Spatial UX: Designing for Apple Vision Pro, Mixed Reality, and the Interfaces That Have No Screen

For forty years, digital design has meant designing rectangles. Screens of different sizes, sure. Landscape and portrait. Retina and standard. But always flat, always bounded, always contained within a frame. Spatial computing removes the frame. The interface is no longer on a surface. It is in the room with you. And almost every UX convention we have built our discipline around stops working the moment that happens.

This Is Not a Gaming Story

When people hear about mixed reality headsets, the first association is usually games. That framing has kept the technology at arm's length from most product teams, who reasonably concluded that spatial computing was not their problem to solve yet.

That conclusion is becoming harder to maintain. Apple Vision Pro launched in early 2024 and has been expanding to more countries through 2025 and 2026. Analysts tracking the AR and VR headset market project 87% growth in 2026 following a temporary contraction in 2025. Forty percent of Global 2000 manufacturers are already maintaining production digital twins. Enterprise applications across healthcare, architecture, logistics, remote collaboration, and financial visualization are in production. visionOS 26, announced by Apple in mid-2026, introduced new spatial experiences that extend Vision Pro's capabilities significantly and unified the design language across iPhone, iPad, Mac, and Vision Pro into a single coherent system.

The organizations treating spatial computing as a future concern will find themselves designing for it under time pressure when that future arrives. The ones building spatial design literacy now will have a genuine advantage. This post is a starting point for the second group.

What Spatial Computing Actually Is

Spatial computing is computing that exists in three-dimensional space rather than on a flat surface. The user interacts with digital content that is anchored to, or overlaid on, their physical environment. Instead of looking at a screen, they look through one or wear one, with digital elements appearing to exist in the world around them.

The term covers three overlapping experiences that are worth distinguishing precisely.

Augmented Reality (AR)

Digital content is overlaid on the real world. The user can see their physical environment clearly, with digital elements appearing on top of it. AR glasses, smartphone AR, and the passthrough mode on Vision Pro all fall into this category. The physical world is the primary canvas. Digital elements are additions to it.

Mixed Reality (MR)

Digital objects interact with the physical environment in a more integrated way. A virtual object might appear to sit on a real table, cast a shadow on a real floor, or be occluded by a real wall. Vision Pro's default operating mode is mixed reality. The distinction from AR is the degree of integration: in mixed reality, digital and physical elements genuinely interact with each other rather than simply coexisting on the same visual plane.

Virtual Reality (VR)

The physical environment is fully replaced by a digital one. The user sees only what the headset renders. Meta Quest's immersive mode and Vision Pro's full immersion mode fall into this category. The physical world is temporarily absent. The digital environment is everything.

Apple's particular contribution to the spatial computing conversation is a design philosophy that treats the transition between these states not as a mode switch but as a continuous spectrum. visionOS lets applications exist as flat windows in a shared space alongside other apps, as three-dimensional volumes that users can view from any angle, or as fully immersive environments. The design challenge is deciding which state serves each moment in the user experience, and making the transitions between them feel natural rather than jarring.

Why Flat-Screen UX Conventions Break in Spatial Contexts

This is the part that most design teams underestimate. It is tempting to assume that spatial computing is just another form factor, the way desktop to mobile was a form factor transition. The same principles apply, just in a new shape.

That assumption is wrong. The peer-reviewed research on spatial interface design is consistent on this point. A Frontiers in Computer Science study published in June 2025, examining spatial UX for Apple Vision Pro content creation tools, found that users' interaction logic in XR environments differs fundamentally from desktop environments. Not incrementally. Fundamentally. The specific conventions that break down are worth understanding one by one.

The cursor disappears

Every flat-screen interaction model is built around a pointing device. Mouse, finger, stylus: the user moves a cursor to a target and selects it. In spatial computing, there is no cursor in the traditional sense. On Vision Pro, the primary pointing device is the user's eye. The interface tracks where the user is looking and uses gaze as the selection mechanism, confirmed by a hand gesture.

This changes the fundamental grammar of interaction. Hover states, which carry enormous information weight in traditional interfaces, no longer work the same way. An element that highlights when the user looks at it must be carefully designed not to feel like it is responding to accidental glances. The precision expectations around target sizes change completely: Fitts's Law, which governs the relationship between target size, distance, and selection accuracy, requires significant recalibration for three-dimensional pointing tasks where depth perception adds an entirely new variable.

There are no edges

Flat-screen design is fundamentally about the edge of the screen. Navigation lives at the edges. Content is scrolled within them. The fold is a concept defined entirely by the screen boundary. In spatial computing, there are no edges. Content can exist anywhere in a 360-degree field around the user.

This sounds liberating. It is also deeply disorienting for users and designers alike. Without edges to orient content, the visual hierarchy that traditionally guides the eye through a layout has to be rebuilt from scratch using spatial positioning, depth, and lighting rather than proximity to the top-left corner of a rectangle. Content that is out of the user's current field of view is effectively invisible, which creates new navigation challenges that scroll bars and page indicators cannot solve.

Physical comfort is a design constraint

On a flat screen, the user's body is largely irrelevant to the design. They sit in front of it or hold it. On a spatial computing platform, the user's body is central. Content placed too high causes neck strain over time. Content placed too far to the periphery requires users to turn their heads repeatedly, which becomes fatiguing. Content that moves too fast causes disorientation that in extreme cases produces nausea.

Apple's visionOS design guidelines make this explicit: people should be able to use your app for extended periods without getting tired or uncomfortable. That requirement eliminates entire categories of motion design that work perfectly on flat screens. Parallax scrolling, rapid transitions, and content that moves against the user's head movement are all contraindicated in spatial contexts for physiological reasons, not aesthetic ones.

The environment is always there

On a flat screen, the design team controls everything the user sees within the interface. In a mixed reality context, the user's physical environment is always present. A dark interface element that reads beautifully against a standard background may become invisible against a cluttered desk. A layout that feels spacious in one room becomes cramped when the user is in a different space.

Apple's design response to this is the use of glass materials and transparency as primary interface aesthetics. Semi-transparent backgrounds allow interface elements to remain legible across varying real-world backgrounds while preserving the user's awareness of their physical surroundings. The design of visionOS 26, which unified Apple's visual language around a liquid glass aesthetic across all platforms, was explicitly designed with this spatial computing requirement in mind. The glass interface that feels elegant on an iPhone will, by design, look natural when floating in three-dimensional space around a Vision Pro user.

The New Interaction Vocabulary

Spatial computing introduces a set of interaction primitives that have no direct equivalent in flat-screen design. Understanding them is the starting point for designing interfaces that feel native to the medium rather than ported from a different one.

Gaze

On Vision Pro, gaze is the primary pointing device. The interface tracks where the user is looking with a precision that Apple describes as sub-degree accuracy. Designing for gaze means understanding that unlike a mouse cursor, which a user moves deliberately, the gaze moves constantly as part of natural visual behavior. Interface elements must not respond to incidental glances. The selection gesture, typically a pinch of the fingers, provides the confirmation that distinguishes intentional selection from casual looking.

The design implication is that gaze-sensitive elements need clear affordances that communicate their interactive nature without requiring hover confirmation. Color, depth position, subtle animation, and spatial audio all become tools for communicating interactability in the absence of a traditional hover state.

Indirect gesture

Apple's visionOS prioritizes indirect gestures: the user looks at a target and performs a gesture with their hands in their lap or at a comfortable natural position, rather than reaching out to physically touch a floating interface element. This is a deliberate departure from the reach-and-grab interaction model common in earlier VR and AR systems, which required users to extend their arms to interact with virtual objects. Sustained arm extension is physically fatiguing and impractical for productivity applications.

Designing for indirect gesture means that the spatial relationship between where the user looks and where they gesture is more important than the physical proximity of hands to content. A button that is three meters away can be selected as comfortably as one that is an arm's length away, as long as the gaze and gesture pairing is clearly established by the design.

Spatial audio

Sound in spatial computing is directional. Audio comes from the location of the object producing it. A notification from a window to the user's left arrives from the left. A music player behind the user sounds like it is behind them. This is not an audio feature. It is a navigation and information architecture feature.

Spatial audio can communicate the location of content that is outside the user's current field of view, guide attention toward relevant elements without visual interruption, and reinforce the sense that digital objects have physical presence in the space. Designing for spatial audio requires treating sound as a primary interface element rather than an enhancement, which is a significant shift for most UX teams whose audio design experience is limited to notification sounds and interaction feedback tones.

Windows, volumes, and immersive spaces

visionOS organizes applications into three distinct presentation modes. Windows are flat panels that float in the user's space, essentially traditional app interfaces that exist in three-dimensional space but contain two-dimensional content. Volumes are three-dimensional objects that the user can view from any angle. Immersive spaces replace the user's environment entirely with a digital one.

Choosing the right mode for each part of an application experience is a core spatial design decision. A productivity tool might use windows for its main interface and a volume to display a three-dimensional data visualization. A training application might use windows for instructional content and an immersive space for the practice environment itself. The transition between modes is a moment of significant UX consequence and needs to be designed as deliberately as any other key interaction in the flow.

Flat-Screen UX vs. Spatial UX: What Changes

Where Spatial UX Is Already Being Applied

The enterprise applications emerging in 2025 and 2026 are the clearest early signal of where spatial computing creates genuine value over flat-screen alternatives. These are not experiments. They are production deployments solving real problems.

Healthcare and surgical planning

Surgeons using spatial computing can examine three-dimensional patient anatomy data at actual scale, rotate structures, isolate systems, and plan procedures with a spatial understanding that flat medical imaging fundamentally cannot provide. The interaction model, gaze and gesture in a sterile environment, is also practical in a way that keyboard and mouse are not in an operating theater. Several hospital systems in the United States and Europe have moved from pilot to clinical use of Vision Pro for pre-surgical visualization.

Architecture and built environment

Walking through a building that does not yet exist is a fundamentally different evaluation experience than reviewing a floor plan or a rendered video. Architects and clients using spatial computing can assess scale, light, spatial relationship, and material in a way that resolves ambiguities that conventional visualization always leaves open. Forty percent of Global 2000 manufacturers already maintain production digital twins, many of which are being accessed through spatial computing platforms.

Remote collaboration and virtual presence

Vision Pro's Persona feature creates a real-time three-dimensional representation of the user that other Vision Pro users see in their environment during a call. The collaboration experience this creates is qualitatively different from video conferencing in ways that are immediately apparent to people who try it. Eye contact, spatial awareness of where participants are relative to shared content, and the sense of shared presence all change the dynamic of distributed work in measurable ways. As hardware becomes lighter and less expensive, the addressable market for this application category will grow substantially.

Training and skills transfer

Procedures that require spatial understanding, physical manipulation, or safety-critical practice benefit from spatial training environments that flat-screen simulations cannot replicate. Equipment maintenance, hazardous material handling, clinical procedures, and precision manufacturing are all categories where spatial computing training is moving from novelty to standard practice in organizations that have tried it.

What UX Teams Should Be Doing Right Now

Most UX teams do not have access to Vision Pro hardware and are not actively building spatial applications. That does not mean spatial computing is irrelevant to their practice. The skills, mental models, and design thinking that spatial UX requires can be developed before a specific project demands them.

• Understand visionOS design guidelines. Apple's Human Interface Guidelines for visionOS are publicly available and describe the design principles, interaction patterns, and technical constraints of the platform in detail. Reading them changes how you think about information hierarchy, motion, and interaction in ways that will improve your flat-screen work even if you never design a spatial application.
• Start thinking in three dimensions. Design tools like Gravity Sketch and Shapr3D allow teams to prototype spatial layouts directly in three-dimensional space. Figma's ecosystem of spatial plugins is growing. The mental shift from designing in two dimensions to designing in three is significant and takes time. Starting that process now, even without a specific spatial project, builds capability that will matter.
• Rethink what information hierarchy means without edges. Take a current project and ask: if this interface existed in three-dimensional space with no screen edges, how would users know what to look at first? How would navigation work? What would replacing the scroll interaction look like? These are thought experiments, but they surface assumptions in your current design thinking that are worth examining.
• Follow the enterprise deployments. The most practical spatial UX learning happening right now is in the documented deployments in healthcare, manufacturing, and architecture. Reading case studies from organizations that have moved spatial computing to production gives a clearer picture of where the genuine value lies and where the genuine friction points are than any speculative analysis can provide.
• Pay attention to what Apple is doing across its entire product line. The liquid glass design language introduced in visionOS 26 and unified across iPhone, iPad, and Mac is not just an aesthetic decision. It is Apple training the billion-plus users of its flat-screen products to read and expect spatial design conventions. By the time the next generation of spatial hardware reaches mainstream pricing, the visual and interaction language of spatial computing will already feel familiar to most users. That is a deliberate strategy, and it is worth understanding its implications.

The Honest Limits of Where We Are

Intellectual honesty about spatial computing requires acknowledging what is not yet ready alongside what is.

As of mid-2026, Vision Pro remains a premium product primarily serving developers, creative professionals, and enterprise early adopters. The weight, the battery life, and above all the price limit its reach. A more affordable Apple Vision model is expected, based on widely-reported leaks, and Meta continues iterating on lower-cost hardware with faster cycles. But mass market spatial computing is still a few hardware generations away.

The software ecosystem for visionOS remains thinner than that of iOS or macOS. The design patterns for spatial applications are still being established through trial and the sometimes uncomfortable experience of discovering that a convention that seemed obvious turns out not to work in three dimensions. The research base is real but young.

None of this means the preparation should wait. The iPhone launched in 2007. Teams that started building mobile design literacy in 2007 and 2008 were significantly better positioned when mobile became the dominant computing platform than teams that waited until 2012 to take it seriously. The window for getting ahead of the spatial computing curve is open. It will not stay open indefinitely.

The Frame Is Gone. The Work Is the Same.

Spatial computing does not change what good UX is. Users still need to understand what a product does, find what they are looking for, complete tasks without unnecessary friction, and trust that the product will behave predictably. Those goals do not change when the screen disappears.

What changes is the set of tools and conventions available to achieve them. Gaze replaces cursor. Depth replaces the fold. Spatial audio replaces the notification banner. A field of view replaces the viewport. The underlying discipline, understanding users, designing for their actual context, and testing against their real behavior, is exactly the same.

The UX teams that will design the best spatial experiences are not the ones with the most VR experience. They are the ones with the strongest foundation in why good design works, applied with the humility to learn that some of what they know only worked because of the rectangle they were designing inside.

The rectangle is leaving. The work continues.