Sim-ready 3D data for robotics

Sim-ready 3D assets at scale. Validate in sim; finally.

Turn images into photorealistic, physics-tagged, tested 3D assets, plus thousands of variations, on demand. Test against the maximum range of objects your robot might encounter. Image or text in, sim-ready 3D out.

▶ Open the Lab API docs Early access GitHub

Not just a mesh that looks right in a viewer

Every asset ships with the properties the physics engines need: mass, friction, collision mesh, scale, watertight geometry. Orbit a variant, grab a flap, drawer, door or caster, then drop it into your stack.

Cardboard box with articulated flaps in projectsim Lab
Boxes, cages and pallet stacks
Desk with sliding drawers in projectsim Lab
Desks, ovens and wardrobes
Pallet racking in projectsim Lab
Racking and gas cylinders
Bedroom scene in projectsim Lab
Full rooms with Gaussian-splat backdrops

Start here

Drop it into your stack
  1. Pick a class and variant in the Lab.
  2. Copy the download snippet from the side panel, or pull one variant yourself: 
    pip install -q huggingface_hub
python -c "from huggingface_hub import snapshot_download as d; \
d('projectsim/warehouse-manipulation-and-states', repo_type='dataset', \
  allow_patterns=['assets/logistics_cage/v01_*/*'], local_dir='.')"
  3. Load it. Isaac Sim, Isaac Lab, MuJoCo, Unreal Engine, etc. No manual cleanup step.