metadata
license: apache-2.0
task_categories:
- robotics
tags:
- LeRobot
- RoboTwin
- bimanual
- simulation
configs:
- config_name: default
data_files: data/*/*.parquet
This dataset was created using LeRobot.
Dataset Description
Bimanual manipulation demonstrations for the Stack Blocks Two task, generated in the RoboTwin simulator with domain randomization.
- Robot: TRLC DK1 bimanual (
bi_dk1_follower) - 2x 6-DOF arms with parallel grippers - Task: Stack Blocks Two
- Episodes: 1000
- Total frames: 217,335 (213.1 min @ 17 fps)
- Avg episode length: 12.8s
Cameras
| Camera | Resolution | Codec | FPS |
|---|---|---|---|
observation.images.head |
640x360 | h264 | 17 |
observation.images.left_wrist |
640x360 | h264 | 17 |
observation.images.right_wrist |
640x360 | h264 | 17 |
Action & State Space
action: float32[14] - joint position targets (6 joints + 1 gripper per arm)observation.state: float32[14] - current joint positions + gripper state
Task Descriptions
- "Shift red block and green block to the middle, then stack green block over red block."
- "Bring red block to the center, then stack green block over it."
- "Center red block using the right arm, then stack green block with the left arm on top."
- "Shift red block and green block to the center and stack green block on red block."
- "Place red block and green block in the center, then stack green block above red block."
Loading the Dataset
from lerobot.datasets.lerobot_dataset import LeRobotDataset
dataset = LeRobotDataset("andreaskoepf/robotwin_dk1_realcam_stack_blocks_two_1000")
frame = dataset[0]
print(frame["observation.state"].shape) # torch.Size([14])
print(frame["action"].shape) # torch.Size([14])
Citation
@misc{robotwin2025,
title = {RoboTwin: Dual-Arm Robot Benchmark with Generative Digital Twins},
author = {Chen, Yao Mu et al.},
year = {2025},
url = {https://github.com/TianxingChen/RoboTwin}
}