File size: 3,332 Bytes
f00d746
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
f532dc6
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
---
license: apache-2.0
task_categories:
- robotics
- reinforcement-learning
tags:
- lerobot
- reachy2
- bimanual
- mobile-manipulation
- teleoperation
- kitchen
language:
- en
size_categories:
- 10K<n<100K
---

# Reachy2 Kitchen Multimodal Dataset

Teleoperation dataset for **Reachy2 mobile bi-manipulator** performing kitchen tasks in the **SAI platform**. Collected using [`sai-zoo`](https://github.com/ArenaX-Labs/sai-zoo.git) teleoperation tool and converted to [LeRobot](https://github.com/huggingface/lerobot) format.

## Dataset Statistics

| Metric | Value |
|--------|-------|
| Episodes | 219 |
| Total Frames | 23,644 |
| Tasks | 9 |
| FPS | 20 |
| Camera Resolution | 256×256 |

## Tasks

| Task | Description |
|------|-------------|
| Drawer | Open/close left drawer, Open/close right drawer |
| Stove | Turn on/off front-left, front-right, rear-left burners |

## Features

**Observation State** (36 dims):
- `base` (3): Mobile base position
- `torso` (1): Torso joint
- `left_arm` / `right_arm` (7 each): Arm joint positions
- `left_arm_eef_pos` / `right_arm_eef_pos` (3 each): End-effector position
- `left_arm_eef_quat` / `right_arm_eef_quat` (4 each): End-effector orientation (quaternion)
- `left_hand` / `right_hand` (1 each): Gripper position

**Action** (18 dims):
- `base` (3): Base velocity command
- `torso` (1): Torso command
- `left_arm_eef_pos` / `right_arm_eef_pos` (3 each): EEF position delta
- `left_arm_eef_rot` / `right_arm_eef_rot` (3 each): EEF rotation delta
- `left_gripper_close` / `right_gripper_close` (1 each): Gripper command

**Videos** (3 cameras):
- `agentview_center_rgb`: Front center camera
- `agentview_left_rgb`: Front left camera
- `agentview_right_rgb`: Front right camera

## Dataset Structure

```
reachy2-kitchen-multimodal/
├── data/
│   └── chunk-000/
│       └── episode_XXXXXX.parquet      # State, action, timestamps per episode
├── videos/
│   └── chunk-000/
│       ├── observation.images.cam_robot_0:agentview_center_rgb/
│       ├── observation.images.cam_robot_0:agentview_left_rgb/
│       └── observation.images.cam_robot_0:agentview_right_rgb/
│           └── episode_XXXXXX.mp4      # Video per camera per episode
├── meta/
│   ├── info.json                       # Dataset metadata & feature schema
│   ├── tasks.jsonl                     # Task index → description mapping
│   ├── episodes.jsonl                  # Episode index, task, length
│   └── modality.json                   # State/action/video index mapping
└── raw_dataset/                        # Raw sai-zoo teleoperation data
    ├── Reachy2IkKitchenDrawerClose-beta/
    ├── Reachy2IkKitchenDrawerOpen-beta/
    ├── Reachy2IkKitchenStoveTurnOff-beta/
    └── Reachy2IkKitchenStoveTurnOn-beta/
        └── YYYYMMDD_HHMMSS/            # Timestamped episode folder
            ├── env_metadata.npy        # Environment metadata
            ├── episode_metadata.npy    # Episode metadata
            └── episodes.hdf5           # Raw episode data
```

## Usage

**Load with LeRobot:**
```python
from lerobot.common.datasets.lerobot_dataset import LeRobotDataset

dataset = LeRobotDataset("CompeteSAI/reachy2-kitchen-multimodal")
```