comma.ai Openpilot Dashboard Design
Openpilot Dashboard Design - Dark Theme Left Driver

Openpilot Dashboard Design

Redesigning confidence communication and system awareness for safer autonomous driving. A clean, audio-first approach that prioritizes driver safety over visual spectacle.

Innovation: Driver safety and accessibility-first design


comma.ai Design Challenge Submission

Project Overview

Addressing critical UX gaps in openpilot's confidence communication and system limit awareness through automotive-grade interface design that reimagines the user-facing layer while leveraging openpilot's robust technical foundation.

display_settings 2160x1080 OLED Display
Qt Framework Constraints
schedule Weekend Sprint
(Design Challenge)
badge UX Designer, UI Designer
Automotive Safety Focus
car_crash Windshield-mounted device
Driver assistance context
accessibility Colorblind-friendly design
Audio-first approach
security Safety-critical interface
Government regulation compliance

The Challenge

Current openpilot users face critical gaps in understanding system status and limitations:

visibility_off

Confidence Opacity

Users can't easily distinguish between "reliable" and "about to fail" system states.

warning

Reactive Alerts Only

System limits hit without warning - alerts come too late to be useful.

tv

Visual Clutter from Live Road Footage

Existing solutions overlay complex UI on camera feeds, creating visual noise.


Research: Understanding the Problem Space

Before designing solutions, I needed to understand the current state of automotive interface safety and user comprehension. This research directly shaped the interface requirements and design constraints.

Core Research Question

What are the current failure modes in automotive interface design that lead to safety issues and user mistrust, and how can interface design address these specific problems?

1
Industry-Wide "Clear When Safe to Use" Gap

Consumer Reports' 2020 evaluation of 18 driver assistance systems found that "Clear When Safe to Use" represents a critical industry gap. Even top-performing systems like Comma Two's OpenPilot (ranked #1 overall) scored only 6/10 in communicating safe usage conditions to drivers.

Consumer Reports 2020 evaluation • Comma Two ranked #1 overall (78/100)

Source: Consumer Reports Active Driving Assistance Systems Evaluation, November 2020

2
Driver Distraction is a Growing Crisis

Current automotive interfaces are contributing to a worsening distraction problem. Device manipulation refers to drivers observed touching/operating handheld electronic devices while driving. The data shows both fatalities and risky behaviors are increasing despite technological advances.

Source: NHTSA Fatality Analysis Reporting System (FARS), National Occupant Protection Use Survey (NOPUS)

3
Users Don't Understand ADAS
33%
Don't understand system limitations
85%
Satisfaction when systems provide clear feedback

Research shows a critical gap: users who don't understand their ADAS make dangerous assumptions, but users who receive clear feedback report high satisfaction and appropriate usage.

Source: AAA Foundation for Traffic Safety, Consumer Reports Owner Satisfaction Survey

4
Color-Only Communication Excludes Users
~4%
of adults have color vision deficiency
(~8% of men, ~0.5% of women)

Traditional red/yellow/green status indicators exclude a significant portion of drivers. In safety-critical systems, this creates unequal access to crucial information.

Accessibility Requirement

Safety-critical information must be communicated through multiple channels: shape, pattern, audio cues, and high contrast - not color alone.

Source: National Eye Institute, Colour Blind Awareness Organization

Research Synthesis: Design Requirements

The research revealed four specific problems that guided the interface design approach:

Problem: Industry Communication Gap

Solution: Transparent system state communication that shows exactly when the system is safe to use and when it's approaching limits.

Problem: Visual Distraction

Solution: Clean, minimal interface separate from road view. No live footage overlays. Optimize for 2-second glances.

Problem: System Opacity

Solution: Precise percentage displays showing exact system state. Users can anticipate limits rather than react to failures.

Problem: Accessibility Barriers

Solution: Audio-first design with shape/pattern cues. Multiple ways to convey the same critical information.

Insights from Automotive UX Professionals

My design approach was informed by insights from industry professionals working in safety-critical interfaces.

Key Findings

Safety Distinguishes Automotive UX from Mobile UX

Unlike mobile interfaces where engagement and retention drive design decisions, automotive interfaces must prioritize immediate comprehension and driver safety above all else.

Reduce Driver Distraction Through Design

  • High contrast - Essential for glanceability in various lighting conditions
  • Large chunky buttons - Industry professionals call this "Fisher Price look and feel" but it's necessary for safe operation
  • Screens understood at a glance - Complex information hierarchies create dangerous cognitive load

Transparency Builds Trust in Autonomous Systems

For self-driving functionality specifically, "functionality without transparency fails to build consumer trust." Users need to understand system capabilities and limitations to make informed decisions about when to trust the automation.

"[The] concern for safety is what distinguishes automotive UX from mobile UX. To this end, screens should be understood at a glance."
— Hyundai Motor Group UX Team

This philosophical strategy shaped every design decision, from the choice of percentage-based displays to the emphasis on audio-first alerts and accessibility considerations.

Design Decision: Transparency Over Simplification

My Core Hypothesis:

Drivers don't want to see video of the road they're already looking at. They want to know if the system is working. Just like glancing down at their speedometer, they should be able to glance over and instantly know system status.

While the current openpilot interface serves developers and provides valuable reference footage, the driving context requires a fundamentally different approach. Users shouldn't be looking at video of the road they're already seeing - they need instant system status comprehension. This design leverages openpilot's robust backend while reimagining the user-facing layer for safety-critical operation.

✈️ Aviation Precedent

Aircraft autopilot systems don't show pilots video of the sky they're already seeing. They display critical data: altitude, heading, airspeed, system status. Data drives decisions, not decoration.

⚡ Split-Second Reality

Driving requires split-second decisions. Live footage is useful for demos and post-drive analysis, but during operation, drivers need instant status comprehension.

The Result: Trust Through Transparency

By showing exactly what the system knows (78% steering torque, 92% confidence) rather than abstract representations, users can make informed decisions about when to trust the system and when to take control.

Transparency builds trust. Trust enables adoption.

The interface serves the driver, not the other way around.

Design Philosophy

accessibility

Universal Accessibility

Colorblind-friendly design using shape, pattern, and audio cues

shield

Safety Over Spectacle

No live footage overlay - drivers should look at the road, not the screen

volume_up

Audio-First Design

Critical information communicated through sound when eyes are on the road

trending_up

Proactive Transparency

Show system limits before they become emergency alerts

verified

Automotive Compliance

High contrast, large targets, minimal animations per government regulations

security

Trust Through Clarity

Clear confidence states build user trust in autonomous systems

A critical design choice emerged when determining how to display confidence levels and system limits. We considered simplified approaches:

Considered: 3-Level System

Good / Caution / Danger

Considered: 5-Level System

Excellent / Good / Fair / Poor / Critical

✓ Chosen: Percentage-Based Transparency

Why percentages? Users deserve to see exactly what the system sees - 78% steering torque usage vs. "high." This builds trust through transparency rather than hiding complexity behind abstracted categories.

This decision aligns directly with Hyundai's insight: "Functionality without transparency fails to build consumer trust." For Level 2 systems where drivers remain ultimately responsible, complete visibility into system state is essential for informed decision-making.

Information Hierarchy

Since openpilot is Level 2 (partial automation) where "the driver must remain attentive and ready to take over at any time," the layout prioritizes critical information based on natural eye movement:

Left-Hand Drive
Left-hand drive interface layout

Driver sees confidence state first

Right-Hand Drive
Right-hand drive interface layout

Layout adapts for right-side drivers

  • Left Side (60-70% of screen): Primary confidence state - first thing drivers see
  • Right Side (stacked): System limits - important but secondary information
  • Accessibility consideration: Option for right-hand drive markets to flip layout

Design Solution

Final Thoughts: Automotive UX Beyond Mobile Patterns

This design challenge reinforced a critical insight: automotive UX requires fundamentally different thinking than mobile or web design. When lives are at stake, accessibility isn't optional, transparency builds trust, and driver safety must trump visual spectacle.

Automotive UX Beyond Mobile Patterns

  • Trust through transparency: Percentage-based displays show exactly what the system knows
  • Universal accessibility: Works for all users regardless of visual capabilities
  • Safety-first hierarchy: Critical information positioned for natural eye movement
  • Proactive awareness: Early warnings prevent emergency situations
  • Regulatory compliance: Designed within automotive UI constraints from day one

As autonomous driving technology evolves, the interfaces that communicate system state become increasingly critical to user safety and adoption. This design provides a foundation for that evolution.

Project Context & Next Steps

This design challenge focused on strategic problem-solving and concept development within weekend time constraints. The emphasis was on demonstrating automotive UX expertise, accessibility considerations, and systematic design thinking.

In a Real-World Production Environment:

  • User Testing: Comprehensive validation with actual openpilot users in controlled driving environments
  • Accessibility Validation: Testing with colorblind users and drivers with varying visual capabilities
  • Driving Simulator Testing: Validation of information hierarchy and reaction times during simulated driving scenarios
  • Safety Metrics: Analysis of driver behavior, takeover times, and system trust indicators
  • Iterative Refinement: Data-driven design improvements based on real-world usage patterns and safety outcomes

The foundation established in this challenge - transparency, accessibility, and safety-first design - provides a solid framework for evidence-based iteration and improvement in a production development cycle.

Submitted to comma.ai Design Challenge • Focus on solving real problems, not creating portfolio pieces

Previous
Previous

FurSure Case Study

Next
Next

CorrectPay UX Redesign