---
pretty_name: AgiBot World
size_categories:
- 100K<n<2000K
task_categories:
- other
language:
- en
tags:
- real-world
- dual-arm
- Robotics manipulation
---

<img src="assets/agibot_world.png" alt="Image Alt Text" width="70%" style="display: block; margin-left: auto; margin-right: auto;" />


<div align="center">
  <video src="assets/agibot_film.mp4" controls width="700">
    Your browser does not support the video tag.
  </video>
</div>



# Agibot World Colosseo (Agibot) 🌍🤖
- **Challenges with Current Benchmarks:**
  - Low data quality and limited sensing capabilities.
  - Short-horizon tasks in controlled environments.
- **First Large-Scale Robotic Learning Platform for Multi-purpose Robotic Manipulation.**
- **Includes data, models, benchmarks, and ecosystem.**
- **Aims to democratize real-robot data for the academic community.**
- **Aspires to trigger the “ImageNet moment” for Embodied AI in the near future.**

# Key Features 🔑
- **One Million+ Demonstrations from 100 robots.**
- **100+ real-world scenarios across 5 business sectors.**
- **Tasks involving:**
    - Fine-grained manipulation
    - Long-horizon planning
    - Dual-robot collaboration
- **Cutting-Edge Hardware:**
    - Visuotactile Sensors
    - 6-DoF hands
    - Mobile dual-arm robots with full-body control

# Research Potential 🌱
- **Supports multimodal imitation learning and multi-agent collaboration, and beyond.**
- **State-of-the-art hardware for scalable robotic systems in production.**
- **Researchers and practitioners are invited to leverage this newly open-source platform to shape the future of Embodied AI.**

# Platform Release 📅
- **Agibot Alpha: ~100,000 trajectories of real-robot data.**
- **Full platform suite coming by end of Q1 2025.**
- **Agibot-World Robot Manipulation Challenge**

<!-- # AgiBot World
Current robot learning datasets fall short of addressing real-world challenges, hindered by low data quality, limited sensing capabilities, and a focus on short-horizon tasks in controlled environments. 

Enter **AgiBot-World**: the first large-scale dataset designed to support advanced bi-manual manipulation and scalable real-world robotics. With over **a million high-fidelity demonstrations** collected from 100 standardized robots, it offers unprecedented diversity and complexity. Spanning **100+ real-world scenarios** across **five major industries**, it addresses challenging tasks featuring **fine-grained manipulation**, **tool usage**, and **dual-robot synergistic collaboration**. 

With advanced hardware, curated task design, and high-quality data, AgiBot-World sets a new benchmark for large-scale robot learning and drives the future of general-purpose robotic systems. Researchers and practitioners are invited to leverage this open-source resource to shape the future of robotics.

## Cutting-edge sensor and hardware design
AgiBot World features cutting-edge multimodal hardware, including array-based visuotactile sensors, durable 6-DoF hands, and mobile dual-arm robots with full-body control. 
## Wide-spectrum of Scenario Coverage
AgiBot World faithfully replicates five major types of scenes in a 1:1 scale: home, industrial logistics, office, supermarket, and catering. It also designs many tasks that frequently occur in real life but are not yet well solved by robots.

## Quality Assurance with Human-in-the-Loop
Through expert training, manual review, and failure annotation processes, AgiBot-World employs a comprehensive human-in-the-loop framework to ensure the production of high-quality data. -->


# Get started
## Download AgiBot Dataset
## Installation
## Dataset Structure
```
data
├── task_info
│   ├── task_374.json
│   ├── task_256.json
│   └── ...
├── 374[task_id]
│   ├── 64825[episode_id]
│   │   ├── depth
│   │   ├── parameters
│   │   ├── videos
│   │   ├── meta_info.json
│   │   └── aligned_joints.h5
│   ├── 67832
│   │   └── ...
│   └── ...
├── 321
│   └── ...
```
In the `aligned_joints.h5` file, we organize the information in this format.
```
aligned_joints.h5
├── action
│   ├── joint
│   ├── effector
│   └── ...
├── state
│   ├── joint
│   ├── effector
│   └── ...
```