If you’ve been to AWE before, you’ll know the pattern: flashy demos, big promises, headsets that are always just eighteen months from mass adoption. AWE 2026, held at the Long Beach Convention Center from June 15 to 18, felt different. Not dramatically so — but there was a shift in tone that anyone who’s followed XR for a while would notice. The booths were less about showing off hardware specs and more about showing proof: real deployments, real workflows, real ROI numbers.
The unofficial theme of the event, if you had to pick one, was spatial AI. Not AI bolted onto an AR headset as a selling point, but AI woven into the spatial layer as the reason the whole thing makes sense for enterprise.
From Showcase to Workflow Tool
Here’s the thing about XR adoption in enterprise: it’s been stuck in a cycle of pilots that never scale. A logistics company tries smart glasses for picking assistance, it works well in the pilot, and then… it stays at one warehouse. The friction of managing headsets, retraining staff when the device changes, and integrating with existing systems has been enough to kill most rollouts before they reach the second site.
What changed at AWE 2026 was that several exhibitors were showing multi-site, multi-year deployments. Not ten users in one warehouse — hundreds of users across multiple locations, with measurable dwell time data and error rate reductions they were willing to put in writing. That’s different. That’s a product, not a pilot.
The categories where this was most evident: field service and maintenance, surgical and clinical assistance, and skills training. These three verticals have something in common — the task is complex enough that a visual aid genuinely changes outcomes, and the workforce is hard to scale without technology. Spatial AI fits that profile well.
AI Is Now the Delivery Mechanism
The evolution in how AI showed up at AWE 2026 compared to previous years is meaningful. In 2024 and 2025, AI in XR mostly meant “we’ve added a chatbot to the headset” or “generative AI can create training scenarios.” Useful, but secondary.
At AWE 2026, AI was the primary value proposition in a much more integrated way. The use cases that got the most floor time weren’t about chatbots — they were about contextual understanding. A smart glasses system that watches a technician work on a piece of equipment and, without prompting, surfaces the relevant section of the service manual when the technician opens the right panel. A surgical AR system that uses AI to recognise the anatomy in view and overlay the planned incision path in real time, adjusted for the individual patient’s imaging data loaded pre-op. An on-device AI model that can identify which component a warehouse worker is looking at and flag if they’re reaching for the wrong bin before they make the error.
The common thread is that the AI understands context spatially, not just textually. It sees what the user sees, in three dimensions, and adds a layer that makes the task easier. That’s fundamentally different from a chatbot you query.
The Smart Glasses Hardware Picture
On the hardware side, the landscape is more interesting than it’s been in years. You’ve got a fairly clear segmentation developing:
Consumer/lifestyle smart glasses — Ray-Ban Meta, the incoming Samsung Galaxy Glasses — are doing well in a narrow use case: ambient AI assistance and photography. They’re not enterprise tools but they’re normalising the form factor, which matters.
Android XR glasses are the story that’s going to define the next two years. Multiple OEMs are building on Google’s Android XR platform, which means software compatibility across devices and a development ecosystem that doesn’t require betting on a single hardware vendor. Xvisio’s enterprise-class MR glasses, which launched at AWE, are built on this model.
Standalone enterprise headsets — HoloLens legacy, the XREAL Aura, Snap Spectacles 5 for developers — remain relevant for specific industrial deployments where the Android XR ecosystem doesn’t yet cover the use case.
Snap’s Spectacles got a lot of floor time. The liquid-crystal-on-silicon display with a 51° field of view and 16 million colours is genuinely impressive hardware for what it costs developers to access. The open developer platform means AWE 2026 saw a solid cluster of enterprise applications built specifically for Spectacles — a development the last two AWEs couldn’t claim.
The Enterprise Takeaway
If you’re evaluating XR for enterprise deployment in 2026, a few things are clearer now than they were a year ago.
Android XR is the platform to build on. The fragmentation problem that plagued enterprise XR — where committing to one headset meant stranding your software investment when the hardware was discontinued — is being addressed by a common platform layer. Building for Android XR now means your application runs on future devices without a complete rebuild.
AI integration needs to be native, not bolted on. The deployments at AWE that were generating real enterprise interest had AI running on-device, understanding spatial context, not making API calls to a chatbot. The hardware now exists to support this (on-device NPUs have caught up to what’s needed for inference on lightweight models) and the business case for contextual spatial AI is cleaner than it’s ever been.
The ROI conversation is maturing. MSPs and systems integrators exhibiting at AWE are showing three-to-five-year cost modelling, not just “here’s what the headset costs.” Error reduction, training time compression, and remote assistance cost savings are the numbers that get sign-off from enterprise finance teams. The industry is finally learning to speak that language.
AWE 2026 didn’t announce the product that makes XR mainstream. It showed that a specific slice of enterprise XR — contextual spatial AI for complex knowledge work — has already crossed from promising technology into quietly proven one. That’s a different kind of milestone, and probably a more durable one.