SimReady assets for
industrial automation
Train and validate factory robots against simulations of your specific production floor: fixturing, conveyors, parts bins, custom tooling. Rigyd converts 3D, images, or text into physics-enabled OpenUSD and MJCF in minutes per asset.
The problem
Why existing workflows fall short.
Generic asset libraries do not cover the factory floor
Off-the-shelf SimReady libraries cover generic objects (pallets, crates, basic furniture). Industrial automation depends on the specific parts, fixtures, and tooling running through a real production cell, none of which curated catalogs contain.
CAD-to-sim conversion loses physics
Factory parts originate in CAD (STEP, IGES, .glb exports). Converting to a simulator-ready format usually drops mass, friction, and collision data, leaving physics to be re-authored manually for every SKU.
Catalogs scale faster than engineers do
A medium-sized production line catalogs 5,000 to 50,000 distinct parts and fixtures. Manual SimReady authoring at ~4 engineer-hours per asset is months of work for one robot cell, before any iteration on the policy itself.
How Rigyd helps
AI-native infrastructure that automates the hard parts.
Convert CAD-derived geometry directly
Rigyd ingests .glb, .fbx, and .obj exports from any CAD tool, plus reference images and text descriptions for parts where geometry is not yet authored. Output is validated OpenUSD with USDPhysics applied, ready for Isaac Sim, Omniverse, or Gazebo Sim via USD imports.
Physics estimated per part
Mass is computed from volume × material density (steel, aluminium, ABS, glass, rubber). Friction is keyed to surface material. Inertia is derived consistently from the mass distribution. Calibration accuracy is inside the domain-randomization variance bands production training pipelines already wrap around.
Catalog-scale API ingestion
Enterprise API processes catalogs of tens of thousands of SKUs in batched pipelines. New parts entering ERP or PLM flow into simulation as quickly as the geometry is exported, so the sim catalog tracks the real production floor.
per part, from CAD-derived input to SimReady asset
SKUs handled per catalog-scale ingestion
estimated saving per 1,000-asset production-cell project
Build production-cell simulations that match your real factory
Drop in CAD-derived 3D models, images, or text descriptions and get SimReady assets in minutes per part.
Starts at $29/month. 30 credits included.
Frequently asked questions
Does Rigyd handle CAD formats directly?
Rigyd ingests .glb, .fbx, and .obj. Most CAD tools (SolidWorks, Fusion 360, Inventor, Creo) export to these formats directly or via Onshape, OpenCascade, or FreeCAD conversion. STEP and IGES are not ingested directly today; the standard pattern is to convert via the CAD tool's native exporter or a STEP-to-glb converter. The geometry, materials, and assembly hierarchy carry through; physics is added by Rigyd in the next step.
How accurate is the estimated physics for industrial parts?
Mass calibrates within 15-20% of measured values for known industrial materials (steel, aluminium, ABS, glass, rubber). Friction coefficients land within ±0.1 of measured material values. Both bands are inside the domain-randomization variance most production training pipelines already wrap around, so policies trained on Rigyd assets transfer to real production at parity with policies trained on hand-tuned assets. Override any value when you have catalog masses or lab-measured friction.
Can I keep my real catalog of parts and sim catalog in sync?
Yes. Enterprise API integrates into CAD pipelines and PLM systems so a new part entering ERP or PLM triggers a Rigyd job that emits the SimReady OpenUSD asset, references it into the simulation scene, and version-tags both the source CAD revision and the generated SimReady asset. Drift between the real and sim catalog stops being a manual housekeeping problem.
Which simulators does the output work in for industrial robots?
OpenUSD output drops into NVIDIA Isaac Sim and Omniverse for visual-rich training, Gazebo Sim via USD imports for ROS 2 stacks, and Unreal Engine and Unity for rendering-heavy workflows. MJCF output drops into MuJoCo and Genesis for fast contact-rich RL training. The same source asset works across all simulators without re-authoring physics.
Does this work for fixtures and end-effectors, not just parts?
Yes. Rigyd produces physics for any 3D geometry: parts, fixtures, end-effectors, conveyors, jigs, and inspection rigs all use the same pipeline. Articulated tooling (parallel grippers, clamps, slides) needs the articulation hierarchy described in URDF or MJCF; Rigyd outputs MJCF natively for this case and OpenUSD with PhysicsArticulationRootAPI when the downstream stack consumes USD. Joint stiffness, damping, and limits are configurable per articulation.
Related reading
In-depth guides on robotics simulation, OpenUSD, and SimReady assets.
21 MAY 2026
Synthetic data generation for computer vision in robotics
Real-world labeled robotics data is expensive and slow to capture. Synthetic data is fast and unlimited, but only useful if the underlying simulation has correct physics, semantic labels, and domain randomization. Here's the complete pipeline.
18 MAY 2026
Digital twin creation pipeline for manufacturing
A factory digital twin needs every object to behave physically, not just render. This is the end-to-end pipeline: CAD intake, BIM merge, physics layer, semantic labeling, simulation runtime, at the asset volumes (10K+ unique SKUs) real factories actually contain.
13 MAY 2026
How to set up mass, friction, and joint properties for robot training
The three pillars of robot physics setup, mass, friction, joints, determine whether your trained policy transfers to real hardware. Here's the calibration target for each, the schemas, and the common mistakes that quietly break training.
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