Yahboom ROSMASTER M3 Prois ahighly integrated intelligent robot platformdesigned forROS education, scientific research, and AI application teaching. Equipped withMecanum wheels, apendulum suspension chassis, a6DOF robotic arm, and abinocular structured-light depth camera, it offersomnidirectional movement,visual recognition,3D grasping, andhigh-precision handling. Built on theROS2 Humble systemwithdual TOF LiDAR, it supportsstable SLAM mapping,autonomous navigation,LiDAR-based obstacle avoidance, andintelligent path planning, ensuring smooth and efficient robotic operation across complex environments.

At its core, theROSMASTER M3 Prodeeply integratesadvanced AI models, includingspeech recognition,natural language processing,voice command control, andmultimodal interactionthroughtext, image, and voice. It supportstask planning,environmental perception, anddynamic feedback reasoning, making it apowerful, scalable platformforAI education,robotics research, anduniversity-level development projects. This all-in-one solution accelerates innovations inrobotics, SLAM algorithms, AI learning, andintelligent automation.

Top-Level Hardware Configuration

• Raspberry Pi 5, Jetson NANO B01, Jetson ORIN NANO SUPER, Jetson ORIN NX SUPER development boards, four main control boards to choose from.
• An aluminum chassis, 80mm Mecanum wheels, and a rear pendulum suspension structure enable easy navigation across varied terrains.
• In combination with a depth camera and a 6DOF robotic arm, enabling 3D grasping, precise handling, and MoveIt simulation. Four TOF ranging LiDARs front left and rear right provide 360 scanning, improving mapping and navigation accuracy.

Fully Integrates Intelligent Perception and Semantic Understanding

• Built-in speech recognition and natural language processing, combined with speakers, enable easy voice commands and Q&A interactions.
• Multimodal interaction capabilities, including text, image, and voice, allow real-time action adjustments based on environmental changes, supporting free dialogue interruptions and dynamic feedback reasoning.
• Integration of a large model and an extensible RAG knowledge system enhances task awareness and complex problem-solving capabilities.

A complete robotics platform for teaching and research

• Based on ROS2 Humble, compatible with a variety of robotics algorithms and AI course content.
• Integrated hardware and software, with comprehensive documentation and teaching resources, enabling efficient progress from beginner to scientific experiments.
• From AI and SLAM to visual recognition and robot controlwidely applicable to university teaching, research projects, and development of robotics competition platforms.