The Realman DAP-D-65 is a dual-arm embodied intelligence development platform positioned as a data-collection device for large embodied models. Its standard configuration includes two servo master arms, two RM65-B-V slave arms, three Intel RealSense D435 depth cameras, two grippers, and a fixed bracket. A master-slave workflow lets an operator guide the master arms while the slave arms follow, enabling synchronised bimanual manipulation and structured dataset capture for learning-based control. Realman reports that with imitation learning plus static-data training, 50 task demonstrations can raise task success to up to 90%.

Multi-camera sensing is central to the platform: RealSense D435 cameras handle both global monitoring and close-range task perception, providing depth and colour imagery across an effective range for everyday manipulation scenes and visual learning experiments. The system is framed as a combined teaching and research platform, supporting virtual simulation and real-scene experiments for robotics curricula such as robot operating systems and kinematics, as well as lab development and application prototyping. Secondary-development interfaces span C, C++, C#, Python, MATLAB, ROS, and JSON-based integration.