Too much HUD content occludes the road
Aliases: HUD clutter · roadway occlusion · conformal clutter
What it is
A head-up display sits on the forward road, so more information starts to hide the world that has to be seen. Occlusion is literal: luminous graphics cover a pedestrian, a lane mark, a lead-vehicle lamp. It is also attentional: the fuller the symbols, the less the road looks like the picture and the more it looks like a backdrop. Placement keeps the eyes near the visual axis; blocking the road reverses that gain—people look “ahead” and see an interface. This entry is only about amount and density covering the road. It is not about virtual-image distance, and not about which of several cabin displays should carry what.
Why it happens
The windscreen is the driving visual channel, not another monitor. Luminance, area, and opacity of symbols cut road contrast directly. By day the HUD has to be bright enough to see, so it blocks more; at night a bright graphic makes the road darker. Conformal arrows and boxes that are solid and tightly tracked cover the very object they mark. Attentional capture joins in: moving, high-contrast layers pin gaze to the symbol, and a pedestrian in the periphery is less likely to be processed. Stacked status—speed, navigation, phone, driver assistance, music—can each be reasonable, and together they become an opaque patch on the glass.
Studying it
Treat HUD layer density as the independent variable and probe targets in the forward scene (pedestrians, lamps, sudden vehicles) for misses. The fraction of road pixels covered by graphics, and whether gaze sits on symbols or on pavement, are further measures. Contrast a speed-only layer with navigation-plus-assistance-plus-media.
Independent variables: number of simultaneous elements, area, opacity, whether graphics track moving objects, day/night luminance. Dependent variables: misses and latency for forward targets, gaze on symbol versus road, reports that the HUD “gets in the way.”
Lab scenes have designed contrast and target placement. On a real road a pedestrian steps out at the edge of a graphic; the cost of occlusion depends on where the layer sat in that frame, and a mean will understate the worst frame.
Where it stops holding
Snow, glare, and tunnel mouths already starve road contrast, so the same HUD area blocks more. Conformal augmented reality occludes little on an empty motorway and fights the scene in a dense city, where boxes and pavement want the same patch. Legally required speed or tell-tales cannot be dropped for cleanliness, but always-on is not five-items-always-on. A full-screen navigation HUD while parked is acceptable. Some drivers run brightness at maximum as decoration; occlusion then is a use problem, not only a default-layer problem.
Applying it
- In motion, default to a few symbols tied to the current manoeuvre. Detailed turn lists, album art, and social notifications stay off the glass.
- Draw conformal graphics as outlines or translucent skins, not solid blocks over the object they point at. Remove them the moment the object leaves the hazard zone.
- Let brightness follow the scene, with a cap: visible symbols, not symbols that out-compete the road.
- Verify by playing the layer back on real-road or high-fidelity footage and hunting frames where a pedestrian, a bicycle, or a lead lamp is covered. If it is covered, shrink or drop that layer—do not add a caption that says “watch the road.”