Misty Cripps · Vaquita DesignDual-Screen Android Phone
Designing an interaction language for a form factor that didn’t have one.
I co-led the design for a dual-screen Android phone, including the system-level framework and the core application suite. Every convention assumed in single-screen mobile interaction had to be reconsidered. We understood that the underlying system could be very complex, but the user should never feel that complexity. When two equally good options existed, we ruled in the user’s favor.
How do you design for a form factor that doesn’t have any conventions yet?
A dual-screen phone doesn’t just double the screen real estate; it fundamentally changes the relationship between the user and their applications. Every assumption needs to be reconsidered: How do apps launch? Where do they appear? What happens when you rotate the device? How do you move content between screens? How do you avoid overwhelming users with complexity while still delivering on the promise of two screens?
Once those questions had system-level answers, the framework had to prove itself by actually working, which meant designing the applications that would live inside it. A framework that worked in the abstract but couldn’t carry a contacts list or a music app wouldn’t be a framework at all. It would be a deck.
A system-level framework, four application models and five applications.
The Window Positioning Model handled application placement across single and dual modes: repositioning, maximizing, minimizing, and stacking application cards. Four Application Models gave developers a clear framework for how any app could leverage dual screens. A custom gesture language introduced three new gestures alongside the standard Android ones, carefully limiting them to avoid cognitive overload. Orientation guidelines, built on a “gravity” model, kept rotation behavior simple and predictable. Global interaction patterns included text entry, copy/paste, drag-and-drop, and a universal clipboard designed for a two-screen context.
We designed five core applications (Phone, Contacts, Messaging, Photos, Music), each putting different application models to work. We documented the system with the same rigorous structure: design concepts and rationale, application maps showing every screen and state, wireframes for all orientations and screen configurations (portrait single, portrait dual, landscape single, landscape dual), and step-by-step task flows annotated with interaction details.
The places where “give the user more options” would have been a worse design.
The framework’s strength came from constraint. Each rejection is a place where exposing more capability and more configuration would have multiplied the user’s cognitive load without giving them anything they actually wanted.
System-level rules. Application-level proof.
A lightweight windowing system for mobile. Managed application placement across single (closed) and dual (open, side-by-side) modes, with rules for repositioning, maximizing, minimizing, and stacking application “cards.” A user who learned how an app behaved when closed shouldn’t have to relearn it when opened.
Parent-Child handled hierarchical navigation, list on left, progressive depth on right, and context preserved at every level. Parallel View showed simultaneous views of the same content, such as a map plus turn-by-turn directions side by side, which we identified as the device’s key differentiator. Exposé surfaced controls and shortcuts on one screen, freeing the other entirely for content. Expand treated landscape orientation as a special view for dual-screen apps, deliberately constrained to reinforce its uniqueness.
Three new gestures alongside standard Android. Each was mapped to specific behaviors in both single and dual-screen modes. The gesture set behaved one way when the device was closed and another (related but distinct) way when it was open, in a way the user didn’t need to think about. The set stayed small to avoid cognitive overload.
A “gravity” model governed what happened when the device rotated between portrait and landscape, with rules designed to keep the experience simple even as the physical relationship between the two screens changed. Rotation is one of the most disorienting moments in a multi-surface device, but the gravity model made it predictable.
Text entry, copy/paste, drag and drop, and a universal clipboard designed for a two-screen context. These patterns had to feel native regardless of which app was active or how the device was held. Moving content between screens is the gesture that distinguishes dual-screen work.
Phone used the Exposé model to transform the second screen into a contextual workspace during a call (quick notes, contact details, search). The app handled 15+ distinct call management flows across every orientation, tab-based navigation mapped to a Parent-Child hierarchy underneath. Contacts used Parent-Child to show a contact card alongside the contact list, with a Quick Communication Bar (call, text, email directly from any thumbnail) and an Activities stream that aggregates communication history per contact across apps. Messaging used the dual-screen layout for simultaneous conversation list and active thread visibility. Photos leveraged dual-screen real estate for browsing, editing, sharing, and organizing. Music leveraged the Exposé model for playback controls and library browsing side by side.
A framework that anticipated what later foldable and dual-screen devices would face.
Preserved
- Mental models from single-screen Android. Single-screen behavior was an understood model, not a fallback.
- Simplicity. The system complexity stayed invisible.
- Direct manipulation and user control. The gesture vocabulary was small, but expressive.
- Predictability across rotation, mode change, and app switching.
Now possible
- Parallel views of the same content side by side: map plus directions, contact list plus contact card, conversation list plus active thread.
- Contextual Exposé patterns that surfaced controls without burying content in menus.
- Parent-Child navigation that preserved context at every level of depth.
- A universal clipboard and drag-drop model.
- The framework anticipated many future challenges companies would later face with their own foldable devices. These patterns port to any multi-surface device: multi-monitor productivity tools, AR and mixed reality, and accessibility-driven interfaces.
Four patents resulted from this work.
We were named on four patents for this work, each addressing a specific dual-screen problem that didn’t have a published solution at the time.
The specs that detailed the framework.
Each document contains the rationale, application maps, every wireframe across every orientation and screen configuration, and step-by-step task flows. The case study above is a synthesis of this work. These are what the engineering team built from.