RoboAuto: Autonomous Robot Program Search through Multi-Level Physical Understanding
RoboAuto: Autonomous Robot Program Search through Multi-Level Physical Understanding
This paper presents RoboAuto, a framework for autonomous robot program search that explicitly incorporates physical understanding at the object, scene, and interaction levels. It combines a hierarchical knowledge base with scene grounding and multi-stage planning to improve performance on robotic manipulation tasks. RoboAuto achieves an 89.1% success rate on short-horizon LIBERO-PRO tasks and outperforms existing methods on long-horizon real-robot tasks.
RoboAuto introduces a framework for autonomous robot program search that integrates multi-level physical understanding, improving success rates on both simulated and real-world tasks compared to existing methods.
Source: arXiv Robotics — research abstracts · Read original article ↗
Loading article text…
What the source reports
Publisher-reported claims, with original evidence. These results have not been independently verified by RoboSignal.
Reported numbers
Reported success rate
89.1%
View original evidence
physical effects and actively tests competing hypotheses to resolve physical uncertainties. RoboAuto achieves superior performance across five benchmark families, attaining an 89.1% success rate on short-horizon LIBERO-PRO tasks and more than twice the mean success rate of ASPIRE on LIBERO-PRO Long and LIBERO-Plus. It also outperforms ASPIRE and $\pi_{0.5}$ on three challenging long-horizon real-robot tasks requiring
Open source S5
- paper: not_reported
- code: not_reported
- weights: not_reported
- dataset: not_reported
What remains unknown
Not established in the collected evidence: Environment, Control, Data origin.
Reported performance applies to the described task. It does not establish general autonomy or deployment readiness.
Source excerpts and review record
Automatically extracted; no manual editorial approval recorded.
physical effects and actively tests competing hypotheses to resolve physical uncertainties. RoboAuto achieves superior performance across five benchmark families, attaining an 89.1% success rate on short-horizon LIBERO-PRO tasks and more than twice the mean success rate of ASPIRE on LIBERO-PRO Long and LIBERO-Plus. It also outperforms ASPIRE and $\pi_{0.5}$ on three challenging long-horizon real-robot tasks requiring
Open source S5
Source：arXiv Robotics — research abstracts · arxiv.org