Revolution
Problem
Self-driving cars are nearly here — but people don't trust them. Drivers fear how an AV will handle edge cases, and they aren't ready to give up control of the wheel.
Solution
A dealership test drive that builds acceptance for Level 4 AVs. Through a personalized setup and a curated driving simulation, Revolution confronts each fear directly — showing, risk-free, exactly how the AV responds.
Revolution
Spring 2021 (10 weeks)
Information Architecture
Revolution
Spring 2021 (10 weeks)
Information Architecture
Problem
Self-driving cars are nearly here — but people don't trust them. Drivers fear how an AV will handle edge cases, and they aren't ready to give up control of the wheel.
Solution
A dealership test drive that builds acceptance for Level 4 AVs. Through a personalized setup and a curated driving simulation, Revolution confronts each fear directly — showing, risk-free, exactly how the AV responds.
Project Manager
I led the project by assigning work, checking in with teammates, and determining the direction of the project.
UX Strategist
I mapped out the user journeys and defined product roadmaps based on business and user needs.
Interaction Designer
I designed how the user interacts with the products through input and output.
UX Engineer
I wired, tested, and developed physical and digital prototypes to test hypotheses and enable rapid iteration.
User Tester
I designed and executed user tests for rapid feedback and iteration on hypotheses.
Silver
Product or Gadget Design
Indigo
Silver
Technology
Indigo
Bronze
UX, Interface & Navigation
Indigo
Self-driving cars are nearly here. The people who would ride in them are not.
The stakes
Why autonomous vehicles matter — and why people still won't trust them.
Every year in the U.S.
37,000
people die in car crashes
Behind
94%
of serious crashes is human error
Still,
3 in 4
Americans say AVs aren't ready
And
48%
would never ride in a driverless car
A safer, self-driving future is almost here — but the people who would ride in it aren't ready. Across our research, one fear surfaced again and again: not the technology failing in the abstract, but how the car would handle the messy, real moments — an ambulance racing up behind you, a lane closed for construction, an animal in the road.
Six weeks listening to how people really feel about giving up the wheel.
How we listened
A mix of first-hand experience, depth interviews, and a public probe.
Tesla test drive
First-hand time in a Level 3 vehicle to feel how the AV experience is actually delivered to driver and passenger.
10 interviews
Seven users and three experts — including PMs and engineers from Cruise and Zoox — on mental models and comfort with emotion recognition.
Cultural probe · n=32
Two hours at Forsyth Park asking strangers what they really think about self-driving cars.
What we heard
Voices from Forsyth Park
First, unfiltered reactions to “What comes to mind when you think of self-driving cars?” On average, people rated their comfort with in-car emotion recognition a 5.8 out of 10 — wary, but open.
One sharpened question, two competing concepts.
How might we foster a sense of control and inspire confidence in an AV's ability to handle any scenario it might encounter?
We split that into three sub-questions: how to overcome conditioned fear, how to communicate the AV's actions without overwhelming people, and how to use emotion-recognition technology consensually — so it never feels creepy.
Two concepts
We developed two directions far enough to compare them head-to-head.
Concept 1 — The Test Drive: an in-showroom simulator that surrounds you with LCD windshields and runs curated driving scenarios. No risk of crashing; you leave with a personalized preference card.
Concept 2 — The AV Ecosystem: a seamless in- and out-of-car experience. Summon the car with a key fob, log in by fingerprint, and let emotion recognition personalize the ride over time.
Designing the in-vehicle interface and the simulation that frames it.
The experience runs on a triple display: a behind-the-wheel instrument cluster, a center IVI (in-vehicle infotainment) screen, and a steering display — all wrapped in a driving simulation projected across the windshield and side windows.
From lo-fi to mid-fi
Setup, home, navigation, and edge-case alerts — wireframed, tested, and refined.
Status bar
Battery, network, time, and temperature — the always-on essentials, card-sorted onto both displays.
Emotion-recognition sensors
Eye gaze, facial expression, and heart rate, each with a clear on/off state so consent is always visible.
Map + lane view
Swappable views — a bigger map when you're driving, the AV's lane view when it is.
AV driving profile
Speed-limit and space-cushion controls surfaced on the home screen, not buried in settings.
Difficult-situation alerts
Construction zones and emergency vehicles get a dedicated, hard-to-miss alert region.
Tactile control panel
Climate, wipers, blinkers, and lighting moved to physical buttons users can feel without looking.
Turning the concept into something you can actually drive.
A drivable simulation in Unity
To make the test drive feel real, I built the simulation in Unity. A custom path-follower drove the AV along Bézier routes through a city environment, while three cameras fed three displays for a seamless wrap-around view.
A speedometer that actually works
I wrote a working speedometer in C#, tied to the vehicle's rigidbody velocity and mirrored to a fourth display over the local network. Custom construction and ambulance scenarios let users watch the AV detect, decide, and respond in real time.
Inside the rig
The Unity build, the three-camera setup, and the physical simulator we drove.
Two rounds of Wizard-of-Oz testing, plus open and closed card sorts.
What testing changed
Revolution, from the showroom to the open road.
Revolution, end to end
A personalized preference card
The onboarding flow ends in a card that carries every setting into ownership — so the showroom experience follows you home.
Welcome and preference card
Emotional-recognition setup
Setup complete
Home, navigation, and alerts
The live IVI: compose your home screen, route through Savannah, and handle edge cases as they happen.
Default home screen
Routing to a destination
Emergency-vehicle alert
The instrument cluster
Speed, AV mode, navigation, and telltales — refined directly from the card-sort results.
Default cluster
Speed-limit warning
Construction alert