Rigyd
rigyd.comTurn images into simulation-ready assets in minutes
AI Toolsroboticssimulation3d-assetsopenusdphysicssynthetic-dataembodied-ai
About
Rigyd is an AI-native 3D infrastructure platform that converts raw 3D models, images, and text into physics-enabled OpenUSD and MJCF assets for robotics simulators like NVIDIA Isaac Sim, MuJoCo, Gazebo, and Unreal Engine. The platform automatically adds collision meshes, mass, friction, and semantic labels without requiring manual annotation or physics expertise. It targets robotics teams that need scalable, physics-accurate simulation data for training embodied AI systems.
Problem
Creating physics-accurate, simulation-ready 3D assets at scale is extremely time-consuming and expensive, creating a data bottleneck for robotics simulation and training.
For
Robotics engineers and AI teams building simulation data pipelines for embodied AI
How it works
Users upload 3D models, images, or text descriptions; the platform's AI generates or optimizes geometry, estimates physical properties, and outputs validated OpenUSD assets ready for major simulators.
Business model
freemium
Status
launched