RealMan Unveils RealBOT Embodied Open Platform at IROS 2025: Next-Generation Human-Robot Collaboration

At the recently concluded IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2025), RealMan Robotics made a significant breakthrough with the launch of the RealBOT Embodied Open Platform. The company conducted a cross-regional teleoperation demonstration spanning 1,200 kilometers between Beijing and Hangzhou, showcasing the real-time collaboration capabilities of embodied intelligent systems.

Through a two-city setup connecting the Beijing Humanoid Robotics Data Training Center with the IROS exhibition booth in Hangzhou, a RealMan operator in Beijing remotely controlled humanoid robots at the booth, successfully performing complex interactive tasks such as handing towels and offering fruit.

Accelerating Embodied AI Innovation

RealBOT is a comprehensive open platform designed to accelerate embodied artificial intelligence innovation. Integrating advanced motion control, multi-dimensional perception, and precision manipulation capabilities, the platform enables researchers and developers to experiment with, prototype, and develop embodied AI applications.

Designed with a focus on high-quality data collection, RealBOT leverages over one million multi-modal data samples gathered from ten real-world application scenarios at the RealMan Data Training Center. This approach strengthens model training while accelerating cross-industry deployment.

Key Technical Features

Fully Integrated In-House Development: Proprietary actuator and control technologies deliver optimized performance and reliability.

Flexible Computing Infrastructure: Full compatibility with NVIDIA Jetson Orin and Digua RDK S100 platforms.

Multi-Sensor Fusion Perception: Integrates depth and wide-angle cameras, LiDAR, IMU, and microphone arrays in a unified system.

Compact Design for Confined Spaces: Superior mobility and precise operation capability in restricted environments.

High Dexterity: 21 active degrees of freedom (DOF) providing advanced hand manipulation and adjustable gripper support.

Open Ecosystem: Compatible with various mainstream vision systems and gripper models.

RealBOT’s ultra-lightweight architecture establishes a robust foundation for scalable humanoid systems and high-quality data generation. This enables embodied artificial intelligence technology to progress from laboratory settings to homes, factories, and service sectors worldwide.

Strengthening Global Collaboration in Embodied AI

The Beijing Humanoid Robotics Data Training Center, established by RealMan, forms the cornerstone of high-quality data production and AI training. Housing over 100 robots across ten different real-world environments, the center ensures continuous collection of large-scale multi-modal datasets supporting RealBOT and global research partners. This work creates an open, collaborative ecosystem for embodied intelligence.

For more information: www.realman-robotics.com