WidowX AI
Trossen Robotics
WidowX AI
Description
WidowX AI: The intelligent robotic arm for your AI projects
WidowX AI by Trossen Robotics is a 6-degree-of-freedom (DoF) articulated robotic arm designed for experimentation, training, and intelligent robotics research. With a 700 mm reach, 1.5 kg payload, and 1 mm accuracy, it enables reliable execution of complex manipulations.
Available in various configurations (Base, Leader, or Follower), the WidowX AI can be equipped with precision grippers (Base version) or ambidextrous pinchers (Leader version). The Follower version includes precision grippers and an Intel® RealSense™ D405 camera mounted on the arm for 3D vision.
It is powered by the iNerve controller and supports a wide range of software: Hugging Face, LeRobot, ROS2, MuJoCo, Gazebo, NVIDIA Isaac, and more.
Its modularity and software compatibility make it an excellent platform for AI model training, simulation, teleoperation, and the development of intelligent robotic behaviors in real or simulated environments.
Applications of WidowX AI
- Academic research in robotics and artificial intelligence
- Training in robotic manipulation and advanced control
- Development of 3D vision systems and object detection
- Data collection and annotation for machine learning
- Prototyping of human-robot interfaces (HRI) and robotic assistants
- Dynamic simulation (MuJoCo, Gazebo) and reinforcement learning
- Integration with AI pipelines: Hugging Face, LeRobot, Google Colab
Technical Specifications of WidowX AI
- Degrees of Freedom: 6 DoF
- Payload: 1.5 kg
- Reach: 700 mm
- Total span: 1400 mm
- Repeatability: 1 mm
- Accuracy: 1 mm
- Weight: 4 kg
-
Available configurations:
- Base: Precision gripper with silicone tips
- Leader: Ambidextrous hand with sliding rail pinchers
- Follower: Precision gripper + Intel® RealSense™ D405 camera with integrated mount
-
Box contents:
- 1 x WidowX AI (chosen config)
- 1 x 24V / 25A power supply
- 1 x iNerve controller + PC interface
- 1 x accessory bag
- Software compatibility: Hugging Face, LeRobot, ROS2, MuJoCo, Gazebo, NVIDIA Isaac