Physical Controls · Automotive HMI · Research-led Design

Bringing Physical
Controls Back to Cars

As automotive interiors shifted to all-screen interfaces, something was lost: the confidence of pressing a real button. This project made the case — with data — for where physical controls belong, what interactions they should use, and what form they should take.

Role
UX Designer
Collaborators
Interior Design · Ergonomics · User Research
Methods
Survey · EUNCAP · 3D Prototyping · User Testing
Output
Function Map · Interaction Guide · Ergo Spec · Validated Model
Context

The All-Screen Interior and Its Costs

Over the past decade, automotive interiors have undergone a dramatic consolidation: physical buttons replaced by touchscreens, knobs replaced by sliders, tactile certainty replaced by visual confirmation. The logic was clean — fewer parts, infinite flexibility, a premium aesthetic signal.

But the human cost accumulated quietly. Drivers took their eyes off the road longer to find controls. Error rates climbed. The elderly and users with motor differences struggled with targets that offered no resistance or feedback. Regulators began scrutinizing distraction metrics. The all-screen interior was solving a manufacturing problem while creating a usability one.

"Every button we removed was one fewer thing to look at. Every button we removed was also one fewer thing you could find without looking."

Phase 01
Function & Position Map

What Should Be Physical, and Where

Before designing anything, we needed a defensible answer to the hardest question: which functions deserve a physical control? This required moving beyond gut instinct and establishing a principled framework, then validating it against quantitative survey data, regulatory guidance, and vehicle positioning strategy.

The Four Principles

🛡
Safety Functions

Eyes-off, hands-on critical

Controls that must be operable without visual attention during a safety event. EUNCAP scoring directly penalizes glance-demand — physical controls are non-negotiable.

🔁
Habit & Frequency

High use or strong physical preference

Survey data revealed functions used multiple times per trip where users expressed strong physical preference: volume, climate, drive mode. Muscle memory is an efficiency asset.

Premium Highlight

Physicality as brand expression

Certain controls carry emotional weight when physical: a dedicated privacy mode button, a solid-feel window control. The tactile richness signals quality in ways pixels cannot.

📍
Spatial Contextual

Location is the interface

Controls whose physical position is itself intuitive feedback: roof curtain on the headliner, seat heat on the seat bolster, door lock on the door panel. Spatial logic reduces cognitive load.

Research Inputs

📊
Quantitative Survey
Usage frequency × physical preference mapped across 40+ functions, cross-referenced with glance duration data.
🏛
EUNCAP
EU New Car Assessment Programme criteria informed mandatory physical requirements for safety-critical functions.
🎯
Car Positioning
Premium segment positioning determined which functions merited tactile premium signaling vs. digital efficiency.

Cabin Zone Map

Physical Control Anchors by Zone
Headliner
Roof curtain · Reading lights · Sunshade
Door Panel
Window · Lock · Mirror · Seat heat
Center / Steering
Volume · Drive mode · Climate temp · Hazard · Privacy mode
Door Panel
Window · Lock · Mirror · Seat heat
Screen (Digital)
Media selection · Navigation · Settings · Comfort fine-tuning
Physical
Digital
Phase 02
Interaction Model

The Right Gesture for the Right Control

Deciding that a control should be physical is only the first question. The second is: what kind of physical interaction? Automotive controls must operate without precise motor behavior — drivers are glancing at the road, wearing gloves, operating one-handed, or reaching across zones under vibration.

Interaction Types Mapped

Interaction Suited For In-Car Suitability
Press / Tap Binary toggle, confirm, activate Primary — discrete, eyes-free, no precision needed
Rotate Continuous value (volume, temp, fan) Primary — infinite encoder; strong haptic feedback possible
Press + Hold Mode lock, confirmation of intent Secondary — useful for preventing accidental triggers
Slide / Rocker Directional incremental (seat position) Secondary — spatially intuitive, needs more surface area
Swipe / Gesture Scroll, navigate lists Avoid — precision-dependent; fails under vibration
Pinch / Multi-touch Zoom, resize Avoid — requires bimanual focus; incompatible with driving

Priority Guide

1

Discrete actuation — Press / Rotate

No precision required. Eyes-free operable. Clear tactile start and end state. Default choice for all physical controls.

2

Directional / incremental — Rocker / Slide

Use when spatial direction maps to function logic. Requires ergonomic clearance and reasonable surface area.

3

Held / modal — Press + Hold

Reserve for functions where preventing accidental activation matters more than speed. Never the only method for safety-critical functions.

Phase 03
Design

Collaborating with Interior Designers

In automotive UX, the physical form of a control is not ours to design in isolation. Interior designers own the surface, the material language, and the spatial composition. Our role was to be rigorous and generative collaborators — bringing UX requirements into their process rather than imposing finished specs.

Mood Board as Shared Language

Rather than delivering a list of requirements, we built a mood board to establish shared aesthetic and tactile direction — communicating texture vocabulary, feedback character, and premium references from adjacent industries like watchmaking, professional audio, and architectural hardware.

Machined metal · precision detent
🎛
Studio knobs
🏛
Arch hardware
🪵
Soft-touch · matte contrast

Ergonomic Requirements

We translated interaction model decisions into a concise ergo spec — clear constraints to design within while preserving interior design's creative latitude on form, material, and finish.

Phase 04
Prototype & Test

3D Printed Variants,
Real Users, Real Decisions

With interaction types defined and ergo constraints set, we moved into rapid physical prototyping. Rather than waiting for production-intent samples, we 3D-printed dirty prototypes of solution variants — deliberately rough, focused on testing the interaction logic rather than the aesthetic finish.

What We Prototyped

🎚
Variant A — Rotary Cluster

Three stepped encoders for volume, temp, fan. Grouped center console placement.

Variant B — Button Array

Flat-panel discrete buttons with clear tactile separation. High-contrast labeling.

Variant C — Hybrid Rocker

Rocker switches for directional functions, rotary for continuous values.

Testing Approach

We tested with a broad user group — specifically including elderly participants, as their needs expose clearance, force, and legibility failures that younger users compensate for. Sessions combined task completion (find and activate control without looking), preference rating, and qualitative interview.

Key Findings

  • Rotary encoders consistently preferred for continuous values

    Across age groups, rotary interaction was more satisfying and more error-tolerant for volume and temperature. The detent feedback gave confidence that the action had registered — especially important for elderly users who doubted whether a press had been felt.

  • Button arrays require larger clearance than spec'd

    Elderly participants and participants with larger hands frequently mis-pressed adjacent buttons in the 16mm variant. Testing drove a revised recommendation of 20mm minimum, 24mm for steering wheel clusters.

  • ⚠️

    Hybrid rocker caused mode confusion

    Variant C's mixed interaction vocabulary created uncertainty about which gesture to apply. Users paused, mis-operated, then expressed lower confidence — even when task completion was technically successful. Consistency of interaction type within a cluster matters more than optimal matching per function.

  • Tactile differentiation enables eyes-free use

    When control surfaces had distinct textures, participants identified controls without looking in over 85% of trials. Uniform surfaces dropped accuracy to 52%. Texture is not aesthetics — it is function.

  • 💡

    Elderly users surfaced what everyone benefits from

    The clarity, force, and spacing requirements that emerged from elderly testing made every variant better for all users. Inclusive design here wasn't about accommodation — it was about revealing the universal baseline that good haptic design should meet.

  • Test-Informed Decision

    Rotary-first interaction model. 20mm minimum clearance. Mandatory tactile surface differentiation per control type.

    Variant A's rotary cluster was selected as the primary interaction paradigm. Hybrid vocabularies were removed from the recommendation. Ergo spec was updated with revised clearance minimums and fed back into the interior design process as hard constraints for the next prototype round.

    Reflection

    What This Project Taught Me

    On cross-disciplinary collaboration

    The most valuable skill in this project wasn't knowing what good physical interaction feels like — it was knowing how to communicate that to interior designers in a language that respected their process. A requirements document delivered too late or too prescriptively becomes a constraint document. Delivered as a mood board and a set of functional principles, it becomes creative input.

    UX in hardware contexts is fundamentally a translation practice. You translate user behavior and safety data into the vocabulary of material, force, and form — and then you listen as that vocabulary comes back transformed by people who understand surfaces in ways you don't.

    On testing with the edges of the user population

    Including elderly participants wasn't an afterthought — it was a methodological decision that paid off disproportionately. Every failure mode they surfaced was a latent failure mode for the entire user population. The clearance revision, the force range, the mandatory tactile differentiation: none of those would have been caught with a standard adult sample.

    Design for your most constrained user and you will design something better for everyone — not because of altruism, but because constraints are the clearest signal of where a design is actually fragile.