README.md

---
task_categories:
- robotics
tags:
- LeRobot
- robotics
- robot-learning
- imitation-learning
- manipulation
- pick-and-place
- so101
- domain-randomization
configs:
- config_name: default
  data_files: data/*/*.parquet
---

This dataset was created using [LeRobot](https://github.com/huggingface/lerobot).

## Dataset Summary

- **Task:** Move a **red block** into a **white box** (pick-and-place / block-in-bin).
- **Scene setup:** The **white box remains fixed** in the workspace.
- **Variations across episodes:**
  - **Red block position randomization:** The block starts at different positions within a **small neighborhood** around the nominal start location (small translations).
  - **Lighting variation:** Ambient illumination changes slightly between episodes (e.g., intensity and/or direction), while the scene layout remains the same.
- **Learning goal:** Encourage policies that are robust to **small spatial shifts** and **appearance changes** from lighting.

## Supported Tasks and Use Cases

This dataset is suitable for:

- **Imitation learning** with moderate domain randomization.
- **Visual robustness experiments** (sensitivity to illumination changes).
- **Generalization evaluation** for pick-and-place under small distribution shifts.
- **Baseline comparisons** against:
  - fixed block position / fixed lighting datasets,
  - block-rotation-only datasets,
  - position-randomized-only datasets.

## Task Description

**Instruction:** “Place the red block into the white box.”

A typical episode includes: approach the block → grasp → move to the box → release into the box → optional return to rest.

With **position randomization** and **lighting variation**, successful policies should:
- localize the block under changing appearance,
- plan motion conditioned on current block position,
- maintain grasp reliability when shadows/highlights vary.

## Dataset Structure

[meta/info.json](meta/info.json):
```json
{
    "codebase_version": "v3.0",
    "robot_type": "so_follower",
    "total_episodes": 50,
    "total_frames": 20724,
    "total_tasks": 1,
    "chunks_size": 1000,
    "data_files_size_in_mb": 100,
    "video_files_size_in_mb": 200,
    "fps": 30,
    "splits": {
        "train": "0:50"
    },
    "data_path": "data/chunk-{chunk_index:03d}/file-{file_index:03d}.parquet",
    "video_path": "videos/{video_key}/chunk-{chunk_index:03d}/file-{file_index:03d}.mp4",
    "features": {
        "action": {
            "dtype": "float32",
            "names": [
                "shoulder_pan.pos",
                "shoulder_lift.pos",
                "elbow_flex.pos",
                "wrist_flex.pos",
                "wrist_roll.pos",
                "gripper.pos"
            ],
            "shape": [
                6
            ]
        },
        "observation.state": {
            "dtype": "float32",
            "names": [
                "shoulder_pan.pos",
                "shoulder_lift.pos",
                "elbow_flex.pos",
                "wrist_flex.pos",
                "wrist_roll.pos",
                "gripper.pos"
            ],
            "shape": [
                6
            ]
        },
        "observation.images.gripper": {
            "dtype": "video",
            "shape": [
                1080,
                1920,
                3
            ],
            "names": [
                "height",
                "width",
                "channels"
            ],
            "info": {
                "video.height": 1080,
                "video.width": 1920,
                "video.codec": "av1",
                "video.pix_fmt": "yuv420p",
                "video.is_depth_map": false,
                "video.fps": 30,
                "video.channels": 3,
                "has_audio": false
            }
        },
        "observation.images.top": {
            "dtype": "video",
            "shape": [
                1080,
                1920,
                3
            ],
            "names": [
                "height",
                "width",
                "channels"
            ],
            "info": {
                "video.height": 1080,
                "video.width": 1920,
                "video.codec": "av1",
                "video.pix_fmt": "yuv420p",
                "video.is_depth_map": false,
                "video.fps": 30,
                "video.channels": 3,
                "has_audio": false
            }
        },
        "timestamp": {
            "dtype": "float32",
            "shape": [
                1
            ],
            "names": null
        },
        "frame_index": {
            "dtype": "int64",
            "shape": [
                1
            ],
            "names": null
        },
        "episode_index": {
            "dtype": "int64",
            "shape": [
                1
            ],
            "names": null
        },
        "index": {
            "dtype": "int64",
            "shape": [
                1
            ],
            "names": null
        },
        "task_index": {
            "dtype": "int64",
            "shape": [
                1
            ],
            "names": null
        }
    }
}
```


📊 Interested in the dataset? Contact Unidata to discuss purchase terms: [https://unidata.pro](https://unidata.pro/datasets/lerobot-so-101-manipulations/?utm_source=huggingface-Martsinian&utm_medium=referral&utm_campaign=record-red-block)

## Citation

**BibTeX:**

```bibtex
[More Information Needed]
```