Luminary Cloud in the Onshape App Store: Physics AI Meets Parametric CAD

Onshape’s open platform philosophy means engineers aren’t limited to what ships out of the box. Through integrations and add-ins from technology partners, users can extend the cloud-native CAD and PDM platform beyond core functionalities into new areas of simulation, manufacturing, and data management.

The Luminary Cloud integration is a prime example – bringing Physics AI directly into the Onshape workflow to collapse the traditional barrier between design and simulation while leveraging the power of AI.

How Luminary Cloud Accelerates Engineering Workflows

Traditional simulation typically happens after design – you finish a concept, hand it off to a specialist, wait for meshing and solver runs, then iterate based on results. This process can take days or weeks, creating a bottleneck in product development.

Luminary Cloud's Physics AI platform breaks this bottleneck by making it practical to generate simulation data at scale and train models that predict results in near real-time.

Physics AI: From Simulation to Inference

Luminary’s Physics AI platform serves teams anywhere fluid flow and heat transfer matter – automotive, aerospace, electronics, HVAC, and industrial machinery. The platform supports CFD and thermal analysis with more physics in development.

The core idea? Traditional simulation solves physics equations numerically, which requires meshing geometry and running computationally intensive calculations. This takes time – often hours or days per simulation. That’s fine for validating a final design, but it’s too slow for real-time design feedback or exploring thousands of early variants.

Physics AI flips this model. You generate a dataset of simulations upfront, train a machine learning model on that data, and then run inference – where the trained model predicts results for new geometry without solving the physics equations. Inference takes seconds instead of hours.

Luminary’s platform handles the entire pipeline. GPU-native simulation generates training datasets faster and more efficiently than traditional solutions. Automated orchestration manages the complexity of simulation and model training workloads, so engineers can focus on design rather than infrastructure. Physics AI model training delivers performance without requiring specialized machine learning expertise. And flexible inference APIs let you integrate model predictions directly into your existing toolchain.

This means engineers can run inference on design changes and get accurate physics feedback almost immediately. R&D teams can explore thousands of design variants in the time it once took to evaluate one.

To help teams get started, Luminary offers pre-trained SHIFT models like the SHIFT-Wing for aircraft geometries, SHIFT-SUV for automotive, and SHIFT-Pump for turbomachinery. More models are in development. These work out of the box for geometries within their training distribution, or you can fine-tune them with your own data to build proprietary models for your unique design space. The Onshape integration played a key role in generating the SHIFT-Wing dataset – Onshape’s parametric modeling capabilities enabled the Luminary Cloud team to produce thousands of geometry variants while maintaining design intent and version history.

The predictions aren’t just simple outputs like lift and drag. Luminary’s models generate full flow fields – pressure distributions, velocity fields, and other quantities across the entire geometry – so downstream teams can use them for thermal, structural, or other analyses.

Connecting Onshape and Luminary Cloud

Both Luminary and Onshape are cloud-native, which creates natural synergy – no software to install, no IT approvals, and collaboration through shared links rather than zipped files. For teams adopting both platforms, it’s a way to stay at the forefront of what’s happening in CAD and simulation, particularly as Physics AI continues to reshape engineering workflows.

The integration maintains a live connection between platforms. You authenticate, then browse your Onshape documents, workspaces, and versions directly within Luminary. When you import geometry, Luminary creates a version in Onshape to track the exact state at import time.

Variable Studio parameters from Onshape are pulled into Luminary and stay connected, so you can modify dimensions directly in Luminary and pull updated geometry without switching applications. This makes parameterized CAD particularly powerful – set up your configurations in Onshape and drive design exploration at scale from Luminary. If a parameter change breaks Onshape's regeneration, those warnings propagate into Luminary so you're not unknowingly working with invalid geometry.

Whether you’re generating training data or running inference for real-time feedback, the integration keeps your CAD and physics results connected throughout the design process.

Strength Through Partnership

The Onshape App Store brings together leading solutions in simulation, design, manufacturing, data management, business system integration, and more – all built to work seamlessly with Onshape’s cloud-native architecture.

Browse integrations from trusted technology partners, read verified user reviews, and implement apps directly from your browser. Every integration is designed to extend Onshape’s core capabilities and help your team deliver better products faster.

Interested in integrating your product with Onshape? Reach out to onshape-developer-relations@ptc.com to get started.

Getting Started with Luminary Cloud

Luminary Cloud is available in the Onshape App Store as a public partner app. Currently, it works as a connected cloud app – you import geometry from Onshape into Luminary’s environment for simulation and inference.

Visit the Onshape App Store to connect Luminary Cloud and explore other simulation partners that can bring new capabilities to your product development process.