MOVIN Raises 10 Billion Won Series A for Physical AI Motion Data

Markerless capture, now moving into robotics

Founded in 2023, MOVIN converts body motion into precise 3D data in real time. Its main product, MOVIN TRACIN, combines a LiDAR sensor with in-house computer vision algorithms, so performers need no dedicated suit or markers. According to the company, this lowers setup complexity and cost. It has supplied mainly global game, film and virtual human studios, and 70% of its revenue came from overseas within a year of its official launch.

The technology is also reaching robotics. The company says Boston Dynamics and the RAI Institute have applied it to robot teleoperation and behavior learning, and a dedicated solution, MOVIN TRACIN R1, has been deployed on site. At the 2026 NAB Show in the US, MOVIN also unveiled MOVIN TRACIN Stage, a multi-person tracking solution, and demonstrated integration with media platforms Disguise and Notch.

Where the money goes

MOVIN plans to build a motion R&D center of about 100 pyeong (roughly 330 square meters) and use it to collect and process large volumes of multimodal motion data in real physical environments for training physical AI world models. The strategy is to move beyond single-person tools toward industrial motion data infrastructure covering ultra-light on-device AI and large-space, multi-person settings.

Kim Hong-chan of IMM Investment said the firm valued the company's proven, stable revenue base in content production along with concrete robotics reference deployments. CEO Choi Byul-yi said generative AI is evolving beyond text and 2D images toward physical AI that understands real-world movement, and that MOVIN aims to build the 3D motion data infrastructure that industrial robots and autonomous control systems need.