Product designers have been building physical prototypes to catch problems for as long as manufacturing has existed. A prototype reveals what a CAD model can’t: whether a component is awkward to reach during assembly, whether the ergonomics feel wrong at actual scale, whether two parts that look fine in a 3D model will actually interfere in the real thing.
Physical prototypes are expensive and slow. A machined prototype for a complex automotive component can cost tens of thousands of pounds and take weeks. Spatial computing can’t replace physical testing entirely, but it’s increasingly useful for the earlier, cheaper question: is this design worth building a physical prototype of at all?
Where XR Actually Fits in the Design Process
The clearest wins from XR in product design fall into a few specific categories, and they’re not the ones the headset vendors tend to lead with.
Scale review at 1:1. Looking at a CAD model on a screen gives you a distorted sense of scale — a desktop or monitor representation of a vehicle interior or a large machine component doesn’t convey how it’ll actually feel to sit in or work on. Loading that model into a VR environment at true 1:1 scale gives designers and engineers a spatial intuition that no amount of screen-based review provides. “We didn’t realise the handle clearance was that tight until we walked through it in VR” is a comment you hear consistently from teams that have adopted this workflow.
Assembly and maintenance path checking. Can a technician actually reach that fastener with their hand and a socket wrench in the available space? Does the assembly sequence work given the geometry of surrounding components? These are questions that are genuinely difficult to evaluate from a static CAD view and trivially obvious when you’re standing inside the assembly in VR. Automotive, aerospace, and medical device manufacturers are all using this — the ROI is direct: finding an impossible maintenance path in VR costs nothing; discovering it after tooling has been cut is expensive.
Multi-site design reviews. A product team distributed across London, Munich, and Detroit can load the same CAD model into a shared virtual environment and review it together without travel. The collaboration isn’t perfect — current enterprise XR collaboration has latency and fidelity limitations — but for design reviews where the goal is “does this design have obvious problems,” it’s substantially better than a video call with screen sharing.
The Toolchain in 2026
The major CAD vendors have all built XR integration, with varying levels of maturity:
PTC Vuforia and Creo have the most mature enterprise AR pipeline. Creo models export directly into Vuforia Studio for AR/VR review, and Vuforia’s industrial AR authoring is used by GE Aviation, Bosch, and similar manufacturers for assembly guidance that sits alongside the CAD-to-VR review workflow.
Dassault CATIA and 3DExperience integrate with VR headsets through 3DExperience’s immersive virtuality module, supporting collaborative VR design review for teams already in the CATIA ecosystem. The integration is deep but the costs are high — this is enterprise-grade pricing.
Siemens NX and Teamcenter support XR review through Siemens Xcelerator and partnerships with Virtalis and similar enterprise VR platform vendors.
For smaller teams using SolidWorks or Fusion 360, the integration is less seamless. SolidWorks has a VR plugin (XVR), and Autodesk Fusion exports to formats readable by enterprise XR platforms — but you’re looking at more manual export steps rather than native workflow integration. The open-source alternative is to export to GLTF or USD format and use Unity or Unreal Engine as the VR review environment, which works well for teams that have development expertise in-house.
Hardware Choices for Design Teams
The headset choice depends heavily on the design review use case:
Apple Vision Pro is the best available display for static or slow-moving design review — the display quality and resolution are unmatched, and the passthrough mixed reality lets you blend the CAD model with the physical environment. The drawback is cost (£3,499+) and the lack of controller input for detailed model manipulation. Teams are using it primarily for executive and customer design presentations where display quality matters.
Meta Quest 3 (or Quest 3S for budget-constrained teams) is the practical choice for everyday design reviews. Lower display quality than Vision Pro but adequate for most engineering review purposes, controllers for model interaction, and a price point that makes multi-seat deployment feasible. The Quest’s hand tracking has improved enough to be useful for basic manipulation.
Varjo XR-4 sits between the two for industrial use — near-retinal-resolution display, excellent passthrough AR, and hardware built for heavy enterprise use rather than consumer markets. Used by automotive and aerospace teams where display quality is critical for detail review. Expensive but significantly below Vision Pro’s cost.
What Doesn’t Work (Yet)
Honest assessment: there are real limitations to the current technology in product design contexts.
Haptic feedback is still inadequate for evaluating ergonomics. You can see whether a handle looks like it’ll fit your hand at the right scale, but you can’t feel the weight distribution or surface texture. This remains a gap — haptic glove technology is progressing (covered in our haptics article) but isn’t mainstream in design workflows yet.
Very large assemblies (50,000+ polygons) still cause performance problems in real-time VR environments without significant polygon reduction. The workflow typically involves creating a “VR-optimised” model with simplified geometry for review, which means the review is on a somewhat degraded version of the design. Teams deal with this by being selective about which assemblies they review in VR rather than defaulting to it for everything.
That said, the ROI calculation for the use cases where it does work is clear enough that adoption is accelerating. The most common report from teams that have integrated XR into design review: “We don’t understand how we spotted these problems before.” For interference checking and scale verification, the answer is: you often didn’t spot them until the physical prototype.