The "Optical Heart" of AR Glasses: Why Is It So Hard to Make?
2026-9-14
A Nanoscale War Inside a Pair of Glasses
If you take apart a pair of AR glasses, the most valuable component isn't the display panel, nor the chip inside—it's the seemingly ordinary lens right in front of your eyes.
Hidden within this lens is an "optical heart"—the optical waveguide. Its function sounds simple: to transport the image produced by the optical engine to your eyes through total internal reflection. But it's precisely this "transport" process that has given the world's top optical engineers headaches for over a decade.
Field of View Stuck at 50°: It's Not a Lack of Will, It's Physics
The field of view (FOV) of AR glasses determines how large the virtual image you see can be. But the FOV of mainstream glass waveguides has consistently been stuck between 40° and 50°.
The reason is simple: the angular range of total internal reflection is determined by the material's refractive index. Glass has a refractive index of only 1.5 to 1.7—the ceiling is right there, and no amount of engineering optimization can get around it.
Silicon carbide has a refractive index of 2.7, which in theory could push the FOV all the way to 70°. But at what cost? Silicon carbide's hardness is second only to diamond, and processing it requires semiconductor-grade equipment. A single lens costs 2,000 to 3,000 yuan—three to four times that of an ordinary lens.
Nanoscale "Etching" and "Stacking": Every Step Is a Yield Killer
Geometric waveguides rely on "stacking"—precisely layering multiple semi-transmissive, semi-reflective film systems with alignment accuracy down to the micron level. After stacking, image uniformity may be less than 60%, with dark edges and a bright center.
Diffractive waveguides rely on "etching"—carving gratings with a period of a few hundred nanometers onto the lens surface. If a single speck of dust lands on it, it gets illuminated and magnified during imaging, becoming a visible defect to the naked eye.
Producing one good sample in the lab is one thing. Continuously producing tens of thousands of units on a production line, each meeting specifications, is another matter entirely.
The Better the Performance, the Harder to Manufacture
This industry has a cruel rule: the solutions with the best optical performance are often the hardest to mass-produce.
Geometric waveguides offer uniform color and no rainbow artifacts, but the multi-layer bonding process has extremely low yields. Diffractive waveguides are thin, lightweight, and highly transparent, but light efficiency can衰减 to less than 1%, putting enormous pressure on power consumption and heat dissipation.
Silicon carbide waveguides attempt to break through. Some teams have used nanoimprint lithography to bypass the difficulty of etching silicon carbide, achieving mass production on 4-inch wafers, with light efficiency improved by 72% over mainstream products and rainbow artifacts essentially eliminated. But silicon carbide grinding costs four times that of glass, and controlling warpage on 8-inch substrates remains a challenge.
The Forging of the Heart Determines the Pulse of AR
Optical waveguides are hard to make because they must simultaneously satisfy a set of conflicting requirements: large field of view, thinness and lightness, high light efficiency, color uniformity, low cost, and high yield.
Each one individually has a solution, but bundling them all together into a pair of glasses becomes a systems engineering nightmare.
The difficulty of forging this "optical heart" is, at its core, humanity's ultimate test of manipulating light at the nanoscale. When AR glasses truly become mainstream, the hardest part won't be making virtual images appear—it will be making it all look as if nothing ever happened.
If you take apart a pair of AR glasses, the most valuable component isn't the display panel, nor the chip inside—it's the seemingly ordinary lens right in front of your eyes.
Hidden within this lens is an "optical heart"—the optical waveguide. Its function sounds simple: to transport the image produced by the optical engine to your eyes through total internal reflection. But it's precisely this "transport" process that has given the world's top optical engineers headaches for over a decade.
Field of View Stuck at 50°: It's Not a Lack of Will, It's Physics
The field of view (FOV) of AR glasses determines how large the virtual image you see can be. But the FOV of mainstream glass waveguides has consistently been stuck between 40° and 50°.
The reason is simple: the angular range of total internal reflection is determined by the material's refractive index. Glass has a refractive index of only 1.5 to 1.7—the ceiling is right there, and no amount of engineering optimization can get around it.
Silicon carbide has a refractive index of 2.7, which in theory could push the FOV all the way to 70°. But at what cost? Silicon carbide's hardness is second only to diamond, and processing it requires semiconductor-grade equipment. A single lens costs 2,000 to 3,000 yuan—three to four times that of an ordinary lens.
Nanoscale "Etching" and "Stacking": Every Step Is a Yield Killer
Geometric waveguides rely on "stacking"—precisely layering multiple semi-transmissive, semi-reflective film systems with alignment accuracy down to the micron level. After stacking, image uniformity may be less than 60%, with dark edges and a bright center.
Diffractive waveguides rely on "etching"—carving gratings with a period of a few hundred nanometers onto the lens surface. If a single speck of dust lands on it, it gets illuminated and magnified during imaging, becoming a visible defect to the naked eye.
Producing one good sample in the lab is one thing. Continuously producing tens of thousands of units on a production line, each meeting specifications, is another matter entirely.
The Better the Performance, the Harder to Manufacture
This industry has a cruel rule: the solutions with the best optical performance are often the hardest to mass-produce.
Geometric waveguides offer uniform color and no rainbow artifacts, but the multi-layer bonding process has extremely low yields. Diffractive waveguides are thin, lightweight, and highly transparent, but light efficiency can衰减 to less than 1%, putting enormous pressure on power consumption and heat dissipation.
Silicon carbide waveguides attempt to break through. Some teams have used nanoimprint lithography to bypass the difficulty of etching silicon carbide, achieving mass production on 4-inch wafers, with light efficiency improved by 72% over mainstream products and rainbow artifacts essentially eliminated. But silicon carbide grinding costs four times that of glass, and controlling warpage on 8-inch substrates remains a challenge.
The Forging of the Heart Determines the Pulse of AR
Optical waveguides are hard to make because they must simultaneously satisfy a set of conflicting requirements: large field of view, thinness and lightness, high light efficiency, color uniformity, low cost, and high yield.
Each one individually has a solution, but bundling them all together into a pair of glasses becomes a systems engineering nightmare.
The difficulty of forging this "optical heart" is, at its core, humanity's ultimate test of manipulating light at the nanoscale. When AR glasses truly become mainstream, the hardest part won't be making virtual images appear—it will be making it all look as if nothing ever happened.
