Resources
synthetic-data
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.
digital-twin
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.
physics
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.
collision
Best practices for collision meshes in robotics sim
The collision mesh is the single most-tuned property in robotics simulation. Wrong choice = phantom interpenetration, slow physics, or both. Here's how to pick between primitives, convex hull, V-HACD decomposition, and mesh simplification, with concrete defaults per asset class.
openusd
How to convert GLTF/FBX/OBJ to OpenUSD for simulation
Most 3D pipelines export to GLTF, FBX, or OBJ. Most simulators expect OpenUSD. This guide covers the conversion paths, Omniverse, Blender, command-line, and AI-automated, plus what each format preserves and what gets lost.
sim-to-real
Sim-to-real transfer: why physics accuracy matters more than visual fidelity
The sim-to-real gap is overwhelmingly a physics problem, not a rendering problem. Here's the research, the failure modes, the calibration ranges that actually matter, and why investing in physics accuracy beats investing in photorealism for most robotics policies.
isaac-sim
Isaac Sim asset requirements and best practices
NVIDIA Isaac Sim expects assets in a specific format: OpenUSD with USDPhysics schemas, calibrated mass and inertia, convex collision meshes, semantic labels, and validated material bindings. This is the practical checklist.
simready
How to create SimReady assets without manual modeling
Building a SimReady asset by hand takes ~4 hours and requires Blender, V-HACD, and USDPhysics expertise. AI-driven asset preparation collapses the workflow to about 5 minutes per asset. Here's how the automated pipeline works and when to use it.
openusd
What is OpenUSD and why does it matter for robotics
OpenUSD is the 3D scene description format Pixar built for film and NVIDIA scaled for robotics. This guide explains the architecture, the USDPhysics schemas, and why USD beats SDF/URDF/MJCF for modern simulation pipelines.
physics
How to add physics properties to 3D models for simulation
Adding physics, mass, friction, collision meshes, inertia, is the bottleneck in robotics simulation. Here's the manual workflow, the AI-automated alternative, and a step-by-step guide for both.