MOVIN raises $7.5M to turn human motion into robot training data

The South Korean startup's markerless motion-capture system is already used by Boston Dynamics and the RAI Institute for robot teleoperation.

MOVIN, a South Korean startup that turns human movement into real-time 3D motion data, has raised about $7.5 million in a Series A round led by IMM Investment, taking its total funding past $10 million. IBK Industrial Bank of Korea and Voyager Ventures joined as new investors, while Atinum Investment, NAVER D2SF and Schmidt followed on from earlier rounds.

The company, founded in 2023 by four doctoral candidates from KAIST, builds MOVIN TRACIN, a system that pairs LiDAR sensors with on-device AI to capture full-body motion in real time without suits, markers or a studio. MOVIN said it generated more than $750,000 in revenue within a year of launch and has sold more than 200 units across more than 20 countries, with North America its largest market and overseas sales accounting for about 70 percent of the total.

The robotics connection is the reason the round matters beyond motion capture. Boston Dynamics and the Robotics and AI Institute use MOVIN's technology for robot teleoperation and training, and MOVIN says it is the only South Korean launching adoption partner for NVIDIA SOMA. It also supplies a robotics training version of the system, MOVIN TRACIN R1, to partners.

The new money will fund global expansion, on-device multimodal AI, and a new research center in South Korea of roughly 330 square meters, focused on collecting real-world datasets for world-model development. The round follows a seed investment from NAVER D2SF in 2023 and a 4 billion won pre-Series A round in March 2025.

Motion data has become a bottleneck for physical AI: robots learn by watching humans move, and clean, precise movement data is expensive to collect. MOVIN's bet is that a cheap, portable capture device can supply that data at a scale studios and laboratories cannot match on their own.