Robert Schultz | Vuzix: Why must waveguide gratings do no work on the light—and what breaks when vector triangles open up?
02:01 - 04:08
Other snippets from this talk
Summary of the clip:
Why must waveguide gratings do no work on the light—and what breaks when vector triangles open up?
Waveguides work by coupling light in and replicating the pupil across the system. Exit pupil expansion laterally translates from the edge of the eye out to the front, relaying the pupil with infinity light in and infinity light out. The system maintains angles throughout. At the projector, individual pixels move from X,Y position into the angle of the lens. Any spread in the pixel creates spread in the waveguide, and angular range encodes and separates one pixel from another in the image.
In regular AR waveguides, the concern is contrast between pixels and how good and clean the display looks. Later, the concern becomes data and corruption of pixels from neighboring pixels in data streams. To first order, we do not want the gratings doing any work on the light. When designing waveguides, we close our vectors and ensure input and output vectors match. A simple design puts input and output on a single 1D grating.
In a 2D grating system, light comes across multiple dimensions, still making sure that all vector triangles close and that no actual work is done on the gratings. As soon as you leave those vectors open, you start getting chromatic dispersion and other issues in the waveguide. This is the first-order rule: gratings must not work on the light, and input and output vectors must match.
In this short video, you can learn:
* Exit pupil expansion laterally translates from the eye edge to the front and expands the pupil while maintaining angles, with infinity light in and infinity light out.
* Projectors convert individual pixel X,Y position into angle, so pixel spread becomes waveguide spread and angular range separates one pixel from another.
* Waveguide design closes vectors so input and output vectors match; open vector triangles cause chromatic dispersion and other issues.
đź“‹ **Clip Abstract** This clip covers waveguide basics: light couples in, the pupil is replicated, and exit pupil expansion relays and expands it while maintaining angles. It then explains angular encoding and the design rule that gratings should do no work, requiring closed vector triangles; open vectors cause chromatic dispersion.
About the speaker:
* Speaker: Robert Schultz
* Company: Vuzix
* Event: Eindhoven 2026
* Location: High Tech Campus, Eindhoven
#ExitPupilExpansion, #AngularEncoding, #ChromaticDispersion, #VectorTriangleClosure, #ARWaveguides, #DiffractiveOptics
This is a highlight of the presentation:
Facing Modern Challenges in Waveguide Design and Manufacturing
More Highlights from the same talk.
07:26 - 08:17
How much waveguide production capacity exists today, and how soon is it doubling?
How much waveguide production capacity exists today, and how soon is it doubling?
The manufacturing campus is in Rochester, New York—actually West Henrietta, just outside Rochester, about 10 minutes up from the University of Rochester. The building is working, and there are now two buildings in total. The facility is used for waveguide production, with pictures from inside showing waveguides handling and the work being done there.
Just added is a second building, and about a third of that building has been taken over as a pure waveguide manufacturing facility. Inside, the team is handling waveguides and running the processes. The current facility is set up to support more than a million units, and an additional machine was added over the last two weeks.
Capacity is now being looked at for doubling into two and a half million a year, in terms of what can be supported. The current facility supports a million plus, and the new machine moves the operation toward that doubled target. This is the present waveguide manufacturing scale: one million-plus today, with two and a half million a year as the immediate capacity ambition.
In this short video, you can learn:
* The manufacturing campus is in West Henrietta, just outside Rochester, New York, about 10 minutes from the University of Rochester.
* A second building was added, with about a third taken over as a pure waveguide manufacturing facility.
* Current capacity is set to support a million-plus waveguides a year, and an additional machine added in the last two weeks points to doubling to two and a half million a year.
đź“‹ **Clip Abstract** The manufacturing campus in West Henrietta, just outside Rochester, New York, now includes two buildings, with about a third of the second building taken over as a pure waveguide manufacturing facility. The current facility supports a million-plus waveguides a year, and after adding another machine over the last two weeks, capacity is being looked at for doubling to two and a half million a year.
About the speaker:
* Speaker: Robert Schultz
* Company: Vuzix
* Event: Eindhoven 2026
* Location: High Tech Campus, Eindhoven
#WaveguideManufacturing, #WaveguideHandling, #PureWaveguideFacility, #AnnualWaveguideCapacity, #ARWaveguides, #AugmentedRealityOptics
16:19 - 18:26
What pixel pitch, angles and gigahertz rise times does Vuzix need from its partners?
What pixel pitch, angles and gigahertz rise times does Vuzix need from its partners?
Compact packaging starts with the grating and built-in lenses; wider-pitch VCSEL arrays pull the focal length in, split emission into individual channels and let the design remove lenses altogether, collapsing the system. Polarization-insensitive designs use power gratings in and/or out. Zone gratings divide and spread light as needed. Crossing optical paths and short terms sit inside the waveguides, and high-temperature materials open the solution spaces under consideration.
From partners the list is stability, low pixel outs, maximum efficiency and low pixel defects that will kill the program. The coming angle out of the pixels matters as much as raw output; another talk proposed lensless microdisplays where pixels are offset to direct light more efficiently inward. Efficiency is the constant target. The same physical questions return whenever these components are taken into real systems, and the answers shape what partners must supply.
One route is two- to three-micron pixels on a ten- to fifteen-micron array, so more light couples into the waveguide. That reframes the problem as how much light over what field of view can be delivered to the system. Rise and fall times need to work ideally in the gigahertz range. Size and geometry need to be as small and fine as possible, which is where the manufacturing questions land.
In this short video, you can learn:
* Wider-pitch VCSEL arrays with integrated gratings let focal length come in, split emission into individual channels and remove lenses to collapse the system.
* Partners must provide stability, low pixel outs, maximum efficiency and low pixel defects, while the coming angle out of the pixels stays the unknown that can kill a program.
* Rise and fall times need to work ideally in the gigahertz range, and two- to three-micron pixels on a ten- to fifteen-micron array are the coupling route.
đź“‹ **Clip Abstract** Waveguide design is being compressed through grating-integrated, lensless packaging built around wider-pitch VCSEL arrays, polarization-insensitive power gratings and zone gratings, with high-temperature materials widening the solution space. Partners must deliver stability, low pixel outs, low defects, maximum efficiency and controlled pixel emission angles, while rise/fall times reach the gigahertz range and two- to three-micron pixels sit on ten- to fifteen-micron arrays.
About the speaker:
* Speaker: Robert Schultz
* Company: Vuzix
* Event: Eindhoven 2026
* Location: High Tech Campus, Eindhoven
#VCSELArrays, #PowerGratings, #PixelPitch, #GigahertzRiseTimes, #ARWaveguides, #MicroDisplayPhotonics
11:20 - 13:07
Why do AR waveguides demand a uniform eyebox while CPO waveguides chase raw coupling efficiency instead?
Why do AR waveguides demand a uniform eyebox while CPO waveguides chase raw coupling efficiency instead?
AR waveguides are not CPO waveguides. AR aims to maximize uniformity in the AR glass; CPO aims to maximize efficiency. If the eyebox is not uniform, you do not have an AR product. CPO instead gets as much light in and out as possible, focusing energy to the eye. AR wants a large eyebox, no pupil expansion on the other side, and really high index. CPO can use high index but is much more worried about coupling.
One image in AR versus CPO breaks down differently: CPO focuses into multiple diodes, multiple lasers, or multiple fibers. But at the end of the day, they are all planar waveguides. The same manufacturing techniques can be used across both. The equations differ, yet they remain related. AR needs uniform output across the eyebox; CPO needs coupling into and out of the waveguide.
The waveguides are made with large, friendly eyeboxes so that when anybody puts them on, they cover the full range of people who can see the image. On the CPO side, if the optics are decoupled from the protectors, you can afford some movement. A lens incorporated in the waveguide can be fixed on one side but float on the other. That accommodates expansion in the board, twist, or shifting of the system.
In this short video, you can learn:
* AR waveguides optimize for a uniform, large eyebox; without that uniformity there is no AR product.
* CPO waveguides optimize for coupling efficiency, focusing as much light and energy as possible to the eye.
* Both AR and CPO use planar waveguides and can share the same manufacturing techniques, even though their equations and targets differ.
đź“‹ **Clip Abstract** AR and CPO waveguides share a planar platform but optimize differently: AR demands a uniform, large eyebox with high index and no pupil expansion, while CPO prioritizes coupling efficiency and energy delivery to the eye. The same manufacturing techniques can span both, and decoupled optics can allow useful movement on the CPO side.
About the speaker:
* Speaker: Robert Schultz
* Company: Vuzix
* Event: Eindhoven 2026
* Location: High Tech Campus, Eindhoven
#UniformEyebox, #PupilExpansion, #HighIndexWaveguide, #PlanarWaveguide, #ARDisplays, #OpticalInterconnect