to this build.
FULL CAD3D print this robot arm, show it a task once, and it learns to repeat it
RoboticsAI●●●●○advanced
FULL CADSO-ARM100: The Low-Cost Robot Arm Behind HuggingFace LeRobot
RoboticsAI●●●○○moderate
FULL CADXLeRobot: A Household Robot on an IKEA Cart
RoboticsAI●●●●○advanced
FULL CADYou can 3D-print your own walking Star Wars BD-X droid - fully open-source
RoboticsAI●●●●●expert
DISCLAIMER
- Comparee is not the author of the projects featured here. All rights to each project belong to its creator — every page links to the original source, and we never host creators’ files.
- Information is provided without warranty and may become outdated as projects evolve. Prices are indicative bands only — always check the creator’s parts list for current costs.
- Building and operating any project is at your own responsibility. Protective equipment, safe workshop practice and compliance with local regulations are the builder’s responsibility.
CompareeTEAM1mo ago
A licensing wrinkle we'd flag for anyone forking: the hardware is CERN-OHL-S-2.0 while the software repos are Apache-2.0, and that split has real consequences for derivatives. The project also sprawls across nine repos plus a Google Drive BOM, so assembling the full picture takes effort before assembling anything physical. Has anyone compared its bilateral teleop to a plain position-replay rig?