---
title: "SimReady assets"
description: "Simulation-grade assets with colliders, mass, inertia, friction, PBR materials and joints, made from photos, scans, CAD or a plain-language description, for NVIDIA Isaac Sim and MuJoCo."
url: https://matrixsim.ai/solutions/simready-assets
---

# SimReady assets, made autonomously.

Objects that look, collide and move like the real thing. Start from photos, a scan or CAD, or just describe the object in plain language. Validated, and ready to drop into Isaac Sim or MuJoCo.

[Book a call](https://cal.com/madhav-chawla/30min)[See examples](#catalog)

Visual

Collision

01 / The problem

## Most 3D models are not built to be simulated.

1. **They look right but behave wrong**  
Meshes from scans and asset stores have no mass, no friction and colliders that do not match the surface. Policies learn the wrong physics.
2. **Fixing them by hand does not scale**  
Simulation engineers spend days per object fitting colliders, tuning mass and rigging joints. A useful training set needs thousands.
3. **The gap shows up on deployment**  
Every shortcut in an asset becomes part of the sim-to-real gap, discovered when the robot meets the real object.

02 / How it works

## From object to validated asset, without the hand work.

Our pipeline does the work a simulation engineer would do, autonomously, and checks the result before it ships.

1. _Real_  
### Input  
Photos, a scan, CAD, or a description in plain language.
2. _Sim_  
### Reconstruct  
Clean geometry at real-world scale with PBR materials.
3. _Sim_  
### Add physics  
Colliders, mass, inertia, friction and joints.
4. _Sim_  
### Validate  
Automated checks and physics tests in simulation.
5. _Sim_  
### Export  
OpenUSD for Isaac Sim, MJCF for MuJoCo.

03 / What you get

## Everything a simulator needs.

Each asset is a complete, physically described object, not just a mesh. Seven layers, all checked before it ships.

* **Visual geometry**  
Clean, watertight mesh at real-world scale with sensible level of detail.
* **Materials**  
Physically based materials (PBR) so sensors and cameras see what they would see in the world.
* **Colliders**  
Collision geometry fitted to the object, so contact happens where the surface is.
* **Mass and inertia**  
Physical properties derived from the object's material and volume.
* **Friction and restitution**  
Contact parameters that make grasping, pushing and sliding behave.
* **Joints**  
Hinges and sliders with axes and limits for doors, drawers, lids and handles.
* **Metadata**  
Category, dimensions and provenance, so assets are searchable and reproducible.
* **Validated**
  * Watertight visual mesh
  * Collision geometry
  * Mass and inertia
  * Joints and limits
  * Rest and contact test

04 / Examples

## From the catalog.

Assets our pipeline has built, as a simulator sees them. Grab a part to work its joint, and switch lens to see the collision geometry and where the mass sits.

01 / 10

### Bench vise

IND-003 · Bench vise

What a camera sees

Grab a part to move it · Drag to turn

1. 1Swivel0° / 65°
2. 2Jaw (screw drive)0 mm / 90 mm

#### Physical

Mass

11.9 kg derived

Envelope

15 × 30.5 × 16.2 cm

Base

Fixed in place

#### Materials

* Carbon / mild steel
* Grey cast iron

#### Simulation model

Rigid bodies

4

Colliders

22 convex

Joints

2 hinge + 1 slider

Visual meshes

11

05 / Platform and catalog

## Make your own, or start from ours.

Upload your objects, or describe them in plain language, and get SimReady assets back. Or browse the catalog for ones we have already made. We are onboarding early teams now.

[Or get a demo ](https://cal.com/madhav-chawla/30min)

MATRIXCatalogMy assetsUpload

Two-drawer steel cabinet, 70 cm tallGenerate

Catalog10 assets

* Bench vise  
11.9 kg3 joints
* Microwave oven  
11.2 kg5 joints
* Manual pallet jack  
48.1 kg10 joints
* Cordless drill  
975 g6 joints
* Chest of drawers  
17.4 kg3 joints
* Digital caliper  
203 g2 joints
* F-clamp  
1.3 kg3 joints
* Open Euro storage tote  
2.7 kgRigid

06 / Compatibility

## Works where you already simulate.

OpenUSD + UsdPhysics**NVIDIA Isaac Sim and Isaac Lab**USD stages with physics schemas, colliders and articulation roots, ready to reference into your scenes.

MJCF**MuJoCo**MJCF models with meshes, inertials, contact parameters and joints, ready to include in your models.

Supported**Rigid**Tools, containers, appliances, packaging and everyday objects.

Supported**Articulated**Doors, drawers, lids, hinges and sliders, with joint axes and limits.

In progress**Deformable**Cloth, cables and soft objects. Working with early partners now.

07 / Questions

## Frequently asked.

What do you need from us?

Whatever you have for the object: photos, a scan, CAD, or simply a description in plain language. Tell us how the robot will interact with it and we will tell you what works best.

Can I describe an object instead of capturing it?

Yes. Describe it in natural language, for example a two-drawer steel cabinet, 70 cm tall, and the pipeline generates a SimReady asset from the description, with the same physics, joints and validation as a captured object.

Which simulators do you support?

NVIDIA Isaac Sim and Isaac Lab through OpenUSD with physics schemas, and MuJoCo through MJCF. Ask us if you need another target.

Can you handle articulated objects?

Yes. Doors, drawers, lids, hinges and sliders ship with joint axes and limits. Rigid and articulated objects are supported today.

What about deformables?

Cloth, cables and soft objects are in progress. If deformables matter for your use case, talk to us: we are working with early partners on them.

How do we get access to the platform and catalog?

Join the platform waitlist and we will let your team in as places open. If you need assets sooner, book a call and we will walk you through the platform and the catalog.

Talk to us

## See your objects as SimReady assets.

Book a call, tell us what your robot needs to handle, and we will show you the pipeline on objects like yours.

[Book a call](https://cal.com/madhav-chawla/30min)
