Robots Atlas>ROBOTS ATLAS
FourierSDK

Development · SDKs

FourierSDK

2.1.0·Fourier Intelligence

Active Real-time capable API available
CATEGORYDevelopment · SDKs
READINESSTRL 8
ADOPTION SCALEGrowing Community
LICENSESApache-2.0
FIRST RELEASE2023

FourierSDK is the official developer toolkit from Fourier Intelligence (傅利叶智能), headquartered in Shanghai — one of the leading Chinese humanoid robot manufacturers. The SDK is designed for programming and research on GR (General Robot) series robots, primarily the GR1 and GR2.

The FourierSDK architecture is modular: the communication core is built on gRPC and MQTT (for WiFi connectivity), with a Python API layer and an optional ROS 2 interface. The main components are: FSA Motor SDK (EtherCAT/CAN motor control), GR Body SDK (kinematics, dynamics, gait trajectory planning), Arm SDK (control of 7-DOF manipulation arms with grippers), and Teleop SDK (teleoperation interfaces via haptic devices and VR controllers).

FourierSDK supports control modes: position, torque, impedance, and mixed. Integration with the Fourier Simulation Environment (based on Isaac Sim and MuJoCo) enables virtual prototyping before hardware deployment. The SDK supports ROS 2 through the fourier_ros2 package, enabling integration with Nav2, MoveIt 2 and the broader mobile robotics ecosystem.

In the robot learning context, Fourier Intelligence actively collaborates with research universities in China and globally — the SDK is used to collect demonstration data via teleoperation (haptic arms, VR). The GR1 platform participated in NVIDIA GR00T (cross-embodiment training) projects and is one of the platforms supported by LeRobot from Hugging Face.

FourierSDK is available via GitHub (fftai organisation) as a mixed open/closed source: the communication layer is open-source (MIT/Apache), while dynamic models and advanced control algorithms require an agreement with Fourier Intelligence.

Type & Roles
Software types
Robot manufacturer SDK
API Library

An API Library is a software package that exposes programmatic interfaces for communicating with a device, service, or system. In robotics it typically forms a lightweight integration layer built on top of the manufacturer's official API or an open-source project, abstracting low-level protocol details and providing language-native bindings (Python, C++, Java, etc.).

Control Stack

A Control Stack is the set of software components responsible for control logic, motion planning, command execution, and coordination of a robot's actuators. It receives high-level goals or waypoints (from planning or AI layers) and translates them into low-level actuator commands, closed-loop feedback control, and safety checks. Common implementations include MoveIt (arm manipulation), Nav2 (mobile navigation), and custom whole-body control stacks for humanoids.

SDK

An SDK (Software Development Kit) is a curated set of libraries, interfaces, tools, sample code, and documentation intended for building applications and integrating with a specific hardware device, platform, or service. In robotics, an SDK typically exposes device control, telemetry, sensor access, configuration, and execution functions, significantly reducing the time-to-first-integration for developers targeting a specific robot or platform.

Select an item to see its description.
Main category
SDKsControl & PlanningMiddlewareDeveloper Tools
Roles in robotics ecosystem
Motion PlanningPerceptionTeleoperationDeveloper Enablement
Robot Control

Robot Control denotes the role of software responsible for motion control, command execution, coordination of actuating elements and the direct operational logic of the robot.

Robot Learning

The robot learning role describes software for training a robot's control policies and manipulation/locomotion skills using machine learning methods. It covers: reinforcement learning in simulation with sim-to-real transfer, imitation learning and learning from demonstration, training Vision-Language-Action (VLA) models, and fine-tuning robotics foundation models. It typically uses massively parallelized simulation environments (Isaac Lab, MuJoCo) to generate training data, then deploys the trained policies on a physical robot.

Device Integration

Device Integration denotes the role of software responsible for communication, configuration, initialization and handling of specific devices, sensors, controllers or hardware components within a robotic system.

Select an item to see its description.
Software family
Family

A family of official SDKs and APIs provided by robot manufacturers: Boston Dynamics (Spot SDK), ABB (RobotStudio), KUKA (RSI / KUKA Sunrise), Universal Robots (URCap / RTDE), Agility Robotics (Digit SDK), Apptronik (Apollo SDK).

Maturity & Adoption
8 / 9
Prototype / pilot phase
ResearchPrototypeProduction
Adoption scaleGrowing Community
Maintenance statusActively Maintained
First release2023
Last update20 May 2026
Deployments

NVIDIA GR00T cross-embodiment training — Fourier GR1 is one of the platforms on which demonstrations are collected for GR00T N1 training. Fourier Intelligence announced a partnership with NVIDIA at GTC 2025. Hugging Face LeRobot — GR1 is among the platforms supported by the LeRobot robot learning framework for imitation learning. Peking University robotics lab and Shanghai Jiao Tong University — FourierSDK is used in research projects on whole-body control and bimanual manipulation. Fourier Intelligence R&D — internal deployment of GR1/GR2 at the Shanghai laboratory for collecting household manipulation demonstrations for VLA model training. AI-driven demonstrations at trade shows — CES 2024, World Artificial Intelligence Conference (WAIC) 2024 Shanghai — GR1 with FourierSDK demonstrated autonomous manipulation task execution.

Community

GitHub: fftai organisation — ~800 combined stars, ~150 forks, ~15 contributors (mainly internal team). Repositories: fourier-grx, fourier-ros2. Chinese community — active on WeChat developer groups for Fourier (a few hundred participants). Chinese-language documentation + English for partners. Academic partners — ~10 research groups in China use GR1 + FourierSDK as a platform for whole-body control, bimanual manipulation and robot learning research. Developer forum — GitHub Issues as the main support channel. Fourier Intelligence offers commercial technical support for enterprise customers. Growing global community after the NVIDIA GTC 2025 partnership announcement — increased interest from Western robotics labs.

Integrates with
R2
ROS 2
Open-source framework for building robot software. The successor to ROS 1, built on DDS with native support for distributed, real-time and multi-platform systems. The de facto standard in research and commercial robotics.
→
R2
ROS 2 Humble Hawksbill
LTS release of the ROS 2 framework based on Ubuntu 22.04, supported through May 2027. The most widely deployed release in humanoids, AMRs, and research platforms. Full integration with Nav2, MoveIt 2 and ros2_control.
→
R2
ROS 2 Jazzy Jalisco
ROS 2 Jazzy Jalisco — the latest LTS (May 2024, supported until May 2029) on Ubuntu 24.04. C++17, Fast DDS 3.0, improved executors, RMW Zenoh, better QoS for variable networks.
→
N
Nav2
The professionally-supported successor to the ROS Navigation Stack — a production-grade navigation framework for mobile robots, quadrupeds, and AGVs. Behavior Trees, planners (Smac, NavFn, Theta*), controllers (MPPI, RPP, DWB), AMCL, costmaps.
→
M2
MoveIt 2
Open-source motion planning, manipulation, and kinematics framework for ROS 2 (Foxy → Jazzy). Stewarded by PickNik Robotics. The de facto standard for manipulators in the ROS ecosystem.
→
GH
Gazebo Harmonic (gz-sim)
An LTS release of the new-generation Gazebo (gz-sim), released in September 2023 — the successor to Gazebo Classic. Modular architecture, integrated with ROS 2 Humble/Iron/Jazzy, supported until September 2028.
→
M(
MuJoCo (Multi-Joint dynamics with Contact)
Open-source rigid-body physics engine with accurate contact and friction, by Emo Todorov. Apache 2.0 since 2021, maintained by DeepMind. Standard for learning-based robotics (RL) and sim-to-real.
→
NI
NVIDIA Isaac Lab
NVIDIA open framework for GPU-based robot learning, built on Isaac Sim. Trains RL policies across thousands of parallel environments. Successor to Isaac Gym and Orbit; Apache 2.0 since 2024.
→
L
LeRobot
LeRobot (Hugging Face) — open-source robot learning framework with ACT, Diffusion Policy and TDMPC implementations. Standardises data collection and policy training for manipulators and mobile robots.
→
GN
GR00T N1
NVIDIA GR00T N1 — open foundational model for humanoids: dual-system VLA (fast motor policy + slow semantic reasoning), trained on cross-embodiment data, available through NVIDIA Isaac.
→
Related robotics software
BD
Boston Dynamics Spot SDK
Boston Dynamics' official developer kit for the Spot quadruped robot. Python + gRPC over TLS — full control over pose, motion, cameras, the Spot Arm manipulator, GraphNav maps, and autonomous missions.
→
AR
Agility Robotics Digit SDK
The official SDK for the Digit humanoid robot (Agility Robotics). Agility Arc cloud + JSON-RPC API + Python/C++ clients. Logistics-focused: pick & place, totes, pallets, integration with WMS/MES. ROS 2 support.
→
AA
Apptronik Apollo SDK
An SDK for the Apollo humanoid (Apptronik) — a general-purpose platform for manufacturing and logistics. Skill-based programming, Gemini Robotics integration (DeepMind). Partners: Mercedes-Benz, GXO, NASA.
→
US
Unitree SDK2
Unitree SDK2 — cross-platform Unitree Robotics SDK (C++/Python) for controlling Go2/B2 quadrupeds and G1/H1 humanoids over CycloneDDS. Supports ROS 2, torque/position control, AI-enable payload and autonomy.
→
R2
ROS 2
Open-source framework for building robot software. The successor to ROS 1, built on DDS with native support for distributed, real-time and multi-platform systems. The de facto standard in research and commercial robotics.
→
M2
MoveIt 2
Open-source motion planning, manipulation, and kinematics framework for ROS 2 (Foxy → Jazzy). Stewarded by PickNik Robotics. The de facto standard for manipulators in the ROS ecosystem.
→
L
LeRobot
LeRobot (Hugging Face) — open-source robot learning framework with ACT, Diffusion Policy and TDMPC implementations. Standardises data collection and policy training for manipulators and mobile robots.
→
GN
GR00T N1
NVIDIA GR00T N1 — open foundational model for humanoids: dual-system VLA (fast motor policy + slow semantic reasoning), trained on cross-embodiment data, available through NVIDIA Isaac.
→
ROS supportCompatibility with ROS / ROS 2 ecosystem
Official Vendor ROS 2 WrapperOficjalny wrapper ROS 2 tworzony i utrzymywany przez producenta sprzętu lub oprogramowania
ROS 2 Node (Standalone)Samodzielny węzeł ROS 2 (node) jako główna forma dystrybucji
ROS 2 Hardware Interface (ros2_control)Hardware Interface dla frameworka ros2_control
ROS 2 Launch IntegrationGotowe pliki launch ROS 2 jako forma integracji systemowej
System capabilities
⊙
Open source
Source code is publicly available under an open-source license — enables security audits, custom modifications, and integration without licensing barriers.
×
⚡
Real-time capable
Designed with timing-determinism guarantees — meets the requirements of control loops, safety systems, and tasks demanding low, predictable latency.
✓
⟨/⟩
API available
The software exposes a programmable interface (REST, gRPC, SDK, or language bindings) that enables automation and integration with other systems.
✓
📦
Pre-built / binary
Distributed as ready-to-use binary packages, container images, or installers — no need to build from source.
✓
Programming languages
PythonC++C
Operating systems
Ubuntu 22.04Ubuntu 20.04Debian
Ubuntu 24.04

Ubuntu 24.04 LTS 'Noble Numbat' — supported until April 2029. The host for ROS 2 Jazzy.

Select an item to see its description.
Minimum hardware requirements
Minimum hardware requirements
CPUOn-robot: NVIDIA Jetson AGX Orin 64GB (recommended) or Jetson Orin NX 16GB. Workstation: x86_64 minimum 8-core Intel/AMD with CUDA GPU for simulation. ARM64 supported.
RAM (GB)Minimum 16 GB RAM. Zalecane 32 GB dla środowisk symulacyjnych (Isaac Sim/MuJoCo). On-robot 32 GB dla pełnego stanu czujników.
GPUOptional for robot control. Required for vision (D435i, ZED) and robot learning pipeline. Minimum NVIDIA RTX 3060 for simulation.
Disk (GB)Minimum 50 GB SSD. Środowisko symulacji Isaac Sim ~20 GB, pełne środowisko deweloperskie ~100 GB.

EtherCAT master required on the control PC for low-level motors. Ubuntu 22.04 LTS recommended. 1G Ethernet cable between control PC and robot unit.

Packaging & distribution
Package managers
pip / PyPIapt / debGitHub Releases / GitHub Actions ArtifactsSource – CMake / ament_cmakeDocker / Docker Hub
CPU architectures
x86_64 (AMD64)ARM64 / AArch64NVIDIA Jetson – AArch64 (JetPack)aarch64 / ARM64
Installation difficulty
LevelAdvanced
Protocols and interfaces
Communication protocols
gRPCROS 2 TopicsROS 2 ActionsROS 2 ServicesDDS (Data Distribution Service)MQTT (Message Queuing Telemetry Transport)Ethernet / TCP-IPCAN bus (Controller Area Network)EtherCAT
Hardware interfaces
Ethernet 1000BASE-T (Gigabit Ethernet)Ethernet 10GBASE-T (10 Gigabit Ethernet)EtherCAT (Slave Interface)CAN 2.0A / 2.0BCAN FD (Flexible Data-Rate)USB 3.0 / 3.1 Gen 1MIPI CSI-2
Latency classes
Hard Real-Time (1–5 ms)Hard Real-Time (5–20 ms)Soft Real-Time (20–100 ms)
Deployment types
On RobotEdgeLocal WorkstationContainerized
Supported simulators
MuJoCo
NVIDIA Isaac Sim
NVIDIA Isaac Lab
Gazebo Harmonic
Licenses
Apache-2.0Apache License 2.0v2.0

License family: Permissive

ModificationDistributionCommercial useSublicensingPrivate useROS-compatibleOSI approvedFSF Free/LibreRequires attributionPatent grant
LicenseRef-ProprietaryProprietary – All Rights Reserved

License family: Proprietary – Commercial

Version history
2.1.0Jan 2026

GR2 support, new bimanual manipulation modules, integration with NVIDIA GR00T fine-tuning pipeline.

2.0.0Jun 2025

Refactored to gRPC-first architecture, new Teleop SDK with VR headset support, fourier_ros2 package for Humble/Jazzy.

1.5.0Jan 2025

MuJoCo and Isaac Sim simulators officially supported, GR Body SDK full kinematic chain API.

1.2.0Jul 2024

Arm SDK 7-DOF manipulator control, haptic teleoperation device support (Rokoko, HEBI).

1.0.0Feb 2024

First public release of FourierSDK for GR1. FSA Motor SDK, basic gait API.

0.5.0Sept 2023

Internal beta for research partners. Basic gRPC communication API.