Repeatable Tasks
ROBO-1 is built for repeatable deployment, with the same setup, same inputs, and same behavior from station to station.


A standardized robot arm for teachable automation.
Identical hardware. Reproducible results. Share policies, not setup guides across every bench in the lab.
Unbox to data collection in an afternoon. Skip months of integration before the first real workflow runs.
Train on one station. Deploy to a hundred. No per-unit config as the line scales up.
ROBO-1 is designed to move from pilot to repeated deployment without rebuilding the whole stack every time.
ROBO-1 is built for repeatable deployment, with the same setup, same inputs, and same behavior from station to station.

Roboport combines teleoperation, fleet ops, and policy iteration so your system gets easier to run as you scale.

ROBO-1 connects demonstration, data collection, model training, and real-world execution in one integrated workflow.
Teach
Demonstrations and data collection.

is for companies who want toautomate solderingsort and pack inventoryrun quality inspectionsassemble small partstend CNC machinesload and unload stations
Everything you need to build, deploy, and scale robotic automation - hardware, software, and fleet management in one platform.
ROBO-1
A 6-DOF robot arm with gripper, built for teachable automation. Standardized hardware means every unit behaves identically — train a policy once, deploy it to any ROBO-1.
The questions teams ask when evaluating teachable automation in the real world.
Use the passive leader arm to demonstrate the motion while the follower arm mirrors it and the included cameras record the episode. Those demonstrations can then be stored in Roboport or used in LeRobot-style training and evaluation workflows.
Yes. Most teams begin with direct teleoperation, then move toward higher-level operator supervision as a skill matures. Roboport is built to support that progression while capturing the operator data you need to train the model to take over more of the job.
Roboport covers episode capture, persistent dataset storage, model training and fine-tuning, checkpoint evals, and agent-assisted data workflows. It is the layer that connects teaching, training, validation, and deployment.
You can monitor sessions, manage teleoperation, and roll out policies from one platform instead of building custom tooling for every robot. Because the hardware, cameras, and workspace geometry are standardized, scaling to more stations does not turn into a fresh integration project each time.
ROBO-1 ships with native Hugging Face LeRobot support, a matched MuJoCo digital twin, and a URDF for custom simulation workflows. It is designed to work with common imitation-learning methods like ACT and Diffusion Policy instead of locking you into a closed stack.

Sign up for preorder access and we'll follow up with launch timing, pricing, and next steps for your team.