An exclusive preview of 5 companies exhibiting at AWE USA 2026

Thanks to the amazing team at AWE, I’ve been given the exclusive opportunity to preview five interesting companies that will exhibit at the AWE USA event taking place in Long Beach (LA) from the 15th to the 18th of June, 2026. These companies had pretty cool stuff to show me… and I can show it to you, too!

Oxford Optical Labs

Oxford Optical Labs aims to solve two major problems with XR headsets, especially when used in public settings (e.g., museums).

They have developed an adjustable fluid lens that is not made of glass, but it is like a plastic membrane with inside a fluid. By changing the electricity going through the lens, they can dynamically change the optical parameter of the lens, as you can see in the video below.

This means that if you use these lenses inside a headset, or as an add-on to an existing headset, the same lens can support the prescription of all people using the device. You could theoretically arrive at a VR arcade, give your current prescription glasses to the clerk, and he could copy your prescription from your glasses to the XR device. This visio is pretty cool and would save people a lot of time and money that is currently spent for the setup of prescription add-ons.

But there is more: we all know about the fixed-focus problem in XR headsets. Well, since you can change the optical parameters of these lenses on the fly, you can also change the focus on the fly. This means that if you use eye-tracking together with these lenses, you can make the user focus on the virtual elements at whatever distance they are. Virtual elements could so appear in focus together with physical elements everywhere in the environment around you. The switch of focus happens within 70ms, which is less than the saccadic suppression of the eyes, so most users do not notice anything happening and just see the focus changing naturally.

I asked about the durability of these lenses and the eventual risks of fluid leaks. I’ve been told the lenses are actually very resistant and pass all the tests that standard glasses have to pass. You can even try to scratch the lenses with your keys, and they would just be fine. The company has been using the same technology for more than 15 years (outside of VR), so it has a long experience with it.

I personally think this is a very cool solution. Both prescription and focus have been two big problems in our field for the last few years. Currently, there is no solution for them, so this company is trying to fill a void. The fact that it has a long experience with this technology makes me very positive about the quality of its product. It has to be seen if costs and durability are coherent with the current and future requests of the market, though. In any case, I am very curious to try it at AWE.

SpatialGen

SpatialGen is the platform currently powering 90% of all third-party immersive video on the Apple ecosystem. Now they are announcing at AWE SpatialGen ZEUS, which is a solution to allow the livestream of Apple spatial content with ease.

The company is going to offer the server rack and the software that takes the input from an immersive camera, encodes it, packages it, and distributes it as a live stream via the customers’ CDN. Going a bit technical:

• The expected input is: Apple ProRes over SMPTE (2110/SDI) standards.
• The expected output is: Real-time AIV-compliant MV-HEVC (Multiview High Efficiency Video Coding) HLS segments.

The idea is to offer a solution that works out of the box, with the customer just having to handle the video recording and not caring about anything else. Notice that SpatialGen does not handle the distribution of the content: its purpose is to transform the stream of the camera into a streamable video format, but then the customer has to provide themselves with a CDN from which the users can actually stream the live video.

SpatialGen is looking for big customers (e.g., telcos or Hollywood studios) that are interested in ultra-low latency streaming for live sports, concerts, and breaking news in Apple Immersive Video format. I say “big customers” because this is a professional enterprise solution, whose price starts from $65,000.

The company is also announcing at AWE that it has started a collaboration on the military side with the US Air Force. Using SpatialGen’s existing commercial scale, “the USAF can deploy sophisticated visualization tools faster and more cost-effectively than through traditional defense procurement alone.”

I don’t have much to comment on this solution. It is something with a very clear use case, for a very well-defined set of customers. For sure, the fact that they previously worked with many Vision Pro video shows that this product should be reliable. In any case, I find it interesting that the Vision Pro has reinvigorated the interest in 180-360 videos.

4D Views

4D Views is a company with a long history in the field of volumetric videos. They started doing this work in 2007 (before it was cool) and can record this kind of videos all over the world thanks to more than 20 partner studios in different countries. All over these years they worked on famous immersive projects like Kagami or An Ark.

The announcement for AWE is that they are also starting to record volumetric videos with Gaussian Splatting. They can be one of the first studios to offer both ways to record volumetric videos:

• The traditional way with meshes and textures. This has the advantage that the capture is easier to edit using the company’s software called 4DFX. Besides that, since it is the “old” method, it integrates better with existing pipelines and game engines
• The new way with Gaussian Splatting. This has the advantage of rendering better transparency, props, and details.

4D Views is also going to announce at AWE a plugin to easily integrate these splat videos inside game engines. Interested companies may reach out to them at the AWE booth or via e-mail.

I think that Gaussian Splats-based videos are the future for entertainment, communication, and various other fields (yeah, also that one). So the work that 4D Views is doing is for sure very interesting. Its long experience in the field also shows they can deliver a high-quality service.

MetaNeural

MetaNeural is a company proposing a platform for training and simulation. It worked for many years in B2B, especially in the critical infrastructures sector, so it knows what these companies are looking for and has built a platform that could fulfill their needs. For instance, secretive companies do not want to share their data, not even to other internal departments, so MetaNeural made its platform no-code, so its customers don’t need a dev department they should share their info with. The customers also need to ingest their own data and their own 3D models that may come from various sources, and that’s why MetaNeural platform can support many different sources.

At AWE, they are going to announce a few new features that may be of great interest to the military sector. One of them is a complete simulation engine for drone behaviour. Using it, you can, for instance, simulate an attack from a swarm of drones, which has become pretty common in today’s warfare. You can simulate in the engine your situation (e.g., your company, your military base, etc) and then ask the engine to simulate an attack. The MetaNeural software will calculate what are the most critical infrastructures in your scenario and attack them. From the simulation, you may see if you have protected enough your important resources, and also, you may organize better countermeasures. You can train on what you do when your infrastructure has been attacked, and so on.

All of this happens just by talking to the system. As I said, there is no code. You just naturally talk to the application in natural language, and things happen thanks to AI. I asked the company about the risks related to AI hallucinations, and they told me there are a lot of guardrails in place so that they may have very little effect.

MetaNeural is also going to announce a human behavior simulation engine for critical situations. For instance, what happens if some terrorists enter a factory? The engine can simulate how people behave during the crisis: for instance, if the police don’t intervene fast enough, some people may start getting crazy and do dangerous things. It also simulates what happens during all the phases of the critical situation: e.g., the arrival of special forces, media trying to get information, etc…

I think this can be a very useful platform. I know there are a lot of investments going on in Europe about the military sector now, so the fact that MetaNeural is providing a solution for these kinds of scenarios can be a good way for them to have a new source of revenue and investments. It’s not the funniest ecosystem to work in, but someone has to do that job.

Prehension

Prehension offers a gesture recognition engine. Hand tracking is becoming more and more important as an interaction means in VR and MR. Existing frameworks are pretty good at providing you with the pose of the hands, but almost none of them also provide a good solution for gesture recognition. And this is where Prehension enters the game.

Prehension wants to provide a gesture recognition engine for Unity that is customizable and easy to use. What you do is:

• Run a specific training scene and perform the gesture that you want your application to recognize
• Let the system take all this data and send it to a Prehension online service that will train a recognizer
• When the service has finished its work, you can download a file containing the training ML data useful to recognize that gesture
• You can then use this file inside your Unity project and have your gesture recognized out of the box.

It is interesting to notice that while the machine learning training happens on the cloud, the recognition engine is fully local, and this ensures very low latency.

Prehension claims that its system works with all hand tracking plugins (e.g., Pico, Meta, etc), it has low latency, and it doesn’t consume many computational resources.

It is sold as a monthly fee you have to pay for a seat. The exact amount is still undisclosed.

This seems to me a pretty handy (pun intended) solution. From the description, it seems to be rather straightforward to use. The only big question I have is whether there are enough companies now needing such a solution to sustain their business. For the rest, it seems pretty valid.

And that’s it! I hope you enjoyed this preview (and if it is the case, please subscribe to my newsletter) and if you have questions, feel free to ask them in the comments. See you at AWE in Long Beach!

(Header image by 4DViews)