[ WORK / ARA ]
Gestoos Universitat de Barcelona German Autolabs

ARA — AI AVATAR &
INFOTAINMENT

An empathetic AI avatar and infotainment system for autonomous vehicles. The future of (not) driving — enjoy the ride.

ARA infotainment — curved in-car screen
[ HERO IMAGE — CURVED SCREEN ]
[ CLIENT ]
UNIVERSITAT DE BARCELONA
[ DISCIPLINE ]
UX DESIGN
[ YEAR ]
2021
[ CONTEXT ]
TEAM PROJECT
130+
SURVEYS COLLECTED
5
USER INTERVIEWS
2
EXPERT CONSULTATIONS
7
WIZARD-OF-OZ TEST PARTICIPANTS

The Problem

[ TRUST PARADOX ]

By 2030, 65% of consumers won't trust autonomous vehicles — while the demographic that would benefit most from them trusts them the least.

Aging populations are losing their independence as they become unable to drive safely. Existing in-car infotainment and (semi) autonomous interfaces fail to address their trust issues:

DON'T EXPLAIN DECISIONS
No visibility into what sensors detect or why actions are taken.
LACK EMPATHY
Robotic, cold interactions increase anxiety during critical moments.
COMPLEX INTERFACES
Touch-heavy designs exclude users with declining motor skills.
NOT FOR AGING USERS
Small text, low contrast and single-mode interaction exclude older adults.

Design Principles

[ INSIGHTS → SYNTHESIS ]
( 01 )

TRANSPARENCY GOES FIRST

Show users what the car sees and why it decides — real-time computer-vision overlays and plain-language explanations build trust.

( 02 )

EMPATHETIC COMMUNICATION

The AI acknowledges user emotions, especially in critical moments. A calm, reassuring tone reduces anxiety.

( 03 )

ACCESSIBLE BY DESIGN

Multimodal interaction — voice, gesture, touch, haptic — lets aging users with varying abilities interact naturally, without training.

The Process

[ RESEARCHING INTO THE FUTURE ]

Research targeted Gen X and Baby Boomers (41–75) alongside Gen Z and Millennials (25–40) who drive or ride in cars — to understand fear and distrust of autonomous mobility, perceived loss of control, and other organisations' role in accelerating a self-driving future.

SAFETY

· EMERGENCY PROTOCOLS
· HAZARD DETECTION
· HEALTH MONITORING
· ADAPTIVE DRIVING

TRUST & TRANSPARENCY

· COMPUTER VISION OVERLAY
· DECISION EXPLANATIONS
· AR HUD INTEGRATION
· SYSTEM STATUS INDICATORS

INTERACTION

· MULTIMODAL AI (VOICE + GESTURE)
· TOUCH CONTROLS
· PERSONALISATION
· ENTERTAINMENT SYSTEM

Features were validated for the MVP through brainwriting sessions and card sorting with 10 users.

[ LOW-FIDELITY TESTING ]

WIZARD OF OZ METHOD

In collaboration with Gestoos, we integrated real gesture-recognition tech: a pointing gesture triggers simulated information projected onto the windscreen. 7 participants tested multimodal interactions on a filmed Barcelona route with critical incidents — rain, an ambulance, pedestrians — while a teammate played the role of the AI assistant.

Final Design

[ THE FUTURE OF (NOT) DRIVING ]
01

MULTIMODAL AI ASSISTANT

Voice (German Autolabs) + gesture (Gestoos) + touch + haptic feedback for natural, accessible interaction.

02

REAL-TIME TRANSPARENCY

Computer-vision overlay shows pedestrians, vehicles, traffic signals and hazards as the car sees them.

03

EMPATHETIC AI PERSONALITY

Pixar-inspired animated avatar with expressions, calm emergency handling and health monitoring.

[ HI-FI WALKTHROUGH ]
[ UI DESIGN LANGUAGE ]

Extra

[ DIGITAL MARKETING CAMPAIGN ]

As part of the deliverable, we created a digital marketing video for the product.

お仕舞い・・・

THANK YOU FOR
REACHING THIS FAR.

[ ALL WORKS → ]