KOZA-KZY-M1 Embodied AI Education Robot (Mecanum)
KOZA-KZY-M1 is a Juze KZY-series embodied AI robot for ROS education and research. Its Mecanum-wheel chassis enables 360° omnidirectional motion, integrating an AI voice module, depth camera / HD camera gimbal, and TOF LiDAR for multi-robot sync, SLAM mapping and navigation, AI vision, and voice interaction. A dual-model architecture (decision + execution) supports barge-in dialogue and dynamic feedback, with optional RAG knowledge-base extension and mainstream large-language-model platforms.
Features
- Multimodal AI interaction: language / speech / vision fusion with barge-in dialogue, dynamic feedback, and scene understanding.
- Omnidirectional chassis: Mecanum wheels for 360° translation and rotation in classroom and lab spaces.
- AI + SLAM navigation: mapping, multi-point navigation, object search, and autonomous cruising; visual tracking and following.
- Depth sensing (Deluxe / Flagship): 3D structured-light depth camera for depth Q&A and finer scene understanding.
- Long-horizon tasks: dual-model decision + execution for multi-step embodied workflows.
- ROS2 development stack: Ubuntu + Docker + ROS2 Humble, with Rviz visualization and YOLOv11 detection.
- Multi-robot and teaching scenarios: multi-robot sync; smart-traffic / autonomous-driving labs (signs, lane keeping, parking, and more).
- Cross-platform control: PC, gamepad, and APP.
Key Specifications
| Model | KOZA-KZY-M1 |
|---|
| Series | KZY Embodied AI · Education Robot |
|---|
| Controller | Jetson Nano B01 4GB (approx. 0.5 TFLOPS AI) |
|---|
| OS & software | Ubuntu 18.04 LTS + Docker + ROS2 Humble; Python |
|---|
| Chassis | Mecanum omnidirectional; 1:56 high-torque 520 encoder metal motors ×4 |
|---|
| Body | Anodized sandblasted aluminum alloy |
|---|
| Power | 12.6V 6000mAh Li-ion pack; approx. 3 h runtime |
|---|
| Dimensions | Standard / Deluxe approx. 284.4×231.4×181.4 mm; Flagship approx. 284.4×231.4×188.8 mm |
|---|
| Weight | Approx. 2.68–2.74 kg |
|---|
| Vision & navigation | Rviz; YOLOv11; SLAM mapping & navigation; multi-robot formation |
|---|
Configurations
| Config | LiDAR | Vision | Notes |
|---|
| Standard | T-MINI PLUS | 200W HD camera gimbal (servo) | Entry teaching and ROS2 basics |
| Deluxe | T-MINI PLUS | Nuwa-HP60C 3D structured-light depth camera | Depth sensing, visual follow, scene understanding |
| Flagship | Slamtec C1 | Nuwa-HP60C 3D structured-light depth camera | Higher LiDAR performance; optional aluminum case |
Options
- Autonomous-driving teaching sandbox (approx. 3.2×2.8 m)
- Structured e-course materials (provided after purchase)
- Optional aluminum carrying case for Flagship
