LIBERO-RECOVER: Beyond Task Success Towards Failure Recovery in Robotic Manipulation Models
LIBERO-RECOVER: Beyond Task Success Towards Failure Recovery in Robotic Manipulation Models
This paper introduces LIBERO-Recover, a large-scale benchmark for evaluating failure recovery in robotic manipulation models. Built upon the LIBERO framework, it collects over 1,000 real execution failure scenarios across four recovery levels: Action Retry, Action Adaptation, Object State Recovery, and Environmental Recovery. The benchmark evaluates core capabilities including spatial understanding, object structure reasoning, interaction understanding, and topological reasoning, shifting the focus from task success to robustness after failure.
The authors introduce LIBERO-Recover, a benchmark evaluating failure recovery capabilities in robotic manipulation models across four levels (Action Retry, Action Adaptation, Object State Recovery, Environmental Recovery), addressing the gap between high success rates on ideal benchmarks and real-world robustness.
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
recovery levels
4
View original evidence
largely unmeasured, revealing a critical gap between benchmark performance and real world reliability. To address this gap, we introduce LIBERO-Recover Benchmark, a large scale benchmark for failure recovery in robotic manipulation. Built upon LIBERO, we collect real execution failures from SOTA embodied models and construct 1,000+ scenarios across four recovery levels: (1) Action Retry, (2) Action Adaptation, (3) Ob
Open source S4core capabilities
4
View original evidence
ject State Recovery, and (4) Environmental Recovery. We evaluate four core capabilities: spatial understanding, object structure reasoning, interaction understanding, and topological reasoning. As the first large-scale benchmark for embodied failure recovery, LIBERO-Recover shifts evaluation from \emph{Can the robot succeed?''} to \emph{Can the robot recover after failure?''}, promoting robust and generalizable embod
Open source S5
- paper: not_reported
What remains unknown
Dataset licensing and commercial-use rights are not established here.
Source excerpts and review record
Automatically extracted; no manual editorial approval recorded.
largely unmeasured, revealing a critical gap between benchmark performance and real world reliability. To address this gap, we introduce LIBERO-Recover Benchmark, a large scale benchmark for failure recovery in robotic manipulation. Built upon LIBERO, we collect real execution failures from SOTA embodied models and construct 1,000+ scenarios across four recovery levels: (1) Action Retry, (2) Action Adaptation, (3) Ob
Open source S4
arXiv:2609.05178v4 Announce Type: replace Abstract: Vision-Language-Action (VLA) or World Action (WAM) models have recently demonstrated remarkable performance in robotic manipulation. On LIBERO, SOTA method have achieved nearly 100\% success rates, seemingly suggesting that the models are ready for deployment in real world. However, near perfect performance on existing benchmarks can be misleading: success under ide
Open source S2
ject State Recovery, and (4) Environmental Recovery. We evaluate four core capabilities: spatial understanding, object structure reasoning, interaction understanding, and topological reasoning. As the first large-scale benchmark for embodied failure recovery, LIBERO-Recover shifts evaluation from \emph{Can the robot succeed?''} to \emph{Can the robot recover after failure?''}, promoting robust and generalizable embod
Open source S5
ied agents. The project will be avaible in \textcolor{blue}{https://liulin815.github.io/LIBERO-Recovery/}.
Open source S6
Implications for data suppliers
RoboSignal interpretation and collection questions, not statements of buyer demand.
- Confirm the required data type and collection setting with the buyer; this source does not establish a complete collection specification.
- Compare the reported units and scope before using these quantities in a budget. Recording hours, sensor-hours and trajectories are different measures.
- Validate demand and acceptance criteria with a buyer before scaling. Publication, popularity and a research result do not establish a purchase commitment.
Source：arXiv Robotics — research abstracts · arxiv.org