LIBERO

LIBERO stands for Lifelong Robot Learning Benchmark, a widely used robot manipulation benchmark platform that provides multiple tabletop manipulation environments, standardized task definitions, and demonstration data formats.

LIBERO-PeRM builds a procedural data generation pipeline on top of the LIBERO environment, ultimately constructing five task suites containing 312 tasks, with high-quality demonstrations, rich metadata, and preference satisfaction metrics.

The LIBERO benchmark contains four task suites: Spatial, Object, Goal, and Long; LIBERO-Pro is its enhanced version, increasing task complexity and the number of steps

LIBERO-RIG is a VLA model benchmark test whose design perturbs only the robot hardware and setup, keeping the LIBERO tasks themselves completely unchanged.

LIBERO-RIG covers 11 perturbation dimensions, with 5 severity levels set for each dimension, combining into 210 test conditions across four LIBERO suites.

In simulation environments, RoboRSI achieves the highest success rates on benchmarks such as LIBERO, LIBERO-PRO, LIBERO-Plus, and RoboTwin, exceeding the strongest baseline by 2.7 to 11.0 percentage points.

The LIBERO series is a widely used manipulation benchmark in the robot learning field, containing a large number of language-conditioned long-horizon manipulation tasks, used to evaluate the generalization and continual learning ability of policies.

The paper validates on three platforms: the simulation benchmarks RoboTwin and LIBERO, as well as a real robot platform.

The community is advancing standardized simulation benchmarks such as LIBERO, SimplerEnv, and HumanoidBench, as well as the DROID dataset based on real robots.

LIBERO simulation successfully runs long-horizon tasks on Mac for the first time, with GR00T achieving a 100% success rate, WAM achieving 98%, and a 2.8-billion-parameter VLA small model achieving 96%.