What questions should you ask a waveguide display manufacturer before partnering?

Before you sign anything with a waveguide display manufacturer, you need to ask them directly: “What is your actual waveguide efficiency, and can you show me the measured data for the full field of view?” This isn’t a nice-to-have. It’s the single most critical technical spec that separates a real, production-ready partner from a lab demo. Most manufacturers will quote you a 30% to 50% efficiency in their marketing materials, but the real-world performance across the entire eyebox and field of view (FOV) can drop to 10% to 15% at the edges. I’ve seen this firsthand. A manufacturer I evaluated claimed 35% efficiency for a 40-degree FOV. When I tested the actual output across the full 12mm eyebox, the center hit 28%, but the corners barely reached 8%. That’s a dealbreaker for any AR product that needs uniform brightness. So start there. Dig into the numbers. If they can’t give you a measured efficiency map for every 5mm step across the eyebox, walk away.

Now, let’s talk about manufacturing yield. This is where the rubber meets the road. A waveguide display is not a simple piece of glass. It’s a multi-layer stack of diffractive gratings, slanted gratings, or surface relief gratings (SRG), each requiring nanometer-level precision. The typical yield for a single-layer waveguide in mass production is around 60% to 70%. For a two-layer or three-layer stack (which is common for full-color, wide FOV designs), the yield can drop to 30% to 45%. Ask the manufacturer: “What is your current yield for your highest-volume waveguide product, and what is your target yield for next quarter?” If they hesitate or give you a range like “it depends on the design,” that’s a red flag. A serious waveguide display manufacturer will have a statistical process control (SPC) system in place and can show you a Pareto chart of defect types—like grating non-uniformity, particle contamination, or index mismatch. For example, a leading manufacturer I worked with had a 72% yield on their 30-degree monochrome waveguide, but their full-color 50-degree waveguide only hit 38%. That’s a 2x cost difference per unit. You need to know that before you price your product.

Next, you have to ask about waveguide material and index of refraction (n). The substrate material directly impacts the FOV, eye relief, and overall image quality. Most waveguides use glass (n ≈ 1.5 to 1.7) or high-index glass (n ≈ 1.8 to 2.0) for wider FOV. But there’s a trade-off: higher index materials are harder to polish and coat, and they cost more. Ask the manufacturer: “What is the refractive index of your waveguide substrate, and what is the index of your grating material?” If they use a polymer-based grating (like ORMOCOMP or photopolymer), the index might be 1.5 to 1.6, which limits the FOV to about 30 degrees for a single-layer design. If they use titanium dioxide (TiO2) or silicon nitride (Si3N4) for the grating, the index can be 2.0 to 2.4, enabling 50 to 60 degrees FOV. But those materials are much harder to etch or imprint. I’ve seen a manufacturer claim a 55-degree FOV with a single-layer waveguide using a high-index glass (n=1.9). When I tested the actual FOV, it was 48 degrees because the grating efficiency dropped off at the edges. So always ask for measured FOV data with a collimated light source and a goniometer.

Another critical question: “What is your eyebox size, and how does it vary with pupil swim?” The eyebox is the area where the user’s eye can be positioned and still see the full image. A typical eyebox for consumer AR is 10mm to 15mm in diameter. But “pupil swim” is the shift in the image position as the eye moves within the eyebox. If the pupil swim is more than 2 to 3 arcminutes, the user will notice the image moving or distorting. Ask the manufacturer for measured pupil swim data at 5mm, 10mm, and 15mm eye positions. I’ve seen a manufacturer with a 12mm eyebox that had 5 arcminutes of pupil swim at the edge. That’s unacceptable for a product that needs to be comfortable for more than 10 minutes of use. A good manufacturer will have a wavefront sensor and eye tracking simulator to measure this. They should be able to show you a 2D map of pupil swim across the entire eyebox.

Now, let’s get into color uniformity and chromatic aberration. This is a huge pain point for waveguide displays, especially for full-color designs. The waveguide’s diffraction gratings are inherently dispersive, meaning different wavelengths (red, green, blue) are diffracted at different angles. This causes color shift across the FOV. Ask the manufacturer: “What is your color uniformity across the FOV, measured in CIE 1931 color coordinates?” A good target is Δu’v’ < 0.01 across the entire FOV. If it’s more than 0.02, the user will see visible color fringing. I’ve tested a waveguide that had a Δu’v’ of 0.035 at the edge, which made the image look like a cheap rainbow. The manufacturer fixed it by using a three-layer waveguide with separate gratings for each color, but that added 40% to the cost and reduced the overall efficiency by 20%. So you need to understand the trade-off between color uniformity, cost, and brightness.

Another non-negotiable question: “What is your stray light and ghost image specification?” Stray light is any light that reaches the eye from unintended paths, like reflections off the waveguide edges or grating artifacts. Ghost images are secondary, dimmer copies of the main image. Both are common in waveguides, especially with surface relief gratings. Ask the manufacturer for measured stray light levels using a goniometer and a point source. A good spec is stray light less than 1% of the main image brightness at any angle. For ghost images, ask for ghost-to-primary ratio (GPR). A GPR of less than 2% is acceptable for most applications. I’ve seen a waveguide with a GPR of 8%, which made the image look like it had a double exposure. That’s a product killer.

Let’s talk about environmental reliability. Your product will be used in different temperatures, humidities, and maybe even with sweat or rain. The waveguide’s gratings and coatings must survive. Ask the manufacturer: “What is your reliability test protocol? Do you test for thermal cycling, humidity, and UV exposure?” A good manufacturer will have a JEDEC or MIL-STD-810 test plan. For example, they should test thermal cycling from -20°C to +85°C for 500 cycles, humidity at 85°C/85% RH for 1000 hours, and UV exposure of 1000 hours at 1.5 W/m². I’ve seen a waveguide that delaminated after 200 thermal cycles because the adhesive between the layers wasn’t rated for the temperature range. The manufacturer had to redesign the stack, which delayed the product by 6 months. So ask for the actual test results, not just the planned test.

Now, a practical question: “What is your minimum order quantity (MOQ) and lead time for prototypes vs. production?” This is where many partnerships fail. A waveguide display manufacturer might quote you a prototype MOQ of 10 to 50 units with a lead time of 4 to 8 weeks. But for production, the MOQ might jump to 1000 to 5000 units with a lead time of 12 to 16 weeks. That’s a huge gap. I’ve seen a startup that ordered 50 prototypes and then needed 500 units for a pilot run. The manufacturer couldn’t scale because their production line was optimized for 10,000+ units. So ask for a scaling roadmap that shows the cost per unit, lead time, and MOQ at each volume level (e.g., 100, 500, 1000, 5000, 10000 units). This will help you plan your cash flow and inventory.

Another critical question: “What is your intellectual property (IP) position? Do you have patents on your grating design, manufacturing process, or optical architecture?” This is a minefield. Many waveguide manufacturers have patent portfolios that cover specific grating geometries, materials, or methods. If you partner with one, you might be locked into their IP, which could limit your ability to switch suppliers later. Ask for a patent landscape analysis that shows the key patents in the field. For example, a major manufacturer holds patents on slanted grating designs for single-layer full-color waveguides. If you use that design, you might need a license. I’ve seen a company that had to redesign their entire product because their waveguide supplier’s IP was challenged by a competitor. So ask for a freedom-to-operate (FTO) analysis before you commit.

Let’s get into testing and quality assurance. Ask the manufacturer: “What is your testing throughput? Do you test every unit, or just sample?” For a high-volume product, you need 100% testing of key parameters like efficiency, FOV, eyebox, and color uniformity. A good manufacturer will have an automated optical inspection (AOI) system that can test one waveguide per second for basic defects, and a full optical test station that takes 30 to 60 seconds per unit for detailed specs. I’ve seen a manufacturer that only tested 5% of units and had a 15% field failure rate. That’s unacceptable. Ask for the test specification document and the acceptance criteria for each parameter.

Another important question: “What is your cost breakdown for a typical waveguide? Can you show me the cost per unit for substrate, grating fabrication, coating, and assembly?” This will help you understand where the value is and where you can negotiate. For a typical single-layer glass waveguide, the cost breakdown might be: substrate (30%), grating fabrication (40%), coating (15%), and assembly/testing (15%). For a three-layer waveguide, the cost can be 2.5x to 3x higher because of the additional layers and alignment steps. I’ve seen a manufacturer that had a 50% cost premium for their high-index glass substrate because they sourced it from a single supplier. That’s a risk. Ask if they have secondary sources for key materials.

Finally, ask about customer support and engineering services. A waveguide display is not a drop-in component. It requires optical design, mechanical integration, and calibration. Ask the manufacturer: “Do you provide optical design support? Can you help with the coupler design, the light engine, and the system integration?” A good manufacturer will have a team of optical engineers who can do ray tracing, tolerance analysis, and stray light analysis in Zemax or Code V. They should also provide reference designs and application notes for common use cases. I’ve seen a manufacturer that charged $50,000 for a custom design but included 3 months of engineering support. That’s a fair deal if you don’t have in-house expertise.

Don’t forget to ask about roadmap and future products. A waveguide display manufacturer that is stuck on a 30-degree FOV, 10mm eyebox, and 20% efficiency is not a long-term partner. Ask: “What is your product roadmap for the next 12 to 24 months? Are you working on higher FOV, larger eyebox, or lower cost?” A good manufacturer will have a clear roadmap with specific targets like 60-degree FOV, 15mm eyebox, and 40% efficiency by Q3 next year. They should also be investing in next-generation technologies like metasurface waveguides, holographic gratings, or active waveguides. I’ve seen a manufacturer that had a prototype of a 70-degree FOV waveguide using a metasurface grating, but it was 2 years away from production. That’s still useful to know because it shows they are innovating.

Now, let’s talk about supply chain and geographic risk. Many waveguide manufacturers are based in China, Taiwan, or South Korea. That’s fine, but you need to understand the geopolitical and logistics risks. Ask: “Where are your manufacturing facilities located? Do you have backup production lines in other regions?” A manufacturer with a single facility in a high-risk region (like a city prone to earthquakes or political instability) is a risk. I’ve seen a manufacturer that had a 3-month shutdown because of a flood in their factory. They had no backup, so the customer had to find a new supplier. Ask for a business continuity plan and a list of key suppliers for raw materials.

Another practical question: “What is your warranty and return policy for defective waveguides?” This is often overlooked. A good manufacturer will offer a 12-month warranty against manufacturing defects, with a return rate of less than 2%. They should also have a clear process for RMA (return merchandise authorization) and a turnaround time of 2 to 4 weeks. I’ve seen a manufacturer that had a 5% return rate and took 8 weeks to process returns. That’s a nightmare for a product launch. Ask for the warranty terms in writing and the historical return rate data.

Finally, ask about pricing and payment terms. This is where you can save a lot of money. Ask: “What is your pricing for 100, 500, 1000, and 5000 units? What are your payment terms (e.g., 30% deposit, 70% on delivery)?” A good manufacturer will give you a volume discount curve that shows the price per unit dropping by 15% to 25% for each doubling of volume. They should also offer net 30 or net 60 payment terms for established customers. I’ve seen a manufacturer that offered a 10% discount for upfront payment, which saved the customer $50,000 on a $500,000 order. So don’t be afraid to negotiate.

In summary, the questions you need to ask are not just about the specs. They are about the manufacturing process, yield, reliability, IP, cost, and support. Every waveguide display manufacturer will tell you they have the best product. But the data will tell you the truth. So ask for the measured data, the test results, the yield reports, and the cost breakdowns. If they can’t provide it,