arXiv preprint ·
AutoRSR: Automated Real-to-Sim-to-Real Data Generation for Scalable Robot Learning
- Published
- Source
- arXiv preprint, posted before peer review
- Original
- Read it at arXiv
- Topics
Abstract
Scaling embodied robot learning requires large and diverse manipulation datasets, but collecting real-world demonstrations remains expensive in human effort and robot time. Simulation provides a scalable alternative, yet converting a real workspace into a robot-learning-ready environment still requires substantial manual effort for object extraction, asset processing, coordinate calibration, and trajectory specification. We present AutoRSR, a highly automated real-to-sim-to real framework that converts scanned real-world workspaces into reusable embodied simulation assets and scalable manipulation data. Given multi-view observations, robot states, and open-vocabulary semantic queries, AutoRSR reconstructs the workspace, extracts individual object assets, and automatically registers the reconstructed scene to the robot base frame at metric scale. Starting from a small number of reference motions, AutoRSR then generates executable manipulation trajectories and scales them through trajectory interpolation and domain randomization. After workspace scanning and semantic queries, the remaining pipeline requires no manual object annotation, scene-to-robot calibration, or robot teleoperation. Across three real-world manipulation tasks, policies trained on 1K automatically generated episodes per task achieve 73.3--86.7% real-world success rates, comparable to pi 0.5 trained from 150 human teleoperated demonstrations, while reducing the human time required for data generation by an order of magnitude. These results demonstrate that real-world observations can be converted into metrically aligned, executable, and scalable simulation data with substantially reduced human effort.
The authors’ abstract, as arXiv lists it. arXiv’s listings, abstracts included, are free for anyone to reuse.
Read it on ReadSearch
Each opens after you create an account, and brings you back to this paper.
More in Robotics
NavAble: A Large-Scale Dataset and Synthetic Data Generation Pipeline for Blind Navigation
RoPE-Flow: Monocular Vision-based Robot Pose Estimation via 2D Graph-Conditioned 3D Flow Matching
Probabilistic Physics Digital Twin for Human-Robot Interaction
Slip-Induced Fall Suppression During Cornering for a Remotely Operated Motorcycle with a Rear-Wheel-Swing Mechanism
VLX-Go: Robust Embodied Tracking through Multi-Stage Learning and On-Policy Refinement
Get the few papers worth your time in Robotics every morning.
ReadSearch reads the new papers in your field each day and brings you the few worth reading, each with its key finding in a sentence.