AIRoA MoMa Dataset
AIRoA (AI Robot Association) · 2025 · datasets.bot · datasets.bot page
One-liner. Large-scale hierarchical mobile-manipulation dataset of ~25.5K Toyota HSR teleoperation episodes in LeRobot format.
Setup
- Datasets / benchmarks: The AIRoA MoMa Dataset is a large-scale, real-world dataset for mobile manipulation collected with the Toyota Human Support Robot (HSR) via leader-follower teleoperation in household environments. It contains 25,469 episodes (~94 hours) sampled at 30 Hz, with synchronized RGB images from head and wrist cameras, joint states, six-axis wrist force-torque signals, and internal robot states. A two-layer annotation schema of sub-goals and primitive actions supports hierarchical learning and error analysis across seven primary household tasks. The data is standardized in the LeRobot v2.1 format and released by AIRoA (AI Robot Association) under the GENIAC project (METI/NEDO). License: custom. Download: https://huggingface.co/datasets/airoa-org/airoa-moma.
- Hardware / simulator: Embodiment: not listed. Environment: home, kitchen. Realness: physical.
Schema
LeRobot v2.1: synchronized RGB (head + wrist), joint states, six-axis wrist force-torque, internal states; two-layer sub-goal + primitive-action annotation.
Links