ROSpider is an all-terrain ROS hexapod platform featuring a bio-mimetic structure. Its chassis integrates eighteen 35KG high-voltage bus servos, ensuring exceptional flexibility and adaptability. Powered by advanced bio-kinematic gait algorithms, it performs dynamic gait combinations and posture adjustments to navigate narrow spaces and avoid obstacles with ease.
Powered by Jetson/Raspberry Pi, ROSpider is equipped with LiDAR, a 3D depth camera, a 6-mic array, a 6DOF robot arm, and an OLED display. This hardware suite enables motion control, SLAM mapping & navigation, path planning, obstacle avoidance, 3D grasping, autonomous transport, motion sensing, voice interaction, and sound source localization.
ROSpider integrates large AI models for vision, voice, and text, empowering it to perceive its surroundings, plan actions, and execute tasks flexibly. Whether through online APIs or local deployment, the system combines data from LiDAR, camera, and microphones to unlock advanced Embodied AI tasks—including semantic understanding, emotion perception, natural voice interaction, scene understanding, target tracking, and AI-driven navigation.

ROSpider supports voice command interaction, utilizing SLAM technology to adaptively plan paths and navigate to designated areas. Combined with the large vision model, it identifies and tracks objects within its field of view in real-time, executing precise grasping and transport tasks via high-precision PID control.

Powered by inverse kinematic gait algorithms, ROSpider navigates stably through complex terrains such as rubble and narrow passages. Fusing SLAM navigation with AI visual perception and a high-precision 6DOF robot arm, it enables rapid identification, grasping, and transport of objects.