6DOF SO-ARM101 Pro Dual Follower Arm Robotic Arm with TPU Gripper Optional 2MP Camera for XLerobot

The SO-ARM101 is an open-source robotic arm deeply optimized based on the SO-ARM100 model. It optimizes circuit layout and motor gear matching to completely resolve the joint wire breakage issue. With enhanced performance, it supports real-time follow-up between master and slave robotic arms. It can seamlessly adapt to the Hugging Face LeRobot tool, directly connecting PyTorch models and shared data to easily deploy imitation learning and reinforcement learning. Complete assembly tutorials and DIY kits are provided, allowing students, researchers and makers to engage in learning, research and creation of intelligent robots and embark on your robot innovation journey.

* High-precision Magnetic Encoding

* End-to-End Imitation Learning

* Hugging Face LeRobot

* Teleoperation Cooperative Algorithm

* Precise Data Collection

* ACT Model Training

* On-board Camera Vision (Optional Camera)

* Scene Camera View

* Real-time Precise Position Feedback

* 6DOF

* Master-slave Arm Structure

* 30KG High Torque Bus Servo

* High-definition Wide-angle Camera

* Secondary Development Supported

* Textbook-level Tutorials

The SO-ARM101 robotic arm is easy to operate with simple assembly and no complex debugging required. It supports action demonstration learning and adopts a master-slave arm structure. After the master arm completes grasping, placing and other movements, the slave arm can synchronously imitate in real time, and it can independently complete preset tasks after a small amount of training. It can directly connect to shared AI models and data without developing codes from scratch. It is suitable for experiments such as grasping and simple assembly. Students, researchers and makers can leverage its low cost to experience intelligent robot applications and efficiently carry out embodied intelligent practice.

Upgradeable Flexible Gripper

* It features a hollow design with internal reinforcing ribs, adopting the operating principle of a fin-type gripper. It can grasp objects more firmly and adapt to the shapes of items automatically.

* When the fin-shaped structure contacts and clamps objects, the top and bottom of the fins bend toward the object, enabling the gripper to wrap around the object surface to expand the stress area and achieve stable grasping effects.

* It is specially designed for soft and fragile objects that cannot be gripped by conventional mechanical clamps. It can be used to pick up fruits, glass products, eggs, etc.

* It can adapt to a variety of circular-based profiles, including cylinders, spheres, ovals and polygons.

Master-Follower Collaborative Imitation Learning

* The robotic arm integrates the Hugging Face AI framework. It collects demonstration motions via master-follower teleoperation, supports one-click training of imitation learning models, and enables autonomous optimization and strategy deployment.

* The system can independently complete complex tasks and adapt to varying environments. The full closed-loop workflow improves automation and intelligence, accelerating the rapid deployment of research and engineering projects.

Dual-view Collaborative Global Coverage

* Combined with the on-arm bracket camera (optional) of SO-ARM101, it captures the spatial position, angle and surface texture of target objects at close range, improving the efficiency of refined data collection.

* The desktop stand scene camera (optional) identifies workspace environmental information in real time, rapidly responds to environmental changes, ensures precise robotic arm operation, avoids motion deviation and stalling, and boosts the stability and reliability of task execution.

 

Tutorial Materials

* Complete and detailed full-process reproduction tutorials are provided. Users without programming experience can get started easily, enabling barrier-free reproduction and rapid deployment for all users.

* You may search Juxi-Technology on Github or visit https://www.juxitech.com/. Select Lark under Document to view materials. Please contact customer service if you encounter any issues.

Seamless Connection with Hugging Face Community

* The SO-ARM101 robotic arm achieves seamless integration with the Hugging Face community. It can directly reuse pre-trained models, datasets and algorithm library resources provided by the LeRobot project. It supports collaborative sharing of codes and experience with global developers, lowers technical barriers in research and development, and offers a technical entry point for users to access embodied intelligence, a cutting-edge AI field.

Open-Source Project

* This product is based on the open-source LeRobot SO-ARM101 project. https://github.com/huggingface/lerobot

* It complies with the Apache License 2.0 and is fully compatible with the LeRobot ecosystem. Users can directly integrate and utilize it to reduce development thresholds and accelerate project roll-out.

ROS 2 Simulation Control

* It supports real physical hardware control, MoveIt motion planning and advanced task execution.

Model Edge Deployment

* Compatible with NVIDIA Jetson Orin Nano/Orin NX

* Compatible with D-Robotics RDK X5/S100, supporting end-to-end acceleration via BPU calls

30KG Magnetic Encoder Dual-Axis Bus Servo

* The Feetech STS3215 servo integrates two high-quality magnetic encoder motors internally, equipped with a 12-bit high-precision magnetic encoding angle sensor, supporting 360-degree controllable rotation. It features data read-back and multiple operating modes, delivering millisecond-level coordinated response to enable smooth and seamless movements of the robotic arm.

Optional Camera

* 1080P high-definition 2-megapixel camera modules are optional.

* Optional 86° field-of-view 30FPS zoom camera or 100° wide-angle 60FPS fixed-focus camera

* The camera modules support USB driver-free plug-and-play operation and enables high-precision image capture and processing.

Auto Focus Camera (Optional)

* Switchable between Auto Focus and Manual Focus

* 1080P at 30FPS

* 86° Distortion-Free Lens

* Dual Mode Switching for Dual Scenario Adaptation: 1. Auto Focus Mode: When mounted on the robotic arm gripper, it tracks moving targets in real time without manual focal length adjustment, ensuring consistently clear frames during dynamic operations such as grasping and object detection. 2. Manual Focus Mode: When installed on a fixed bracket, switch to fixed focus to lock focal length, preventing frame shift caused by environmental vibration or accidental touch. It supports stable shooting for static scenarios including quality inspection, monitoring and environment recording.

Fixed Focus Camera (Optional)

* 2MP Fixed Focus

* 1080P at 60FPS

* Smooth Image Capture at 60 High Frames Per Second

* Ultra-Wide 100° Field of View

Servo Driver Board

* This serial bus servo driver board features a concise design and plug-and-play functionality. It is equipped with a TVS voltage stabilizing protection module to prevent high-voltage breakdown. It can be quickly connected to mainstream controllers including Raspberry Pi, D-Robotics RDK and Jetson via a single connecting cable, supporting real-time motor control and encoder data feedback. It is compatible with the LeRobot SO-ARM101 robotic arm and most serial bus servos available on the market, serving as an ideal solution for robot development.

Application Scenarios

* University Scientific Research: For teaching experiments and research projects of majors related to robotics, artificial intelligence, automation and other disciplines.

* Developers/Makers: Applicable to algorithm development, product prototype verification and participation in robotics competitions.

* Educational Institutions: An ideal teaching tool for advanced courses such as artificial intelligence programming.

* Senior Hobbyists: Build a cool desktop robot that enables hands-on practice of AI technology via DIY.

* Product Name: SO-ARM101

* Product Material: Bambu PLA+

* Size of Follower Robotic Arm: 111mm x 173mm x 532mm

* Weight of Follower Robotic Arm: Approx. 690g

* Gripper Material: TPU

* Servo Driver Board: Bus Servo Adapter

* Optional Camera: 2MP HD Camera (1080P)

* Power Supply: DC5521 12V 5A

* Operating System: Windows/Linux

* Control Method: USB Serial Communication

* Programming Language: Python

* 12V Follower Robotic Arm (with 6 STS3215-C018 servo motors and TPU flexible gripper) x2

* 12V 5A Power Adapter x1

* Servo Driver Board x2

* Type-C Data Cable x2

* USB3.0 Expansion Hub x1

* F-type Clamp x4

* Phillips Screwdriver x1

* 12V Follower Robotic Arm (with 6 STS3215-C018 servo motors and TPU flexible gripper) x2

* 12V 5A Power Adapter x2

* Servo Driver Board x2

* Type-C Data Cable x2

* USB3.0 Expansion Hub x1

* G-type Clamp x4

* Phillips Screwdriver x1

* SO-ARM101 Arm-mounted Bracket x2

* Zoom 30FPS Camera Module x2

* 1.5m USB Cable x2

* 12V Follower Robotic Arm (with 6 STS3215-C018 servo motors and TPU flexible gripper) x2

* 12V 5A Power Adapter x2

* Servo Driver Board x2

* Type-C Data Cable x2

* USB3.0 Expansion Hub x1

* G-type Clamp x4

* Phillips Screwdriver x1

* SO-ARM101 Arm-mounted Bracket x2

* Fixed Focus 60FPS Camera Module x2

* 1m USB Cable x2

* Package weight: 4.5kg