Skip to content

UT-Austin-RobIn/oopsieverse

Repository files navigation

OopsieVerse

This is the official codebase for OopsieVerse, a damage-aware, simulator-agnostic framework and benchmark for learning and evaluating safer robot manipulation. Please follow the documentation for getting started.

🌐 Website  ·  📄 Paper  ·  📚 Documentation

Citation

If you use OopsieVerse in your research, please cite:

@inproceedings{balaji2026oopsieverse,
  title={OopsieVerse: A Safety Benchmark with Damage-Aware Simulation for Robot Manipulation},
  author={Balaji, Arnav and Bahety, Arpit and Ambatipudi, Sriniket and Lam, Daniel and Xu, Junhong and Mart{\'\i}n-Mart{\'\i}n, Roberto},
  booktitle={Robotics: Science and Systems (RSS), 2026},
  year={2026}
}

Acknowledgements

OopsieVerse builds on BEHAVIOR-1K / OmniGibson, RoboCasa, and RoboSuite.

About

OopsieVerse provides a damage-aware, simulator-agnostic framework and benchmark for evaluating and learning safer policies

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors