Simulation assets for
warehouse robots
A typical warehouse has 50,000+ SKUs. Rigyd converts your product catalog into physically accurate simulation objects with mass, friction, and collision geometry, ready for pick-and-place training.
The problem
Why existing workflows fall short.
Warehouses have infinite object variety
Every box, bottle, bag, and pallet has different weight, friction, and geometry. Manually modeling physics for 50,000+ SKUs is impossible.
Pick-and-place needs accurate physics
Grasping policies fail in the real world when training objects have wrong mass or friction. Physics accuracy directly impacts sim-to-real transfer.
Seasonal inventory changes constantly
New products arrive weekly. Your simulation environment needs to keep pace with actual warehouse inventory.
How Rigyd helps
AI-native infrastructure that automates the hard parts.
Bulk catalog conversion
Upload your entire product 3D catalog. Rigyd adds physics properties to every object, mass from material density, friction from surface identification, collision meshes from geometry.
Calibrated for manipulation
Physics estimates are tuned for robotic grasping and stacking scenarios. Mass accuracy within 15-20% of measured values, well within domain randomization ranges.
API for continuous updates
Enterprise API integrates with your product pipeline. New SKUs get physics annotations automatically as they enter your catalog.
typical SKUs in a single warehouse
saved per 1,000-object simulation setup
cost reduction vs manual physics annotation
Build warehouse simulations at catalog scale
Convert your product catalog from 3D models, images, or text into SimReady assets in minutes.
Frequently asked questions
Can Rigyd handle a full 50,000+ SKU warehouse catalog?
Yes. Enterprise API processes entire product catalogs in bulk. Each SKU gets physics properties, collision meshes, and semantic labels automatically, enabling simulation at real warehouse scale.
How accurate are the physics estimates for pick-and-place training?
Mass accuracy within 15-20% and friction within 0.1 coefficient, well inside typical domain randomization variance. This provides realistic baselines that improve sim-to-real transfer by ~40% per NVIDIA GR00T N1 benchmarks.
Does the warehouse simulation update when new SKUs arrive?
Yes. Enterprise API integrates with your product pipeline, so new SKUs get physics annotations automatically as they enter your catalog. Your simulation stays aligned with live inventory.
How does Rigyd handle SKU variants (sizes, colors, packaging)?
Rigyd treats each unique geometry as a separate asset, with USD references reusing the master mesh across instances when geometry is identical. Size or packaging variants that share geometry but differ in mass (e.g., empty vs full box) can be expressed as USD variants, one file, multiple physics overrides. Colors and labels carry through via material bindings without re-running geometry analysis.
What's the cost difference between manual SKU modeling and Rigyd at warehouse scale?
Manual SimReady authoring is ~4 engineer-hours per asset. For 50,000 SKUs that's 200,000 hours, roughly $18M at a $90/hr blended rate. Rigyd processes the same catalog in ~2,000 compute-hours (parallelizable to days, not years), at a fraction of a percent of the manual cost. The structural unlock isn't the savings, it's that simulation diversity stops being gated by team headcount.
Related reading
In-depth guides on robotics simulation, OpenUSD, and SimReady assets.
9 JUL 2026
Scaling Simulation Asset Libraries Beyond Curated Inventory
Curated SimReady libraries are a great starting point but a hard ceiling. Here is why fixed inventories limit robotics simulation at scale, and how on-demand asset generation closes the gap.
2 JUL 2026
Domain Randomization for Robotics Training: Asset Diversity at Scale
Domain randomization makes trained policies transfer to the real world, but it only works if your asset library is diverse enough. Here is how asset diversity drives randomization, and how to generate it at the scale training needs.
30 JUN 2026
Building a Scalable Embodied AI Asset Pipeline: From Raw Data to Simulation
A practical look at the stages of an embodied AI asset pipeline, from raw 3D data and reference inputs to physics-ready simulation assets, and what changes when you need to produce thousands of them instead of a handful.
Explore more
SimReady for Isaac Sim
OpenUSD assets validated for NVIDIA Isaac Sim
Models for MuJoCo
Physically accurate assets for robot learning
Assets for Gazebo
Simulation-ready models for ROS 2
Warehouse Simulation
50,000+ SKUs need physics properties
Humanoid Robots
Whole-body interaction training assets
Sim-to-Real Transfer
Close the gap with accurate physics data