Chingmu Vision launched Project Decode, integrating optical, inertial and tactile sensing technologies to synchronously capture multimodal data including motion, video, tactile feedback, object information and environmental parameters in real-world scenarios. It enables enterprises and research institutions to efficiently build high-quality, reusable multimodal datasets for embodied intelligence, in scenarios including humanoid robot training, dexterous hand operation data acquisition, embodied large AI models, industrial teleoperation & safety monitoring, rehabilitation assistance and human‑robot collaboration research. Project Decode addresses the core bottlenecks in industry data collection and provides critical infrastructure-level support for the sector. Its submillimeter positioning with angular error ≤ 1° captures full-body micro-motion trajectories down to finger joints. The magnetometer-free tightly coupled design with active optical coded markers effectively resists visual occlusion and magnetic interference, ensuring continuous, stable and complete motion data even in complex operational environments. CHINGMU-MOCAP, en.chingmu.com, Booth 160
Edge computing gives AI agents the processing power to turn instructions into…
Daimon-TWM world model debuts for the public
Platform helps robots adapt to complex tasks
Look beyond the headlines for the latest in humanoid robots
