Training data without teleoperation: our Agentic Data Flywheel

Teaching a robot a new task takes demonstrations, often hundreds of them. The usual way is for people to guide the robot through the task step by step by remote control. That is slow, expensive and hard to scale.

So we asked ourselves: what if the robot wrote its own lessons?

A pipeline that sets its own tasks

With the Agentic Data Flywheel we have built our own pipeline in which a team of AI agents does all the work that people would otherwise do:

The inventor comes up with new tasks: take a cup off the table, put a towel on the shelf, open a drawer.
The stage builder creates hundreds of virtual worlds for each task. Sometimes the cup is on the left, sometimes on the right; sometimes it is bright, sometimes dim; sometimes other objects are in the way.
The teacher writes the strategy the robot uses to solve the task as a program.
The examiner judges every attempt impartially. Only what truly succeeded becomes training data.

What makes it special: the pipeline learns

This is where the magic starts. When an attempt fails, the failure is not simply thrown away. The teacher works out why the grasp missed, rewrites its own strategy and tries again.

With every round, the generator becomes not just more productive but smarter. Tasks it failed at yesterday, it solves reliably today, and every solved task produces data for the next one. That is the flywheel: the longer it runs, the faster it spins.

When our team goes home in the evening, the pipeline keeps working. The next morning, new verified demonstrations are waiting, without anyone touching a joystick.

Independent of hardware

We developed the flywheel for a mobile manipulator. Right now we are transferring it to a two-armed robot without rebuilding the pipeline. That shows the flywheel is not tied to one robot. It simply learns what each body can do.

Why this matters for our robot

Every new task in a hotel or in care needs training data. If it is generated automatically, weeks of manual work turn into a process that runs overnight. That is how we want to bring new skills to market quickly and affordably.