Spatial adaptability
3D spatial understanding and autonomous mobility bring LUMO toward users, wherever companionship is needed.

DOBOT · Embodied AI
Understanding the world. Connecting with you.
An embodied AI humanoid for companionship, research and commercial interaction.
1.24 m
Height
30
Degrees of freedom
130 N·m
Maximum joint torque
≈ 2 h
Typical motion runtime
Four dimensions of intelligence
3D spatial understanding and autonomous mobility bring LUMO toward users, wherever companionship is needed.
An everyday companion, children's playmate or educational partner, adapting to different social roles.
Vision and voice perception help recognize emotions and intentions for more natural interaction.
Real-world interactions and long-term memory help LUMO adapt to preferences and habits over time.
Features and capabilities are subject to official release.
Designed for interaction

Interactive display, a 5-zone capacitive head touch sensor, and 15+ facial expressions combine with head movements and body language.
Expressive interaction

Happy
15+ facial expressions
Full-stack embodied AI
DobotWAM integrates 3D spatial understanding, robotic motion constraints and real-world data feedback loops, going beyond vision-language-action modelling.

Vision and voice fusion recognize emotions. 3D semantic maps help identify people, movement and objects.
Autonomous learning and long-term memory adapt interactions to preferences and personal characteristics.
Data is stored locally by default. Cloud uploads or other uses require consent; users control personal data management.
Applications

An embodied AI platform for K–12, vocational education, universities and research. Explore motion control, embodied AI algorithms, SLAM, visual perception and human–robot interaction through secondary development.
Interactive play, learning and conversations with family members. First-person interaction experiences support remote connection with loved ones and pets.
Welcome visitors, guide routes, introduce exhibits and answer questions, with interactive experiences for exhibitions and events.
Dancing, handshakes, interactive games and engaging performances for audiences. High-risk movements require professional assessment before activation.
Product launches, exhibitions, shopping mall events and promotional activities that create memorable interactive brand experiences.
Choose your edition
NVIDIA Jetson Thor T4000
1,200 TFLOPS (FP4) · Stereo RGB-D camera · 3D LiDAR
| Specification | Standard | Max |
|---|---|---|
| Dimensions (H × W × D) | 1240 × 560 × 244 mm | 1240 × 560 × 244 mm |
| Weight with battery | ≈ 40.5 kg | ≈ 41 kg |
| Upper + lower leg length | 635 mm | 635 mm |
| Single arm length (excluding hand) | 596 mm | 596 mm |
| Total degrees of freedom | 30 | 30 |
| Head / waist / arm / leg DOF | 1 / 3 / 7 / 6 | 1 / 3 / 7 / 6 |
| Operating temperature | 0–40 °C | 0–40 °C |
| Operating time (typical motion) | ≈ 2 h | ≈ 2 h |
| Joint maximum torque | 130 N·m | 130 N·m |
| Single arm payload (without gripper) | 1.5 kg | 1.5 kg |
| Stereo RGB-D camera | — | Included |
| 3D LiDAR | — | Included |
| Head touch sensor | 5-zone capacitive | 5-zone capacitive |
| Audio | 4-microphone array + speaker | 4-microphone array + speaker |
| Display | Interactive display + RGB light strip | Interactive display + RGB light strip |
| Connectivity | Wi-Fi 6 (STA/AP), BLE 5.2 | Wi-Fi 6 (STA/AP), BLE 5.2 |
| AI computing | Rockchip RK3588 · 6 TOPS | NVIDIA Jetson Thor T4000 · 1,200 TFLOPS (FP4) |
Prototype hand is configuration dependent. Approximate operating time is measured under typical motion conditions. Confirm the final configuration and officially released capabilities when requesting a quote.
LUMO in focus
Get in touch
Talk to our integration experts about deploying AI platforms, quantum solutions, robotics, or educational labs across your organization.