
The uLearn K1 Geek is a high-performance bipedal platform that transforms complex robotics theory into an accessible, out-of-the-box experience. Engineered with 22 degrees of freedom and high-torque joints, it offers a robust environment for mastering Python-based control and bipedal locomotion.
Driven by a 48 TOPS Qualcomm core and a 3D depth camera, the Geek provides the edge-AI power necessary for autonomous vision tasks and real-world stability testing. It is the definitive hardware for students and makers to jumpstart their journey from basic coding to advanced embodied AI.
| Parameter |
Specification |
| Height |
~95 cm |
| Weight |
~19.5 kg |
| Total Degrees of Freedom |
22 |
| DoFs per Leg |
6 |
| DoFs per Arm |
4 |
| Head DoFs |
2 |
| Hip Joint Range |
P: −171°~126°, R: −22°~89°, Y: ±59° |
| Knee Joint Range |
0°~127° |
| Ankle Joint Range |
P: −50°~20°, R: ±20° |
| Max Peak Torque |
60 N·m |
| Joint Encoder |
Dual Encoder
|
|
GPU
|
48 TOPS (Dense) AI performance
|
|
Camera
|
Stereo Depth Camera
|
|
IMU
|
9-axis IMU
|
|
Audio
|
Circular 6-Mic Array + Speaker
|
|
Battery
|
2 Ah
|
|
Battery Life
|
30 min (walking 0.4 m/s)
|
|
WiFi
|
Supported
|
|
Bluetooth
|
5.2
|
|
Buttons
|
Power ×1, Interaction ×3
|
|
Expansion Port
|
Gigabit Ethernet
|
|
Firmware Upgrade
|
Supported
|
|
LLM
|
Integrates with all commercial LLMs
|
|
Secondary Development
|
Supported
|
|
Warranty
|
3 months
|

The uLearn K1 Edu is the gold standard for high schools and undergraduate labs, bridging the gap between basic coding and professional-grade AI research. While maintaining the agile bipedal form of the series, the Edu edition introduces a significant leap in computational intelligence, powered by the NVIDIA Jetson Orin NX platform.
This increased processing headroom allows students to move beyond pre-programmed movements into autonomous navigation and real-world SLAM (Simultaneous Localization and Mapping). With an extended battery life designed for full lab sessions and an integrated 3D sensor suite, the Edu is a future-ready tool for mastering interactive speech recognition and complex computer vision in a dedicated educational environment.
| Parameter |
Specification |
| Height |
~95 cm |
| Weight |
~19.5 kg |
| Total Degrees of Freedom |
22 |
| DoFs per Leg |
6 |
| DoFs per Arm |
4 |
| Head DoFs |
2 |
| Hip Joint Range |
P: −171°~126°, R: −22°~89°, Y: ±59° |
| Knee Joint Range |
0°~127° |
| Ankle Joint Range |
P: −50°~20°, R: ±20° |
| Max Peak Torque |
60 N·m |
| Joint Encoder |
Dual Encoder |
| GPU |
117 TOPS AI performance |
| Camera |
Stereo Depth Camera |
| IMU |
9-axis IMU |
| Audio |
Circular 6-Mic Array + Speaker |
| Battery |
5 Ah |
| Battery Life |
80 min (walking 0.4 m/s) |
| WiFi |
WiFi 6 |
| Bluetooth |
5.2 |
| Buttons |
Power ×1, Interaction ×3 |
| Expansion Port |
Gigabit Ethernet |
| Firmware Upgrade |
Supported |
| LLM |
Integrates with all commercial LLMs |
| Secondary Development |
Supported |
| Warranty |
1 year |

The uLearn K1 Pro is the premier choice for graduate-level research and professional algorithm development. Engineered to push the boundaries of humanoid capability, the Pro edition provides the massive computational headroom required to run high-fidelity simulations and complex Reinforcement Learning (RL) policies directly on the hardware.
With a top-tier 200 TOPS AI engine, this platform is optimized for deploying Large Language Models (LLMs) and advanced transformer-based vision networks without latency. It offers a seamless ROS 2 development workflow, allowing researchers to transition from synthetic environments to real-world physical performance with unmatched processing speed and sensor fusion accuracy.
| Parameter |
Specification |
| Height |
~95 cm |
| Weight |
~19.5 kg |
| Total Degrees of Freedom |
22 |
| DoFs per Leg |
6 |
| DoFs per Arm |
4 |
| Head DoFs |
2 |
| Hip Joint Range |
P: −171°~126°, R: −22°~89°, Y: ±59° |
| Knee Joint Range |
0°~127° |
| Ankle Joint Range |
P: −50°~20°, R: ±20° |
| Max Peak Torque |
60 N·m |
| Joint Encoder |
Dual Encoder |
| GPU |
200 TOPS AI performance |
| Camera |
Stereo Depth Camera |
| IMU |
9-axis IMU |
| Audio |
Circular 6-Mic Array + Speaker |
| Battery |
5 Ah |
| Battery Life |
80 min (walking 0.4 m/s) |
| WiFi |
WiFi 6 |
| Bluetooth |
5.2 |
| Buttons |
Power ×1, Interaction ×3 |
| Expansion Port |
Gigabit Ethernet |
| Firmware Upgrade |
Supported |
| LLM |
Integrates with all commercial LLMs |
| Secondary Development |
Supported |
| Warranty |
1 year |

The uLearn T1 is a sophisticated, mid-size developer platform designed for high-fidelity Human-Robot Interaction (HRI) and complex manipulation studies. Moving beyond the classroom, the T1 provides a human-scale presence, making it the definitive tool for researching social robotics, assistive technologies, and adult-scale gait dynamics.
The T1’s unique dual-compute architecture—combining an Intel i7 processor with an NVIDIA AGX Orin—enables the simultaneous management of high-level logic and intensive AI inference. This allows researchers to integrate extensible arm DoFs and optional grippers for dexterous task performance, all supported by a high-capacity battery built for extended, untethered interaction in real-world environments.
| Parameter |
Specification |
| Dimensions |
118 × 47 × 23 cm |
| Leg Length |
57 cm |
| Arm Length |
45 cm |
| Weight |
~30 kg |
| Total Degrees of Freedom |
23 |
| DoFs per Leg |
6 |
| Waist DoF |
1 |
| DoFs per Arm |
4 (extensible) |
| Head DoFs |
2 |
| Waist Joint Range |
±58° |
| Hip Joint Range |
P: ±118°, R: −21°~88°, Y: ±58° |
| Knee Joint Range |
0°~123° |
| Ankle Joint Range |
P: −50°~20°, R: ±25°
|
|
Max Peak Torque
|
130 N·m
|
|
Joint Encoder
|
Dual Encoder
|
|
CPU
|
Intel i7 1370p
|
|
GPU
|
NVIDIA AGX Orin, 200 TOPS AI performance
|
|
Camera
|
Depth Camera
|
|
IMU
|
9-axis IMU
|
|
Audio
|
Circular 6-Mic Array + Speaker
|
|
LLM
|
Integrates with all commercial LLMs
|
|
Battery
|
10.5 Ah
|
|
Battery Life
|
2 h (walking), 4 h (standing)
|
|
WiFi
|
WiFi 6
|
|
Bluetooth
|
5.2
|
|
Expansion Ports
|
USB, Ethernet
|
|
Firmware Upgrade
|
Supported
|
|
Secondary Development
|
Supported
|
|
Warranty
|
1 year
|

Pepper is the first humanoid robot capable of recognizing principal human emotions and adapting its behavior to the mood of its interlocutor. Whether in a hospital, a retail store, or a classroom, Pepper creates unforgettable human-robot interactions. Includes a year-long subscription of RoboHearts.AI software package.
| Parameter |
Specification |
| Height / Weight |
120 cm / 28 kg |
| DoF |
20 (17 joints + 3-wheel base) |
| Mobility |
3 omnidirectional wheels, 3-5 km/h |
| Processor |
Intel Atom E3845 quad-core |
| Tablet |
10.1" touchscreen (chest) |
| Cameras |
2x HD 5MP + 3D depth sensor |
| Mics |
4-mic array, 2 speakers |
| Touch |
5 capacitive (head + hands) |
| Nav Sensors |
2 sonars, 6 lasers, 3 bumpers, gyro |
| Battery |
30 Ah Li-Ion — ~12 hours |
| OS |
NAOqi 2.9 (Android) |
| Dev |
Android Studio, Java, Python, C++, ROS |
| Languages |
20+ supported |
| Emotion Engine |
Face + voice tone analysis |

Built to assist, not replace, Mirokai is a revolutionary service platform that combines high-precision dexterity with socially aware AI. While its animated, real-time expressive face builds instant trust with humans, its industrial-grade underpinnings—including a 28-DoF architecture and Multi-LLM support—allow it to handle complex verbal interactions and logistical tasks with ease.
By utilizing a patented ball-bot mobility system, Mirokai navigates crowded corridors with 360-degree fluidity, making it the ultimate tool for relieving staff from repetitive delivery duties. Its 8-fingered hands provide a human-like 97% grasp accuracy, ensuring that whether it is in a hospital, hotel, or retail space, the "human touch" remains at the forefront of the automated service.
| Parameter |
Specification |
| Height / Weight |
~123 cm / ~30 kg |
| DoF |
28 |
| Hands |
8 fingers, opposable thumbs — 97% grasp |
| Mobility |
Patented ball-bot, ~0.8 m/s |
| Navigation |
Autonomous VSLAM + social awareness |
| Face |
Animated real-time (projector + 3D engine) |
| Cameras |
2x RGBD, 2x infrared |
| Sensors |
9x ToF, 6x ultrasound, 8x torque, 3x IMU |
| AI |
Multi-LLM, speech, VLM, GDPR face tracking |
| Battery |
~4-8 hours |
| Use Cases |
Healthcare, hospitality, retail, and education |