Nikhil Sharma | Vizon Tech: Why must future AR operating systems be open source rather than controlled by closed ecosystems?
15:18 - 16:46
Other snippets from this talk
Summary of the clip:
Why must future AR operating systems be open source rather than controlled by closed ecosystems?
OpenXR is seen where a lot of standards exist, but products don't support it and a lot of features miss. Clients ask for Unity, Unreal, Godot, or completely random engines, so support everything now, even if it can't be done at scale. For every pair of glasses, a Windows app should work. That solves the biggest XR software problem: lack of content. There are hardly one or two games, two or three things to do, even with good hardware.
But there are thousands and millions of 2D apps and websites that you should be able to use in AR. The hardware carries seven cameras, nine cameras, whatever sensors, IMUs and mics, and it has to be open source. The Apple ecosystem will never be open source. Android now is not going to be open source anymore. Android XR definitely not. So whatever comes next should be, and that matters for developers and consumers alike.
Consumer risk is real. You don't want to see a lawsuit like Meta's, where the feed from your glasses is used by some African kids to train AI. You don't want that. Or you train AI and it wants to show you relevant ads. Imagine ads in AR: pathetic, blocking your field of view. That is why the next platform must be open source.
In this short video, you can learn:
* OpenXR is seen as the place where a lot of standards exist, but products do not support it and a lot of features are missing.
* Supporting Unity, Unreal, Godot and other engines so any Windows app runs on any glasses addresses the biggest XR software problem, lack of content, against thousands and millions of 2D apps and websites.
* Glasses carrying seven or nine cameras, IMUs and mics must be open source, because Apple never will be, Android is not going to be anymore, and Android XR definitely not.
📋 **Clip Abstract** OpenXR standards exist, yet products ship without support and miss features, so the practical answer is to support every engine, Unity, Unreal, Godot, anything a client asks for, and make any Windows app run on any glasses. Because devices carry seven to nine cameras, IMUs and mics, and Apple, Android and Android XR will never be open source, the next platform should be.,Devices carry seven to nine cameras, IMUs and mics, and that sensor load is exactly why closed ecosystems fail the test.
About the speaker:
* Speaker: Nikhil Sharma
* Company: Vizon Tech
* Event: Eindhoven 2026
* Location: High Tech Campus, Eindhoven
#OpenXR, #GodotEngine, #UnityAndUnreal, #AndroidXR, #AugmentedRealityGlasses, #XRSpatialComputing
This is a highlight of the presentation:
What it'll take, to make Augmented Reality Glasses a mass adopted Reality.
More Highlights from the same talk.
03:08 - 05:37
How can AR hardware developers balance the trade-offs between cost, field of view, and form factor?
How can AR hardware developers balance the trade-offs between cost, field of view, and form factor?
The core dilemma was affordability versus form: a $300 device people can buy, or something more like normal glasses that costs far more. The response was to build both, despite startup advice. They share 1920x1080 per eye at 90 Hz, hand tracking, 6DoF, SLAM, light tracking, same resolution and similar cameras. One uses prisms with micro LEDs; the other uses waveguides currently with DLP, while other solutions are sought. Looking slightly more like glasses or sunglasses doubles cost.
Scope expanded beyond two products. The original aim was consumer AR glasses, but industrial clients began reaching out. One client received five VR glasses, then said they needed a VR headset. Using existing expertise, in less than two months the team built what is probably the world's lightest VR headset. It currently weighs 120 grams, with a stated target of less than 90 grams, plus 6DoF tracking and hand tracking, and 5K stereo via 2.5K per eye OLED displays.
A few years ago, the widest field-of-view XR see-through headset was built: optical see-through, huge at 300 grams—large versus optical see-through, small versus VR. It uses five infrared cameras at different lenses, industrial grade. Meta Ray-Bans and Google are about 15–20 degrees; other glasses reach 40, 45, 50 at most. XTAL Aura or Oriens are at 70 degrees. On the desk, a 120-degree field of view covers entire human binocular vision; that is the limit, with no point exceeding it.
In this short video, you can learn:
* Two AR designs share 1920x1080 per eye at 90 Hz, 6DoF, SLAM and hand tracking, but one uses prisms with micro LEDs while the other uses waveguides currently with DLP.
* Making an AR device look slightly more like glasses or sunglasses doubled its cost, exposing the form-factor penalty in the bill of materials.
* A 120-degree optical see-through headset covers entire human binocular vision, beyond 70-degree XTAL Aura or Oriens and 15–20 degree Meta Ray-Bans or Google.
📋 **Clip Abstract** The clip contrasts affordable AR hardware and glasses-like form factor, using two closely matched products to expose cost trade-offs. It also traces a wider product line: a sub-90-gram target VR headset, a 300-gram five-camera optical see-through design, and a 120-degree field-of-view system at the human binocular limit.
About the speaker:
* Speaker: Nikhil Sharma
* Company: Vizon Tech
* Event: Eindhoven 2026
* Location: High Tech Campus, Eindhoven
#MicroLEDPrisms, #WaveguideDLP, #6DoFSLAM, #120DegreeFOV, #AROptics, #XRLightEngines
11:38 - 13:10
Why does micro-OLED currently outperform LCOS and micro-LED for near-eye displays in augmented reality?
Why does micro-OLED currently outperform LCOS and micro-LED for near-eye displays in augmented reality?
A broad display hunt covered 0.1-inch to 2.1-inch panels, including micro-OLEDs, micro-LEDs, and LCOS, with custom PCBs designed for testing. The conclusion was blunt: micro-OLEDs right now are significantly superior to LCOS and micro-LED. They offer much higher resolution, much greater market readiness, much lower cost, and the best color accuracy and clarity. That combination makes micro-OLED the closest display route for a monitor replacement, despite the desire for better options.
For monitor use, field of view matters. A 60-degree field of view is stated as necessary; anything less does not cut it. If the device stays in 3DOF in space, even a current 1920p product is judged insufficient. A 2560p resolution is said to be needed for 3DOF monitor use, with average PPD usually acceptable around 50 PPD.
That resolution gap may drive a product decision: the current glasses might bump up the price and move to a 2560p version, though this depends on the Chinese display guys. The trade-off is not only optical. If glasses cost $700 for just a line of text or notifications, users can already see that on a smartwatch. Songs and translation can run on them, but a monitor replacement demands more. A wide FOV headset was also made.
In this short video, you can learn:
* Micro-OLEDs are currently significantly superior to LCOS and micro-LED on resolution, market readiness, cost, and color accuracy and clarity.
* A 60-degree field of view is treated as the minimum for monitor replacement, with anything less not cutting it.
* For 3DOF monitor use, 1920p is judged insufficient; 2560p is needed, with average PPD acceptable around 50 PPD.
📋 **Clip Abstract** The clip compares micro-OLED, LCOS, and micro-LED panels from 0.1 inch to 2.1 inch after extensive buying and PCB design. It concludes that micro-OLEDs are significantly superior today, then sets out 60-degree FOV, 2560p, and 50 PPD requirements for 3DOF monitor use.
About the speaker:
* Speaker: Nikhil Sharma
* Company: Vizon Tech
* Event: Eindhoven 2026
* Location: High Tech Campus, Eindhoven
#MicroOLED, #LCOS, #MicroLED, #2560pResolution, #NearEyeDisplays, #ARGlasses
06:52 - 08:16
How can AR glasses achieve all-day battery life and high performance without becoming too heavy or socially unacceptable?
How can AR glasses achieve all-day battery life and high performance without becoming too heavy or socially unacceptable?
The four glasses stay comparatively small because processing is offloaded entirely to a handheld unit. That unit began on a Qualcomm processor, then moved to Rockchip because Qualcomm was very expensive. It runs the hand-tracking computation, every algorithm, on-board AI, and the ISP handles all XR and AI 3D rendering. The glasses themselves carry only displays, cameras, the ICs to use them, and audio.
But the handheld was still wired, which the speaker rejected, so the next step is making it completely modular. The compute could be strapped at the back of the head, which also balances the weight, or better still packed inside a headphone form factor, since headphones are the only thing that is huge and still socially acceptable sitting on a head.
Audio quality will probably be bad, because audio needs space, but that is not the goal. One side can be filled with batteries, the other with the best processing unit possible, and it stays socially acceptable. Still tethered to the glasses, the glasses process nothing and carry no battery, so it lasts an entire day. That, he argues, is the only way to sell a million or ten million units — socially acceptable, small and cheap.
In this short video, you can learn:
* Offloading every algorithm, AI inference and XR/3D rendering to a separate handheld unit is what keeps the glasses small.
* Swapping the Qualcomm processor for Rockchip cut cost, and that unit is still wired to the glasses today.
* A modular headphone-form-factor pack, batteries on one side and compute on the other, balances weight and delivers all-day runtime.
📋 **Clip Abstract** This clip explains why a tethered, modular compute pack — handheld today, headphone-shaped next — is the route to small, cheap, socially acceptable AR glasses. Offloading AI, hand tracking and XR rendering off the frame, plus batteries on one side and processing on the other, is what yields all-day runtime and ten-million-unit volume potential.,Offloading AI and XR rendering to a tethered pack, with batteries and compute split across a headphone-form-factor module, is presented as the only route to all-day runtime and million-unit volumes.
About the speaker:
* Speaker: Nikhil Sharma
* Company: Vizon Tech
* Event: Eindhoven 2026
* Location: High Tech Campus, Eindhoven
#RockchipProcessor, #OnboardHandTracking, #HeadphoneFormFactor, #AllDayRuntime, #AugmentedReality, #SmartGlasses