Core Overview
Standing at 1.21 meters and weighing approximately 25 kg, the R1 Edu Pro D is an ultra-lightweight bipedal platform. Unlike the basic models, the U6 is equipped with 38 Degrees of Freedom (DOF), providing it with human-like dexterity and the mobility required for complex tasks like running, stair climbing, and even dynamic maneuvers like flips or “boxing” simulations.
Key Specifications & Features
| Feature | Specification |
| Total Degrees of Freedom | 38 (Legs: 12, Arms: 10, Waist: 2, Head: 2, Hands: 12) |
| Dexterous Hands | 2x BrianCo Revo 2 Touch (5-finger with tactile sensors) |
| AI Computing Power | NVIDIA Jetson Orin (100 TOPS) |
| Vision & Audio | Binocular Stereo Camera, 4-Mic Array, 5W Stereo Speaker |
| Battery & Runtime | 9000mAh Quick-Release Lithium (~1 hour endurance) |
| Connectivity | Wi-Fi 6, Bluetooth 5.2, Secondary Development SDK |
Why the “U6” Version?
The R1 series comes in several tiers (U1 through U6). The Edu Pro D (U6) stands out due to two critical upgrades:
-
Tactile-Enabled Manipulation: It features the BrianCo Revo 2 Touch hands. These are five-finger dexterous hands equipped with advanced tactile sensor arrays, allowing the robot to “feel” objects and perform precision tasks that require delicate force control.
-
Maximum Compute: It utilizes the highest available onboard AI module (100 TOPS), enabling it to run large multimodal models (LMMs) and complex computer vision algorithms locally without relying entirely on the cloud.
Performance & Build
-
Agile Mobility: Powered by low-inertia, high-speed internal rotor PMSM motors, the R1 can reach speeds up to 2.5 m/s. Its hollow-joint wiring and air-cooled design ensure the robot remains sleek and durable during intensive testing.
-
Research Ready: The “Edu” designation means it is fully programmable. It supports ROS, Python, and C++, and is compatible with simulation environments like NVIDIA Isaac Sim, making it a seamless bridge between digital algorithms and physical reality.
-
Smart Interaction: Integrated with a large multimodal language model, the U6 can process both voice and visual inputs simultaneously, allowing for natural, responsive interactions with its environment.
Note: As of early 2026, the Unitree R1 series is in a phased global rollout. Given its complexity, it is recommended for users with a background in robotics or AI research.










Reviews
There are no reviews yet.