Skip to content

Custom digital twins
for robotics simulation.

Bring your facility into simulation with USD environments and interactive assets built around your robot’s tasks.

Warehouse environment with industrial storage racks and pallets
Warehouse 16Library environment

Explore the environments
we author.

These library scenes show our authoring detail. Your digital twin is built from your own site references.

Built around what
your robot needs to do.

Your space

Layouts, dimensions and objects recreated from your references.

Physical interactions

Collisions, mass and moving parts for the tasks your robot needs to perform.

Materials & labels

Surface materials and semantic labels aligned with your perception pipeline.

Your simulator

USD delivery with compatibility checks agreed for your simulator and version.

Generate for where
your robot struggles.

Turn training failures into targeted variations of your digital twin and its assets. Specify the conditions you need, or have your training pipeline request them through the API.

Discuss your training workflow
  1. Identify a gap

    Find the task or condition your policy struggles with.

  2. Generate variations

    Target object differences, clutter, occlusion or lighting.

  3. Train and evaluate

    Use the new assets and scenes. Let the results guide the next request.

From your site
to your simulation.

  1. Share your task

    Bring your site references, robot’s task and simulation requirements.

  2. Agree the scope

    Confirm the spaces, interactions, deliverables, price and timeline.

  3. Build your twin

    Our team authors the environment and assets around your references.

  4. Review & deliver

    Check the agreed behavior, review feedback and receive your USD package.

Before we begin.

What will we receive?

Your agreed USD scene and assets, the dependencies needed to open them, and the materials, physical properties, moving parts and labels your task requires. Delivery also includes the package notes, usage terms and validation findings defined in your scope. We agree the deliverables before authoring begins.

How is my project priced?

Site size, unique objects, reference quality, moving parts, label coverage and simulator requirements determine the quote and delivery schedule. We review these with you before work begins. Digital twins are a custom service, quoted separately from self-serve credits.

Do we need complete CAD or a scan to get started?

Bring whatever you already have: site photos, layouts, dimensions, CAD, scans or existing models. We will assess whether it is sufficient and agree any additional reference capture needed for the scope.

Can you match our simulator and label taxonomy?

Tell us your simulator and version, required formats, class labels and the interactions you need to test. We agree compatibility and acceptance checks for that stack as part of the project.

How will my digital twin be checked?

We agree import checks and task interactions for your simulator and version. Checks can cover scale, dependencies, collisions, joint motion and labels. Validation records identify the selected SimReady profile, version, coverage and remaining findings. Creating a package and passing validation are separate results; runtime testing checks the agreed behavior.

Can we generate variations based on training failures?

Yes. Use the task and conditions your robot struggles with to request targeted asset or scene variations. You can define the augmentation requirements yourself or have your training pipeline call the API. Your evaluation results guide what to generate next, such as different objects, clutter, occlusion or lighting. We can discuss how this connects to your digital twin and training workflow.

What happens if the scope changes?

We review the additional spaces, assets or interactions with you and agree any change to the quote, milestones and acceptance criteria before adding the work.

What usage rights are included?

We agree the project’s usage rights and any third-party content terms in the quote. Tell us whether the content will stay within your team or needs to be shared with partners.

Tell us what needs to be real.

Start with your task and your simulator. We’ll work through the build with you.

Your referencesAn agreed scopeA custom quote