RoboRSI: Stable, efficient, and reusable robot self-evolution in complex real-world environments
arXiv cs.AI · 9 Oct 2026, 03:55 AEST
RoboRSI is a robot self-improvement system designed to enhance robot capabilities through experience. It organizes learning around task structures to ensure that improvements are validated and reusable.
Key facts
- A generalist robot should not only perform diverse tasks but also improve through experience, turning what it learns during execution into capabilities that later tasks can reuse.
- We introduce RoboRSI, a robot self-improvement system built on Top-Down Skill Refinement (TSR).
- On a mobile manipulator, RoboRSI develops multi-object household cleanup over 104 rounds.
- In simulation, it achieves the highest success rate on LIBERO, LIBERO-PRO, LIBERO-Plus, and RoboTwin, exceeding the strongest baseline by 2.7 to 11.0 percentage points.
Read aloud
RoboRSI is a robot self-improvement system designed to enhance robot capabilities through experience. It organizes learning around task structures to ensure that improvements are validated and reusable. A generalist robot should not only perform diverse tasks but also improve through experience, turning what it learns during execution into capabilities that later tasks can reuse. We introduce RoboRSI, a robot self-improvement system built on Top-Down Skill Refinement TSR. On a mobile manipulator, RoboRSI develops multi-object household cleanup over one hundred four rounds. In simulation, it achieves the highest success rate on LIBERO, LIBERO-PRO, LIBERO-Plus, and RoboTwin, exceeding the strongest baseline by two point seven to eleven point zero percentage points.
Sources in this cluster
- RoboRSI: Stable, efficient, and reusable robot self-evolution in complex real-world environments
arXiv cs.AI · 9 Oct 2026, 03:55 AEST