SinjaeKang
fatigue-injection-pi0.5-cube-stacking
Franka Panda cube-stacking rollouts collected with square-wave fatigue perturbations injected into end-effector motion across five perturbation magnitudes. Successful episodes only, generated by π0.5 policy.
Downloads0
Episodes250
Why This Matters for Physical AI
This dataset enables study of how neural policies respond to perturbations and fatigue-like disturbances in robotic control, providing insights into robustness and adaptation mechanisms for physical AI systems.
Technical Profile
- Modalities
- rgbproprioception
- Robot Embodiments
- Franka Panda
- Action Space
- end_effector_delta
- Environment
- lab
- Task Types
- manipulationpick_and_placecube_stacking
- Episodes
- 250
- Data Format
- LeRobot
- Annotation Types
- language_instructions
Access
Need custom rgb data?
Claru builds purpose-built datasets for lab applications with dense human annotations and quality assurance.
Request a Sample Pack