Near-Eye Display Engineering: Architecture to Supplier Transfer

Near-Eye Display Engineering: Architecture to Supplier Transfer

This general engineering note is informed by PAO engineering leadership's prior-role near-eye display experience and documented technical records. It does not disclose a confidential product, identify an individual, or claim a specific product result.

A near-eye display is not a projection lens placed near an eye. It is a coupled system containing an image source, illumination, relay or collimation optics, a combiner or waveguide, mechanical datums, electronics, calibration, and a moving human pupil. A design can look strong in an optical model and still fail as a wearable product if these interfaces are defined too late.

The most useful engineering question is therefore not, "Which optical architecture is best?" It is, "Which architecture can meet the viewing requirement inside the product, survive assembly variation, and be measured by the team that must build it?"

Start with the viewing geometry

Field of view, eyebox, eye relief, image distance, resolution, distortion, brightness, contrast, color, and see-through transmission are interdependent. They should be defined with the intended user population and frame geometry rather than as independent targets.

The requirements also need operating conditions. A display viewed indoors at a fixed pupil location creates a different brightness and stray-light problem from a display expected to remain useful outdoors while the frame moves on the face. Prescription inserts, interpupillary variation, head motion, and fit tolerance can change the usable image even when the optical module itself is unchanged.

A practical first pass establishes:

• the eye-position volume that must receive a complete image;
• the angular field and virtual-image distance;
• the spatial frequencies that carry useful content;
• acceptable distortion and color nonuniformity;
• minimum luminance and contrast in representative ambient light;
• package, power, mass, and thermal limits;
• calibration and service assumptions.

Build one system budget

The microdisplay, optical engine, waveguide or combiner, frame, and calibration algorithm should not carry separate performance promises that cannot be reconciled. A system budget connects component behavior to what the eye sees.

For image quality, that budget may include display pixel structure, relay aberration, waveguide pupil and field nonuniformity, focus error, alignment, assembly stress, thermal drift, and calibration residual. For efficiency, it may include source coupling, polarization loss, aperture clipping, relay transmission, combiner efficiency, and eyebox utilization.

The budget should also identify which errors are deterministic and calibratable, which are build variation, and which change with user fit or environment. This distinction prevents software correction from becoming an assumed solution for errors that are unstable or not observable.

Treat the mechanical datum chain as part of the optics

Near-eye display performance depends on the relationship between the optical engine, combiner, frame, and eye. The datum chain that controls those relationships belongs in the optical architecture.

Important questions include:

• Which surfaces establish the optical engine position and angle?
• How is the combiner located without inducing stress or wedge?
• What assembly adjustment is available, and how is it locked?
• Which errors can calibration remove after assembly?
• How do frame flex, hinge load, temperature, and drop events affect line of sight?
• Can the critical geometry be measured after the enclosure is closed?

If these questions are postponed until industrial design is frozen, the remaining correction options are usually expensive.

Plan prototype bring-up before ordering hardware

A prototype should answer a defined sequence of questions. A useful bring-up fixture often provides more adjustment and measurement access than the final product so the team can separate prescription performance from packaging error.

The sequence can include source and display verification, optical-engine alignment, pupil and field mapping, combiner registration, image-quality measurement across the eyebox, distortion mapping, brightness and color uniformity, and thermal drift. The measurements should use a defined camera aperture and pupil position because a near-eye display can look materially different at different points in the eyebox.

Fault isolation is easier when the team records as-built positions, module serials, calibration versions, and test conditions. Without that record, each prototype becomes an anecdote rather than engineering evidence.

Transfer starts while tolerances are still negotiable

Supplier transfer is not an administrative phase after optical design. The supplier's fabrication, assembly, metrology, and adjustment capability should shape the design before release.

The transfer package should make the following unambiguous:

• optical and mechanical datums;
• coordinate and sign conventions;
• critical-to-function dimensions and their rationale;
• component and assembly acceptance methods;
• allowed adjustment and calibration sequence;
• environmental conditions for test;
• golden units, reference data, and correlation method;
• change-control and deviation process.

Supplier reports should connect to system performance. A component can meet its drawing and still fail in the product if the drawing did not control the interface that matters.

A practical development sequence

For an early wearable program, the lowest-risk sequence is usually:

1. Define viewing geometry and environmental conditions.
2. Compare architectures with one shared performance and package budget.
3. Establish the optical-mechanical datum and calibration strategy.
4. Build an adjustable engineering prototype that supports fault isolation.
5. Correlate model, as-built data, and measured image performance.
6. Freeze supplier-ready requirements only after the measurement method is proven.
7. Transfer with controlled fixtures, reference data, and change management.

The result is not merely a smaller optical module. It is a display system whose observed performance can be explained, measured, assembled, and repeated.

PAO supports wearable and near-eye optical engineering and prototype development from architecture through supplier-ready hardware.