Comment Onshape et ses partenaires innovent grâce à l'IA

Computer-Aided Design (CAD) is undergoing a transformation. Static, manual modeling, and siloed simulation are giving way to a new era of intelligent, connected, and highly automated workflows.

At the forefront of this revolution is Onshape, a cloud-native platform that inherently supports rapid innovation. But Onshape isn't doing it alone. Through a growing ecosystem of specialized partners, AI is being woven directly into the engineering process, dramatically reducing friction, accelerating iteration, and unlocking new design possibilities.

Let's dive into how Onshape and four of its key partners—SimScale, Luminary Cloud, Leo AI, and Adam AI—are leveraging artificial intelligence to reshape how hardware is built.

Onshape: The Cloud-Native Foundation for AI

Before looking at the partners, it's crucial to understand why Onshape is the platform of choice for AI innovation.

Traditional, desktop-bound CAD systems struggle with the sheer volume of data and the seamless connectivity required for modern AI tools. Onshape’s cloud-native architecture changes the paradigm. Because all design data, version history, and PDM (product data management) live in a centralized, secure cloud database, AI applications can interact with the live model in real time. There are no files to export, upload, or translate. Onshape’s open-sourced feature programming language, FeatureScript, is an added advantage. LLMs can learn programming languages easily, and with coding assistants, engineers can develop custom CAD tools faster than ever before.

Through initiatives like "Onshape Labs," the platform is actively developing native capabilities, such as AI Agents and MCP servers that can automate tasks, enforce design standards, and operate alongside engineers with full visibility. By providing a robust, API-driven platform, Onshape allows third-party innovators to plug directly into the design environment, extending core workflows without fragmenting the user experience.

1. SimScale: Democratizing Multiphysics Simulation

SimScale is transforming how engineers validate their designs by bringing powerful simulation out of specialized silos and directly to the cloud.

Through its native integration with Onshape, engineers can sync their CAD models to SimScale with a single click—no downloading or messy geometry translations required. But where SimScale truly innovates is in its open execution infrastructure. The platform is designed to allow AI companies, solver vendors, and method developers to integrate their technologies directly into the SimScale environment.

By combining cloud computing power with AI-driven assistants, SimScale allows hardware startups and enterprise teams alike to run complex fluid, thermal, structural, and electromagnetic analyses in a fraction of the traditional time. This "CAD associativity" means engineers can rapidly iterate designs in Onshape and instantly re-run simulations, drastically cutting down on physical prototyping costs and time-to-market.

2. Luminary Cloud: Engineering at the Speed of Physics AI

Luminary Cloud is answering the call for "Physics AI." Luminary provides a platform for engineering teams to build and operate Physics AI models at scale, essentially encoding physical laws into neural networks.

The innovation here is staggering. Luminary’s models can generate accurate predictions of physical fields up to 1,000 times faster than traditional simulation, with less than a 2% prediction error for leading models.

For Onshape users, this means replacing the slow bottleneck of traditional analysis with near-instant physics prediction. Teams can explore vast design spaces, stress-test designs across a wide range of operating conditions, and catch failure modes early.

By offering APIs and an AI Assistant, Luminary makes it easier to embed these lightning-fast Physics AI predictions directly into the tools and pipelines engineers are already using, enabling rapid iteration that was previously unimaginable.

3. Leo AI: An Engineering Copilot

Leo functions as an engineering copilot, streamlining the journey from conceptual ideation through to final validation for hardware teams. By serving as a unified intelligence layer, it integrates parametric modeling, rigorous design auditing, and intricate technical computations into a cohesive environment that respects proprietary standards.

Leo mitigates manual overhead by automating recurring workflows, assessing assembly integrity, and recommending reusable components from internal libraries and global supplier networks. This synergy of data and automation allows engineers to bypass repetitive tasks, fostering a more informed and frictionless development process.

4. Adam AI: Text-to-CAD & Taming the Feature Tree

Adam AI is an AI-powered CAD copilot designed to optimize and clean up feature trees within Onshape Part Studios. It excels at interpreting existing model structures to apply targeted, structured geometry updates that respect original design intent.

Engineers use plain-language prompts to instruct Adam to perform tasks, such as:

• Extracting hardcoded numerical values into named, shared variables,
• Renaming vaguely titled features with descriptive, geometry-based names, and
• Consolidating redundant modeling operations. This functionality brings clarity and consistency to complex models, significantly facilitating team review, modification, and collaboration.

The Onshape integration is still in Beta. During which time Adam AI will continue to explore new use cases including leading edge use cases for text-to-CAD. For example, the thread body depth for a threaded part, shown below, can be changed simply by prompting Adam AI in plain English. Because Adam AI is feature tree aware, it can consider changes that need to be made to other related parts without propagating errors in the feature tree.

The Future is Collaborative and Intelligent

Built on Onshape's cloud-native architecture, these partnerships represent an expanding ecosystem that proves when design data is accessible and platforms are open, AI in engineering can eliminate bottlenecks, reduce manual overhead, and empower hardware teams to build better products, faster than ever before.