Unitree Dex5: Intelligent Robotic Hand with 22 DOF and 94 Tactile Sensors for Humanoid Robotics
Unitree Dex5: Intelligent Robotic Hand with 22 DOF and 94 Tactile Sensors for Humanoid Robotics
Couldn't load pickup availability
SKU: 1601447473210
Description
- ✔ Unrivaled precision and movement control
- ✔ Each finger with its own smooth reverse drive
- ✔ Total of 22 degrees of freedom for maximum flexibility
- ✔ Independent replacement of all five fingers
- ✔ 94 integrated haptic sensors for detailed feedback
- ✔ Support for advanced development of tactile algorithms
- ✔ Compact yet powerful motors with high power density
Unitree Dex5 is a cutting-edge robotic hand solution for humanoid robotics. Designed with precision and unmatched dexterity in mind, this intelligent hand opens up new possibilities for research, development, and advanced applications. Its design enables smooth and natural movements that were previously the domain of the human hand.
⚙️ Key features that will take you further
- 22 degrees of freedom: Smooth and flexible movements for complex manipulations.
- Independent finger replacement: All 5 fingers can be replaced separately for easy maintenance and adaptation.
- Four-finger lateral swing: Improves adaptability to the surface of the grasped object and increases grip reliability.
- 94 haptic sensors: Provides detailed feedback for precise grip control.
- Algorithm development support: Enables secondary development of tactile algorithms for fine motor skills.
- High-performance motors: High-power density coreless motors ensure speed and agility.
- Transmission mechanism: Micro-force controlled composite articulated gears for low damping and minimal gaps.
🔍 Explore unprecedented dexterity and sensory richness
The Unitree Dex5 is designed for ultimate control and precision. With smooth reverse drive at all joints of each finger, it eliminates “stiff hands” and allows for much smoother operations, facilitating direct reinforcement learning (RL) training. With a total of 22 degrees of freedom and agility, this robotic hand moves with ease and speed, ready for any task.
With 94 haptic sensors on a single hand, Dex5 provides millimeter-precise feedback, enabling fine manipulation and gripping with unprecedented sensitivity. This sensory richness is key to developing advanced robotic applications that require interaction with a variety of objects and environments.
The manufacturer has integrated innovative components such as micro-force controlled composite gears and high-performance motors, ensuring reliability, compactness and long service life. With flexible ranges of motion for thumb joints (e.g. joint 0: -33.5°~39°, joint 1: 0°~100°) and other joints, the Dex5 offers a wide range of gripping strategies.
🤔 Do you struggle with limited dexterity and lack of feeling in robotic hands?
Traditional robotic hands often encounter limitations in dexterity, accuracy, and responsiveness to different surfaces and shapes of objects. Many lack sufficient feedback, which complicates the development of complex manipulation tasks and the effective training of artificial intelligence algorithms. The result is “stiff hands” that cannot fully utilize the potential of modern robotics.
✨ Discover the solution: Unitree Dex5 – The hand that thinks and feels
Unitree Dex5 represents a breakthrough in robotic manipulation. This hand was designed to eliminate the problems of stiffness and lack of feedback. It provides unprecedented dexterity and sensory richness, allowing robots to interact with the world with human-like finesse and precision. Take back control of manipulation and unlock the full potential of your humanoid projects.
💡 More than just features: Real benefits for your project
- 22 degrees of freedom: -> Benefit: Provides natural, smooth and extremely flexible manipulation of a wide range of objects.
- 94 tactile sensors: -> Benefit: Allows precise grip with detailed feedback about the pressure, shape and texture of the object.
- Independent finger exchange: -> Benefit: Simplifies maintenance, allows rapid adaptation to specific tasks and maximizes system modularity.
- Support for direct RL training: -> Benefit: Accelerates and streamlines the development of advanced machine learning algorithms for robotic manipulations.
- Micro-force-controlled gears: -> Benefit: They guarantee robust, quiet operation and fine control of force when interacting with the environment, including non-tactile proprioception.
🎯 For whom is Unitree Dex5 the ideal choice?
- For scientists and researchers in humanoid robotics and AI who are looking for advanced manipulation capabilities.
- For developers of robotic systems and prosthetics who need a precise, sensitive and adaptable robotic hand.
- For academic institutions and universities, as a cutting-edge tool for teaching and experiments in advanced robotics.
❓ Are you worried? We will answer your questions
- Concern: Is integrating such a complex hand difficult? -> Answer: Unitree Dex5 is designed with ease of integration in mind, with open interfaces and rich documentation for developers, making it easy to integrate into existing systems.
- Concern: What is the reliability and durability of the components? -> Answer: The hand uses high-quality components, such as high-power density motors and durable micro-force-controlled gears, ensuring long-term reliability and minimal maintenance.
- Concern: What is the level of grip accuracy and sensitivity? -> Answer: Thanks to 94 tactile sensors and precise gears, the hand achieves millimeter accuracy and the ability to delicately manipulate objects of various shapes and materials.
📊 Technical specifications
| Number of degrees of freedom | 22 |
| Number of tactile sensors (total) | 94 |
| Dimensions (L x W x H) | 217.3mm x 127.5mm x 72.1mm |
| Max. load capacity (palm down, grip 5cm object) | 3.5 kg |
| Max. load capacity (palm to the left, grip 5cm object) | 4.5 kg |
| Thumb joint range of motion 0 | -33.5° ~ 39° |
| Range of motion of thumb joint 1 | 0° ~ 100° |
| Range of motion of thumb joint 2 | 0° ~ 110° |
| Thumb joint range of motion 3 | 0° ~ 92° |
| Range of motion of joint 3 (other fingers) | 0° ~ 81° (connected to joint 2) |
| Number of field sensors (fingers) | 12 (total 94 pressure sensors) |
| Transmission mechanism | 12 micro-force controlled composite articulated gears, 4 micro-force controlled gear gears |
| Maximum receiving capacity (without damage) | 1270 apartments/package |
| Available feedback data | Joint mode, position, speed, torque, temperature, voltage, current, IMU data, sensor pressure, sensor temperature, joint stiffness/damping coefficient |
❓ Frequently Asked Questions (FAQ)
- What is the Unitree Dex5 robotic arm primarily intended for?
- It is intended for advanced applications in humanoid robotics, research, development and situations requiring precise and sensitive manipulation of objects.
- What is the number of degrees of freedom and how does it affect agility?
- Unitree Dex5 offers 22 degrees of freedom, ensuring exceptional flexibility and the ability to perform a wide range of complex and fluid manipulation tasks, with dexterity close to the human hand.
- How do tactile sensors work and what are their advantages?
- The hand is equipped with 94 tactile sensors that provide detailed feedback on the pressure, shape and surface of the object being grasped. This allows the robot to “feel” and adapt its grip with millimeter precision, which is key for fine manipulation.
- Is it possible to replace the fingers on the Unitree Dex5?
- Yes, all five fingers are designed to be independently interchangeable. This simplifies maintenance, repairs, and allows for quick configuration of the hand for specific tasks or experiments.
- What are the programming and integration options with other systems?
- Unitree Dex5 is designed with developers in mind. It supports secondary algorithm development, including reinforcement learning (RL) training, and provides rich feedback data for easy integration into existing robotics frameworks.
- What is the maximum load capacity of the arm?
- The maximum load capacity varies depending on the orientation. With the palm facing down and gripping a 5cm round hard object, the load capacity is 3.5 kg. With the palm facing left, the load capacity is 4.5 kg.
- What types of data does the hand provide for feedback?
- It provides comprehensive data, including joint mode, position, velocity, torque, temperature, voltage, current, IMU data, pressure and temperature sensors, and joint stiffness and damping coefficients, which is essential for advanced control.