G1 WBT Inspire: Robot End-Effector and Hand States for Cloth Manipulation

Unitreerobotics provides a dataset of robot state observations for a task involving putting clothes into a basket. The data includes end-effector states computed via forward kinematics and finger states for both hands, with dimensionality dependent on hand type. The dataset was last updated on March 27, 2026.

Use Cases

• Training reinforcement learning agents for cloth manipulation based on end-effector and hand state observations.
• Developing forward/inverse kinematics models based on the described end-effector state data.
• Benchmarking robotic hand control algorithms based on the finger state data for the Inspire Hand.
• Analyzing time-series patterns in robot state during a specific manipulation task.

Strengths

• Includes end-effector states computed via forward kinematics, incorporating waist contribution.
• Provides hand state data for both hands, with specific details for the Inspire Hand (range 0.0–1.0).
• Dataset is hosted on a reputable platform (Hugging Face) and was updated in 2026.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count, file formats, and sample data are unknown, which may limit suitability assessment.
• Description metadata is limited; actual data quality requires manual inspection after download.

Provenance

Source

unitreerobotics via Hugging Face Datasets.

Collection Method

Likely collected from robot simulations or physical experiments.

Time Range

null

Freshness

Last updated 2026-03-27 12:06:00; freshness should be verified.

Geography

null