| Project | Parameters |
| Maximum Obstacle Crossing Height |
15cm |
| Jump/Roll ability |
Support front/back somersault, high jump, ditch jump |
|
Joint torque |
Increased by 50% (compared to previous generation) |
|
Maximum Load |
7.5kg |
|
Endurance |
90 minutes |
| Cruising Range | 5km |
| Control Frequency | 1 kHz |
| System Calculation Force | Increased by about 3 times |
| Aware Module Support | Lidar, depth camera, IMU |
| Communication Expansion | Supports 5G, RTK, edge computing devices |
| Development Interface | ROS + SDK/API |
| Human-computer Interaction | Lamp status prompt, low delay image transmission |
Lite3 has achieved a significant breakthrough in sports performance. It supports high-difficulty movements such as back somersault, front somersault, twist jump, space walk, turn over and climb up and greet. It can climb steps up to 15cm (close to the limit of quadruped robots of the same size) and has dynamic obstacle crossing capabilities such as gully jump and high jump. The power and endurance are fully enhanced, the joint torque is increased by 50%, and the self-developed, the limit walking load is 7.5kg (up 40%), the continuous exercise time is 90 minutes, and the cruising range is 5 kilometers, which is nearly double that of the previous generation. Equipped with industrial-grade control system, the calculation power is increased by nearly 3 times, equipped with high-stability industrial-grade IMU, the real-time control system of deeply optimized kernel is introduced for the first time, and the control frequency is as high as 1kHz, ensuring high real-time communication; Using open modular design, it supports flexible installation of RTK, 5G, AI host, edge processor, lidar, depth camera and other sensors, and provides advanced perception development interface (SDK/API); In terms of perception and interaction, laser radar and depth camera are integrated to realize autonomous navigation, automatic stop and detour, visual positioning and three-dimensional reconstruction of the environment, and new light status prompt and low delay first view image transmission are added, it also supports front and rear stopping obstacles, forward obstacle avoidance and human body recognition following (some functions depend on the model); At the same time, based on ROS system, it provides open source HDL mapping, positioning and navigation algorithm cases, and fully supports diversified application scenarios such as group control dance, research performance and real machine deployment.