MANUS, the global leader in high-precision data gloves for robotics and Physical AI, and Lightwheel, the leading data and simulation infrastructure company powering Physical AI at scale, today announced a strategic partnership to accelerate the development and deployment of dexterous robotic systems.
The partnership connects two foundational layers of the Physical AI data stack. MANUS contributes high-precision hand motion capture as the source signal that feeds the stack; Lightwheel's data infrastructure spans real-world human demonstration capture and large-scale simulation. Together, the companies deliver an end-to-end Real2Sim2Real pipeline that carries dexterous human demonstrations, captured at millimeter fidelity, through simulation, policy training, and real-world robot deployment.
The agreement, signed by MANUS Co-founder and CEO Stephan van den Brink and Lightwheel Co-founder and President Haibo Yang, establishes a formal framework for technical integration, joint go-to-market, and co-development across the Physical AI data lifecycle. Under the partnership, MANUS' high-precision hand capture integrates directly into Lightwheel's data infrastructure, forming a continuous link from real-world demonstration to large-scale simulation expansion.
At the heart of the collaboration is a shared conviction: the quality of robot behavior policies follows directly from the quality of the human demonstration data used to train them. MANUS' data gloves capture 25 degrees of freedom per hand at millimeter precision, anchoring Lightwheel's large-scale human demonstration collection with high-fidelity source data. Lightwheel then amplifies each demonstration by 100 to 1,000 times through synthetic data generation, letting robot makers train on diverse, physics-accurate scenarios at a fraction of real-world collection cost.
The partnership also builds on a shared foundation in the NVIDIA Isaac ecosystem. As a member of the Newton Technical Steering Committee, Lightwheel defines the SimReady standard, and its assets sit at the center of the Isaac Sim workflow. MANUS is the official data glove of NVIDIA Isaac Teleop, the unified teleoperation framework within the Isaac ecosystem, making the two companies' technologies natively interoperable and letting robotics teams route MANUS-powered teleoperation directly into Lightwheel's data engine.
"The most important bottleneck in Physical AI today is the quality and scale of dexterous manipulation data. MANUS captures what the human hand actually does; Lightwheel scales that knowledge into the millions of training examples that robot policies require. Together, we are building the data infrastructure that the next generation of Physical AI will run on."
—Stephan van den Brink, Co-Founder and CEO, MANUS
"Lightwheel's mission is to close the sim-to-real gap at every layer of the data stack. High-fidelity human demonstration data is the starting point for everything downstream: synthetic scaling, VLA fine-tuning, policy evaluation. MANUS gives us the highest-quality source signal available anywhere, and this partnership is the natural next step in building the world's best Physical AI data pipeline."
—Steve Xie, Co-Founder and CEO, Lightwheel
Founded in 2014 and headquartered in Eindhoven, the Netherlands, MANUS is the global standard in high-precision hand tracking for robotics teleoperation, Physical AI training, VR/XR, and professional motion capture. MANUS gloves capture full hand articulation across 25 degrees of freedom per hand with millimeter accuracy, giving operators the fidelity to record dexterous manipulation and train generalizable robot policies. Trusted by more than 2,000 robotics teams and research institutions worldwide, MANUS delivers the high-fidelity source signal the next generation of Physical AI is built on.
Learn more: www.manus-meta.com
Founded in 2023 and headquartered in Santa Clara, California, Lightwheel is a Physical AI infrastructure company building the data engine for Physical AI at scale. Lightwheel's platform spans SimReady (physics-accurate simulation assets), EgoSuite (egocentric human demonstration data), and RoboFinals (industrial-grade policy evaluation), enabling robotics teams to train, scale, and validate robot policies efficiently. Lightwheel's SimReady assets are featured in NVIDIA Isaac Sim, and its Real2Sim2Real workflow is deployed by world-leading robotics companies.
Learn more: www.lightwheel.ai
