Self-developed light adaptive technology to enable the robot to work in direct light, dim light, and no light scenes with the same performance

Support pure vision or sensor fusion mode under the framework of the self-developed multi-sensor fusion with vision as the main input source, based on different scene operation requirements.

Based on visual perception, VSLAM algorithm, path planning algorithm and multi-dimensional fusion technology, the core capabilities of navigation, positioning, map building, path planning and obstacle avoidance are realized.

Build a 3D semantic map combined with visual perception & cognitive technology to provide basic information support for semantic SLAM, intelligent obstacle avoidance, intelligent interaction, intelligent decision-making, etc.

Able to determine potential risks and make avoidance strategies in advance based on risk classification during the operation.

Adjust the operation strategy independently to optimize the robot's operation performance and efficiency, based on the dynamic operation requirements in the scene.

Accept natural semantic commands from users, design operation strategies independently, and complete personalized interactions.

Based on natural command control and a 3D semantic map, the robot can be summoned in real time to complete the personalized task of "target area + operation logic".

Based on natural command control as well as object recognition and stereo vision technology, the target following is operated, ensuring the tracking stability even in case of target position changes and the temporary loss in the visual field.