Didin Robotics and Moovin to Integrate Human Motion Data into Humanoids

Didin Robotics, an industrial physical AI (artificial intelligence) company, and Moovin, a human motion AI firm, are joining forces to build and utilize datasets essential for humanoid learning. By integrating Moovin's LiDAR-based motion capture technology and multimodal data collection capabilities with Didin Robotics' expertise in humanoid development, the partnership is expected to significantly improve the performance of robot motion learning.

On July 15, a representative from Didin Robotics stated, "The two companies have signed a memorandum of understanding (MOU) to build and utilize datasets necessary for humanoid learning."

In order to enhance the performance of humanoids, it is crucial to have precise data on how real people move through complex environments, maintain their balance, and interact with surrounding objects. Existing motion data was often created for content production or animation purposes, so it poses limitations when directly applied to the robot’s joint structures, motion ranges, and physical constraints.

To address this, the two companies plan to establish a joint data pipeline that captures various real-world human movements as high-precision 3D motion and multimodal sensor data and then converts it into a form optimized for humanoid robot learning.

Moovin will collect LiDAR and camera data from multiple viewpoints and provide it retargeted to fit the robotic structures of Didin Robotics. Didin Robotics will use these datasets to improve the efficiency and precision of learning key behavior models for humanoids, including locomotion, balance maintenance, full-body control, and interactions with objects and environments.

With the signing of this MOU, both companies plan to expand their scope of collaboration to include joint research and business opportunities. Choi Byeolyi, CEO of Moovin, stated, "Through this partnership, I hope that Moovin's motion data technology will grow beyond the content industry to become a core data infrastructure for the robotics sector."

Kim Junha, CEO of Didin Robotics, added, "To enhance humanoid performance, high-quality multimodal data that captures both human movement and information about surrounding spaces is essential. Through collaboration with Moovin, we will secure data optimized for robotic hardware and strengthen learning frameworks that can be applied directly to real robot development."

This content was produced with the assistance of AI translation services.