LLM agents replace teleoperation to generate robot data
What happened
PyRUA-DataFarm proposes replacing teleoperators with coding agents: the agent receives a data request and a single calibration clip of the robot in the target scene, reconstructs the task in simulation, and a video model then renders it into realistic, usable data. The authors report that on five real robot tasks, policies trained solely on this data succeeded 93/100 times, versus 84/100 for teleoperation data; under unseen tablecloths and colored lighting, the former held at 87, while the latter dropped to 46. According to the authors, task adaptation is fully automatic, costing an average one-time API fee of about $21 per task plus a $5 rendering contract; thereafter each data item averages $0.80 for simulation, rendering and review, with no robot or operator required.
AI 根据报道生成 · 6 days ago更新
Reporting timeline
Follow the reporting to explore different perspectives.
- HuggingFace Daily PapersFeaturedPyRUA-DataFarm: Using LLM agents instead of teleoperation to generate real-ready robot data
PyRUA-DataFarm proposes having a coding agent rebuild tasks in simulation, with a video model rendering them into real-ready data, replacing teleoperation for collecting robot demonstrations. On five real-robot tasks, policies trained only on this data achieved 93/100 successes, versus 84/100 for policies trained on teleoperated data; under an unseen tablecloth and colored lighting, the former kept 87 successes while the latter fell to 46.
Activity for this event
- Current
- 3
- Peak
- 4Nov 5 12:00
- Past 24 hours
- –
The trend compares participants observed throughout the same period, so its range may be smaller than the current interest count. Move or tap to inspect each hour; use the arrow keys on a keyboard.