Your cart is currently empty!
DFSS Instructor Prep · Module 7
Layer B — Tier 2 Domain Depth · EI Strategic
HUD & Optical Display Engineering
Yazaki India’s most visible product. The Vision X AR-HUD on the Mahindra BE 6 and XEV 9e showcases Yazaki’s move up the value chain from harnessing into integrated electronic systems. Seven of your 20 participants work in EI, including the AR HUD PM, the System Engineering Lead, and the optical/mechanical designers. This module gives you working literacy in HUD optics, the PGU technology choices, the metrics that drive driver experience, and the validation methods that govern image quality.
What’s in this module
- What an automotive HUD does — the human-factors purpose
- The three HUD generations — C-HUD, W-HUD, AR-HUD
- The optical chain — PGU to virtual image, schematically
- The four optical metrics every HUD engineer talks about
- PGU technologies — TFT, DLP, LCoS, MEMS laser, holographic
- Combiner & freeform mirror design
- The windshield problem — wedged PVB & double-image
- Image quality metrics & testing
- Thermal & environmental challenges unique to HUD
- Software-hardware coupling & AR-specific challenges
- DFSS linkage — where HUD meets DMADV
- Instructor facilitation pattern
- Self-check (10 questions)
1. What a HUD does — the human-factors purpose
Before diving into optics, ground the “why”. A HUD exists to reduce the cognitive cost of driving:
- Driver no longer needs to glance at the instrument cluster — eyes stay on the road
- Reaccommodation (refocusing) cost is reduced — the further out the virtual image, the smaller the focus shift
- Critical information (speed, navigation, warnings) is overlaid on the driver’s primary visual field
- In AR-HUD, information is spatially anchored to real-world objects (lane lines, vehicles ahead, navigation arrows aligned to actual intersections)
The “reaccommodation cost” insight
The driver’s eye is focused at the road (typically 20 m+). The dashboard is at ~70 cm. Looking from road to dashboard requires the eye to refocus — taking ~0.5 second per glance. Over a long drive, this fatigue is real. A HUD with VID at 2.3 m still requires significant refocus; an AR-HUD with VID at 10–20 m brings the virtual image close to the natural road focus distance. This is the human-factors driver of the entire VID push.
2. The three HUD generations
Generation 1 — Legacy
C-HUD (Combiner)
Image source: reflected off a small plastic combiner that pops up from the dashboard.
VID: ~2 m
FOV: small
Where seen: entry-level vehicles, Mini, some MGs.
Limitation: safety concern — the combiner is a rigid object in front of the driver. Largely deprecated.
Generation 2 — Mainstream
W-HUD (Windshield)
Image source: reflected off the windshield itself, requiring a specially wedged windshield to eliminate double image.
VID: 1.8 — 2.5 m
FOV: typically < 10°
Where seen: BMW, Mercedes, Audi premium since 2010s; mainstream OEMs increasingly.
Limitation: small FOV; image floats near the bumper, not spatially anchored.
Generation 3 — The future
AR-HUD (Augmented Reality)
Image source: larger, more complex optical engine projecting onto wedged windshield.
VID: > 7 m (ideally > 10–20 m)
FOV: > 10° (ideally > 15° or 20°)
Where seen: Mercedes EQS, Audi Q6 e-tron, Hyundai Ioniq 5, Volkswagen ID family, Mahindra BE 6 / XEV 9e (Yazaki “Vision X”).
Strategic significance: the only HUD class that delivers true road-anchored information.
1.8 – 2.5 m
W-HUD typical VID (image floats near bumper)
> 10 m
AR-HUD recommended VID for road-anchored content
15° – 20°
AR-HUD target horizontal FOV (covers two lanes)
130 × 40 mm
Typical eyebox size (covers both eyes at ~65 mm separation)
3. The optical chain — PGU to virtual image
Every HUD, whatever the generation, follows the same basic path: a source image is generated, magnified and corrected by mirrors, and projected onto the windshield/combiner to form a virtual image in the driver’s view.
AR-HUD Optical Path — Schematic
Three things are happening simultaneously in this geometry:
- The PGU creates the source image (a small bright screen)
- One or more freeform mirrors magnify the image and pre-compensate for distortion the windshield will introduce
- The windshield acts as a partial reflector — and because of its specific wedge angle and the geometry, the driver’s brain interprets the reflection as an image floating beyond the windshield, at the virtual image distance
Why the windshield wedge matters
An untreated windshield has two reflecting surfaces (inner and outer glass). A flat HUD bouncing off both surfaces produces two slightly offset images — a “ghost” or double image. Vehicle-grade HUD windshields use a wedged PVB interlayer (polyvinyl butyral with varying thickness) that aligns the two reflections into a single image. This is why HUD-equipped vehicles need a HUD-specific windshield; replacements with the wrong glass produce double images.
4. The four optical metrics every HUD engineer talks about
These are the vocabulary you will hear from your EI cohort constantly. Master them.
| Metric | Definition | Typical value | Why it matters |
|---|---|---|---|
| VID — Virtual Image Distance | The apparent distance from driver’s eye to the virtual image — i.e., how far away the image “looks” to the driver | W-HUD: 1.8–2.5 m AR-HUD: > 7 m (ideally > 10–20 m) |
Larger VID = less eye refocus needed = less fatigue. For AR-HUD, must approach road focus distance to enable true overlay |
| FOV — Field of View | The angular extent of the virtual image, measured in degrees horizontal × vertical | W-HUD: < 10° (often 6° × 3°) AR-HUD: > 10° (target 15–20° H) |
Larger FOV = more content shown, can span multiple lanes. AR-HUD needs ≥ 20° H to cover lane+half on each side |
| Eyebox | The 3-D region in which the driver’s eye must be located to see the full virtual image — driver moves head outside the eyebox, image disappears or clips | 120 × 60 mm typical 130 × 40 mm common 200+ mm in premium AR |
Must accommodate driver height variation (5th to 95th percentile) and head movement. Both eyes must fit (~65 mm apart) |
| Brightness / Luminance | Image luminance in cd/m² (nits) | Daytime peak: 10,000–15,000 cd/m² Night: dimmable to ~50 cd/m² |
Must be visible against bright sky (sun visor down); dimmable for night driving to avoid dazzle |
The trade-off that defines HUD design
Larger VID + larger FOV + larger eyebox = larger packaging volume + higher cost + more demanding optics. The single biggest design challenge in HUD engineering is achieving the optical performance targets within the cramped dashboard packaging volume available. The freeform mirrors and the PGU technology choice are the levers used to manage this trade-off.
DFSS angle
VID, FOV, eyebox, and luminance are the four primary CTQs of any HUD product. Every DMADV project should map back to one or more. The CTQ tree for a HUD product centres on these metrics, propagated down to optical surface tolerances, PGU specifications, and mechanical packaging dimensions.
5. PGU technologies — TFT vs DLP vs LCoS vs MEMS vs holographic
The Picture Generation Unit is the most critical technology choice. Each technology has a signature trade-off profile.
TFT-LCD
How it works: Backlit LCD panel, transmissive.
Strengths: Mature, low cost, good colour.
Weaknesses: Limited peak brightness; backlight thermal load; lower contrast.
Dominant in: W-HUD generation; lower-end AR-HUD.
DLP (TI Digital Light Processing)
How it works: Micro-mirror array (DMD) modulates a high-intensity LED light source.
Strengths: Excellent brightness, high contrast, robust to heat.
Weaknesses: Higher cost; rainbow artefact in some implementations.
Dominant in: Premium AR-HUD (BMW iDrive 9, several Mercedes platforms).
LCoS
How it works: Liquid Crystal on Silicon — reflective LC modulator with CMOS backplane.
Strengths: High resolution; better fill-factor than TFT.
Weaknesses: Cost; speed/refresh challenges; polarisation-dependent.
Dominant in: Specialised AR-HUD & some near-eye displays.
MEMS Laser Scanning
How it works: RGB laser sources reflected off a MEMS micro-mirror that scans line-by-line to paint the image.
Strengths: Compact; very high brightness; no fixed pixel grid (resolution scales).
Weaknesses: Speckle (laser-coherence artefact); eye-safety regulatory burden.
Dominant in: Emerging in compact AR-HUDs; major automotive R&D investment.
Holographic / Diffractive
How it works: Holographic optical elements (HOEs) or computer-generated holograms project image directly into the eyebox.
Strengths: Compact volume; can support multi-distance VID natively.
Weaknesses: Most complex; image quality / efficiency still maturing.
Dominant in: Future-generation AR-HUD (research stage, some pilot platforms).
Yazaki / Vision X context
The Vision X AR-HUD on Mahindra BE 6 / XEV 9e demonstrates a specific PGU + freeform mirror choice optimised for Indian conditions. The exact technology is proprietary, but the design space is the one above. Knowing the field lets you ask intelligent questions about a participant’s PGU choice and trade-offs.
6. Combiner & freeform mirror design
Between the PGU and the windshield (or combiner) sit one or more mirrors. These mirrors have two jobs: magnify the small PGU image to fill the desired FOV, and pre-distort the image to cancel out the optical distortion the curved windshield will introduce.
- Flat fold mirrors — used to bend the optical path into the available packaging volume
- Aspheric mirrors — single-axis curvature, modest magnification
- Freeform mirrors — non-rotationally-symmetric surfaces, optimised numerically for each specific windshield curvature. Required for AR-HUD-class performance.
Freeform mirror design is done with optical design software:
Ansys OpticStudio (Zemax) — primary optical CAD
Ansys Speos — photometric / image-quality analysis
CODE V (Synopsys)
LightTools — illumination & eyebox analysis
The freeform manufacturing challenge
Freeform mirrors must be manufactured to optical-grade surface accuracy (typically λ/4 — fraction of a wavelength). This is the boundary between automotive precision and optical precision. Diamond-turned masters; injection-moulded polymer or vacuum-metallised mirrors in production. Tolerance design (DMADV Design phase) on these surfaces is a substantial DFSS project in its own right.
7. The windshield problem — wedged PVB & double-image
The windshield is not just a passive surface — it is an optical element that the HUD must collaborate with. Several quirks make this hard.
| Problem | Cause | Solution |
|---|---|---|
| Double image (ghost) | Windshield has two glass surfaces, each reflects ~4% of light, producing two offset images | Wedged PVB interlayer with varying thickness; aligns the two reflections into one |
| Variable curvature | Windshield is not flat, has compound curvature varying across vehicle width | Freeform mirror in HUD pre-compensates; must be co-designed with windshield supplier |
| Solar load on dash | HUD aperture in dashboard is a sun trap; PGU can overheat from concentrated sun | Solar-load shutter; thermal design; PGU heat management |
| Polarisation interaction | Windshield interacts differently with different light polarisations; LCD output is polarised | Polarisation control in PGU optics; sometimes circular polarisation |
| HUD-specific windshield required | Replacement with non-wedged glass produces visible double image | Service procedure must mandate HUD-grade replacement |
A real-world implication
A HUD product DFMEA must include “wrong-windshield replacement” as a usage scenario. Field complaints of “ghost image after windshield repair” are a known warranty pattern. The mitigation is upstream — clear service labels and dealer-tech training — but the failure mode must appear in the DFMEA.
8. Image quality metrics & testing
HUD validation is a specialised discipline. The metrics are the language of the discipline.
8.1 The IQ metrics that get measured
| Metric | What it captures |
|---|---|
| Luminance & uniformity | Peak brightness across the FOV; how uniformly bright the image is corner-to-corner. Typical spec: > 80% uniformity over usable FOV |
| Contrast ratio | Bright pixel : dark pixel ratio under controlled ambient light. Spec depends on test conditions; > 100:1 is typical |
| MTF (Modulation Transfer Function) | How well the optical system preserves spatial frequencies — i.e., sharpness. Measured at standardised spatial frequencies (e.g., 10 lp/mm) |
| Distortion | Geometric error in the projected image — straight lines should remain straight. Pin-cushion / barrel distortion typical issues. Must stay below ~2% |
| Colour accuracy / gamut | Colour reproduction vs reference. White-point tolerance, ΔE measurements |
| Static image position | Where the virtual image actually appears vs where it should — important for AR overlay accuracy |
| Dynamic / motion artefacts | Latency, jitter, frame drops — particularly important for AR overlays moving with vehicle |
| Eyebox limits | How the image quality degrades as the driver moves away from eyebox centre |
8.2 The testing equipment landscape
Radiant Vision Systems — ProMetric imaging colorimeter, TT-HUD software
OptoFidelity — Ronin AR-HUD test system
Konica Minolta — luminance & colour meters
Custom optical benches per OEM CSR
The standard test setup: imaging colorimeter on a robotic arm or moving stage, replicating the human eye’s position throughout the eyebox, capturing image quality at every position. For dynamic AR-HUD, the test must include time-synchronised content streams.
For the Testing Center Manager in your room
HUD validation requires a substantially different test capability from harness/connector testing. Optical-class environmental cleanliness, controlled ambient lighting, calibrated imaging photometry. If your Testing Center hasn’t built HUD-specific capability yet, the AR HUD PM has — they’re a natural collaboration point.
9. Thermal & environmental challenges unique to HUD
Several stressors are unique to HUD products and don’t appear in conventional cabin electronics.
- Solar concentration: The HUD aperture in the dashboard can act as a passive solar concentrator — sun rays focused down onto the PGU. Some HUDs use a shutter; others rely on solar-load thermal design. Internal temperatures can reach 100 °C+ in a parked car.
- Wide ambient temperature range: -40 °C to 105 °C typical automotive grade; PGU technologies have different temperature limits (TFT-LCD struggles in cold, lasers in heat)
- Optical alignment vs thermal expansion: Plastic mirrors / housings expand with temperature, causing image position shift. Athermalisation is a key design discipline.
- Vibration: Multi-mirror optical paths are sensitive to vibration. Modal analysis and resonance avoidance are critical.
- EMI: HUDs have high-speed video links (LVDS, MIPI, FPD-Link III, GMSL) — must meet CISPR 25 / OEM EMI specs.
- Dust & humidity: The optical path must remain clean; sealing the optics module is non-trivial because it must also breathe to avoid condensation.
Linking to Module 2 mechanisms
Module 2’s failure mechanisms appear in HUD-specific forms: polymer ageing (yellowing of plastic mirrors → reduced reflectance); seal compression set (optics-housing seal hardens → dust ingress → reduced contrast over years); solder fatigue on the PGU driver board (thermal cycling); fretting on LVDS / FPD-Link III connectors.
10. Software-hardware coupling & AR-specific challenges
An AR-HUD is as much a software product as an optical one. Several AR-specific challenges appear that don’t exist in conventional W-HUD.
- Latency budget: Sensor input (camera/radar) → fusion → rendering → display → eye = total < ~100 ms for plausible AR overlay. Each stage must hit its allocation.
- Calibration: The vehicle ADAS knows where objects are in world coordinates. The HUD must render at the correct image pixel so the overlay aligns visually with the real object. Requires precise eye-position estimation (camera-based driver monitoring), world-to-image mapping, and continuous self-calibration.
- Multi-distance content: Some HUD content (speed) belongs near-field; navigation arrows belong far-field. Multi-VID architectures emerging — dual-projector or holographic approaches.
- Driver attention assumptions: Wrong overlay (e.g., misaligned arrow) is worse than no overlay — actively misleads the driver. Functional safety implications (ISO 26262).
- 3D AR-HUD (research): Parallax-barrier or light-field approaches deliver depth cues (3D depth range 1–20 m). Requires precise eye-tracking. Emerging in research-grade demonstrators.
A safety subtlety the room must recognise
AR-HUD content that misleads the driver — for example, a navigation arrow pointing into the wrong lane — has direct safety implications. This means AR-HUD systems often carry ASIL-B classification (occasionally higher) under ISO 26262. Software development must be ASPICE-compliant. The AGM-EI software lead in your room owns this challenge.
11. DFSS linkage — where HUD meets DMADV
| DMADV Phase | HUD content that lands here |
|---|---|
| Define | Generation (W-HUD vs AR-HUD), VID target, FOV target, eyebox spec, luminance spec, packaging volume budget, CSR alignment, ASIL classification per ISO 26262 |
| Measure | CTQs: VID accuracy (e.g., target 10 m ±0.5 m); FOV ≥ 15° H; eyebox ≥ 120×60 mm; peak luminance ≥ 12,000 cd/m²; MTF > 0.3 at standardised frequency; distortion < 2%; AR overlay latency < 100 ms; static image position accuracy < 0.5° |
| Analyze | Concept selection across PGU technologies (TFT vs DLP vs LCoS vs MEMS laser); optical architecture (single mirror vs multi-mirror vs holographic); freeform mirror count. DFMEA against Module 2 mechanisms (solder fatigue, polymer ageing, seal degradation) |
| Design | Tolerance design on freeform mirror surface accuracy; PGU power-vs-thermal trade; mechanical packaging; athermalisation. P-diagram noise factors: ambient temperature, solar load, vibration spectrum, supply-voltage variation, EMI |
| Verify | Image-quality testing (luminance, contrast, MTF, distortion) across eyebox & temperature range; AR-overlay accuracy testing per ADAS data; ALT under solar load; CISPR 25 EMI testing; vibration / shock per ISO 16750; AIS 156 if EV platform; functional safety verification per ISO 26262 |
A complete HUD-product worked example
Consider the Vision X-class AR-HUD design for an Indian premium EV (BE 6-class platform):
- CTQs: VID 10 m (±0.5), FOV 15° H × 5° V, eyebox 130×60 mm, peak luminance 12,000 cd/m², AR overlay latency < 80 ms
- Architecture: DLP PGU (high brightness for Indian solar), two freeform mirrors (athermalised metal-substrate), wedged windshield co-developed with glass supplier
- Top DFMEA modes: Image position drift due to freeform mirror thermal expansion (M2 polymer ageing analog); reduced contrast due to dust ingress past optics seal (M2 seal compression set); double image after windshield replacement (M7 §7); LVDS connector intermittent due to fretting (M5 §4)
- Verification: Radiant ProMetric across temperature -30 to +85 °C; AIS 156 vehicle-level testing; CISPR 25 EMI; ALT against solar-load profile representative of Indian conditions
12. Instructor facilitation by function
| Function | HUD angle that lands |
|---|---|
| EI — AR HUD Project Manager | This is their world. Treat as the cohort SME for HUD. Engage on multi-VID, AR overlay accuracy, ASIL targets, Vision X learnings. |
| EI — System Engineering Lead | System integration: ADAS sensors → fusion → HUD render pipeline. Latency budget allocation. Calibration architecture. |
| EI — AGM-EI Software Lifecycle (SGM) | ASPICE compliance, ISO 26262 ASIL-B/C decomposition for AR overlay, OTA strategy for HUD software updates. |
| EI — Optical / Mechanical Designer | Freeform mirror design, athermalisation, packaging-volume engineering, tolerance design on optical surfaces. |
| EI — Sensor Developer | Driver monitoring camera for eye-tracking (eyebox positioning, 3D AR), forward camera for AR overlay registration. |
| EI — Innovation Cell / Tech Asst | Holographic / waveguide research, 3D AR-HUD, dynamic VID. Future-product trajectory. |
| Shared Service — Thermal/EMI/CFD | Solar-load thermal simulation; LVDS/FPD-Link III EMI compliance; vibration modal analysis. |
| Shared Service — Advance Materials | Mirror substrate & coating selection; polymer ageing for plastic optics; PVB windshield material spec. |
| Testing Center Manager | Optical-class test capability is different from harness/connector test. Capacity to validate HUDs to OEM CSR. |
| WH / CDDC participants | HUDs need a wiring harness with LVDS/FPD-Link III connector; HV cables in EV context. Their harness must meet HUD-grade EMI & signal-integrity specs. |
| SD Coordination / Project Mgmt | HUD programmes have longer optical-development cycles, distinct supplier chains (glass supplier, optical CAD partners), and more critical co-engineering with OEM than typical harness work. |
A high-leverage question for any HUD project
“What’s your VID target, your FOV target, and where in the eyebox does your image quality degrade first under thermal extreme?” A team that can answer all three is operating at HUD-engineering rigour. A team that answers only one or two has gaps to close.
A second framing question for the AR-specific cohort
“What’s your end-to-end AR overlay latency budget — from ADAS sensor capture to photon hitting driver’s eye — and which subsystem owns the largest slice?” Forces system-level thinking instead of subsystem-local optimisation.
Instructor self-check
Ten questions calibrated to the level of HUD-engineering conversation you’ll be in.
Q1. A participant says their HUD design has a VID of 2.3 m. The most important question to ask next is:
A. “Why isn’t it 1 m?”
B. “Is that in inches?”
C. “Is this a W-HUD or an AR-HUD? For AR-HUD, 2.3 m is too close — the road-anchored content won’t work; target should be > 7 m, ideally > 10–20 m”
D. “What colour is it?”
Correct — 2.3 m is canonical W-HUD; for an AR-HUD product it would be a significant design gap. The distinction is foundational.
Q2. Why does a HUD-equipped vehicle need a special windshield?
A. To absorb UV
B. A wedged PVB interlayer of varying thickness aligns the inner-and-outer-surface reflections into a single image, preventing “ghost” double-image
C. To carry the HVIL signal
D. To be tinted orange
Correct — the wedged-PVB windshield is the optical partner to the HUD. Replacement with non-HUD glass produces visible double image.
Q3. The eyebox dimension matters because:
A. It limits the image colour
B. It defines the windshield curvature
C. It controls the PGU brightness
D. The driver’s eye must stay within the eyebox to see the full image; eyebox must accommodate driver-height variation (5th–95th percentile) and both eyes (~65 mm apart)
Correct — eyebox size is a primary ergonomic CTQ. Too small and shorter or taller drivers see clipped images.
Q4. The PGU technology dominant in premium AR-HUD due to its peak brightness and contrast is:
A. DLP (TI Digital Light Processing) — micro-mirror array modulating a high-intensity LED
B. CRT
C. Plasma
D. E-ink
Correct — DLP dominates premium AR-HUD because of its brightness and contrast. TFT-LCD is the volume choice for W-HUD; MEMS laser is emerging for compact AR.
Q5. Freeform mirror design is necessary in modern AR-HUD because:
A. Flat mirrors don’t reflect light
B. The freeform surface pre-distorts the image to cancel out distortion the curved windshield will introduce — and provides the magnification needed within the cramped dashboard volume
C. Freeform mirrors are cheaper
D. Required by regulation
Correct — freeform mirrors are the central optical-engineering tool that makes AR-HUD performance possible in production packaging.
Q6. An AR-HUD navigation arrow appears slightly offset from the actual lane. The most likely root cause is:
A. The HUD bulb has failed
B. The vehicle is in reverse
C. Calibration error in the world-to-image mapping, OR eye-position estimation error, OR ADAS sensor calibration drift — overlay accuracy depends on the entire calibration chain
D. The mirror is dirty
Correct — AR-overlay accuracy is a system-wide property. A misaligned overlay is worse than no overlay because it misleads the driver — direct functional-safety implication.
Q7. The total end-to-end latency budget from ADAS sensor capture to photon-in-driver’s-eye for plausible AR overlay is typically:
A. A few seconds
B. Under ~100 ms (often ~80 ms targeted)
C. Below 1 microsecond
D. Latency doesn’t matter
Correct — AR overlay must keep up with vehicle motion. ~100 ms total budget partitioned across sensing, fusion, rendering, display.
Q8. The HUD aperture in the dashboard creates a unique thermal failure mode:
A. The PGU runs too cold
B. Solar concentration through the aperture focuses sun rays onto the PGU, potentially heating it past 100 °C when parked — solar-load shutter or thermal design required
C. The PGU floods
D. The PGU emits sound
Correct — the aperture is essentially a solar concentrator. Indian solar conditions make this particularly aggressive. Often addressed with a mechanical shutter or thermal-design margin.
Q9. The optical design software ecosystem typically used for HUD design includes:
A. Microsoft Word
B. AutoCAD only
C. Spreadsheets only
D. Ansys OpticStudio (formerly Zemax) for optical design and Ansys Speos for photometric/image-quality analysis; CODE V is also used
Correct — OpticStudio + Speos is the dominant chain for HUD optical design and validation in production.
Q10. The most diagnostic three-part question for any HUD DFSS project is:
A. What colour, what shape, what price?
B. Domestic or imported?
C. What’s your VID target, your FOV target, and where in the eyebox does image quality degrade first under thermal extreme?
D. Tin, silver, or gold?
Correct — these three questions span optical specification (VID), display extent (FOV), and stress robustness (eyebox under thermal stress). Together they test whether a team is operating at HUD-engineering rigour.