Blog

The Big Picture

The technology behind self-driving cars is becoming standard. At a recent NVIDIA conference, they announced that 20+ major car brands (GM, Toyota, Mercedes, Volvo, Hyundai, BYD, NIO, and many more) are now using the same core self-driving hardware and software stack. NVIDIA expects $5 billion in automotive revenue in 2026.

This is great for speeding up development. But it creates a new problem: if everyone uses the same driving brain, how do you make your car feel different?

Not through sensors or computers. The only real difference will be how the car talks to, behaves with, and builds trust with the person inside.

That layer — let’s call it “Trust UX” — is still underdesigned in most cars.

We’re Past the Demo Phase

Self-driving cars are no longer judged by whether they can complete a test drive. The real question now is: Can they run a real service that people trust, operators can scale, and regulators can inspect?

The question is no longer only “can the vehicle drive?” It is:

• Can the system explain itself?
• Can it recover well from surprises?
• Can it keep people confident when something unexpected happens?

Trust UX Is Not Just Polish — It’s Operating Infrastructure

Trust UX is not a nicer voice or a prettier screen. It is the bridge between the AI, remote human operators, and people (riders, safety teams, regulators).

It covers:

• Clear state updates (what is the car doing?)
• Intent signals (what is it about to do?)
• Explanations (why did it do that?)
• Escalation logic (when does it ask a human for help?)

In the demo era, this was “nice to have.” In the real deployment era, it is essential.

At Large Scale, Bad UX Becomes an Operations Problem

Waymo passed 14 million trips in 2025. Aurora expects 200+ driverless trucks by end of 2026.

At that volume, a confusing stop is not just a bad user experience. It causes:

• More support calls
• Lower utilization (cars stuck waiting for help)
• Hesitant passengers

Poor trust design doesn’t stay inside the car. It spills into operations and costs real money.

Remote Assistance Is Now a Compliance Requirement

California law requires driverless passenger services to maintain a live passenger-operator communication link. Records must be kept for one year.

A 2026 investigation found that every major operator (Waymo, Tesla, Aurora, etc.) refused to disclose how often they use remote intervention. There is still no federal standard for how fast that connection must be or what qualifications remote operators need.

This means: the remote-support interface is not separate from compliance. It is part of compliance. How the system escalates, explains, and connects to a human is part of the safety story — and potentially part of the legal record after an incident.

In Self-Driving Cars, Your Brand Shows Up Through Behavior

In a normal car, brand comes from:

• Steering feel
• Engine sound
• Suspension tuning

In a self-driving car, those disappear. What remains is behavior — how the AI acts and communicates. And on a shared platform, behavior defaults to generic unless the carmaker defines it deliberately.

Four Trust Questions Every Self-Driving Car Must Answer

Question	What It Means
State legibility	What is the car doing right now? Does the passenger need to know?
Intent signaling	What is it about to do? How far ahead does it say so?
Uncertainty expression	When confidence drops, what does the system reveal?
Escalation character	When human help is needed, how does the handoff feel?

Generic platforms answer these once, for a generic user. Great brands answer them differently for their specific passengers.

Examples: How Different Brands Should Answer

BMW M (performance)

• State info is precise (speed, distance, time)
• Intent is early and specific: “Merging left in 400 meters — stopped vehicle ahead”
• Uncertainty is stated as fact: “Speed reduced — low visibility”
• Feels like a co-pilot, not a cautious assistant

Volvo (safety)

• System speaks before concern arrives: “Giving extra space to the cyclist ahead”
• Uncertainty becomes deliberate caution: “Taking this section slowly”
• Remote assistance feels like care, not failure

Jeep (capability off-road)

• Silence is the confidence signal
• Intent is decisive: “Taking this steep section at low speed”
• Uncertainty is assessment: “Checking the approach before proceeding”
• Never admits doubt — communicates deliberateness

Family cars (Toyota, Hyundai)

• Can a 10-year-old in the back seat understand what’s happening?
• Visual-first information
• Uses “we”: “We’re slowing down — there’s a red light ahead”
• Human-forward escalation — knowing a person is there builds trust

Waymo: The Best Example Today

Waymo calls itself “the world’s most trusted driver.” They publish independent safety data, communicate proactively inside the vehicle, and have built fleet response infrastructure that handles edge cases without alarming passengers.

But even Waymo has room to improve. When the car slows unexpectedly, passengers see a route screen but rarely understand why. Surfacing what the system saw and decided — in plain language, at that moment — would close the gap between “competent opacity” and “genuine transparency.”

At 500,000 trips per week, that trust investment compounds into repeat use and regulatory confidence.

The Business Case

• Brand loyalty in the US is only 51.1% — nearly half of buyers switch brands.
• 53% of US consumers expect to switch brands on their next vehicle.
• 52% would keep a car 2–3 years longer with meaningful over-the-air updates.

In self-driving cars, weak trust UX increases support costs, reduces repeat use, and makes it easier for customers to switch to a competitor — especially when the underlying technology is already shared.

The Bottom Line

The next big battle in self-driving cars is no longer just about driving intelligence. It is about trust industrialization — building the layer between autonomy, remote operations, and people with the same rigor that companies like Aurora and Waabi are applying to safety.

Shared technology platforms will not erase differentiation. But generic trust UX will.

The winners will make the self-driving experience:

• Easy to understand
• Easy to recover with
• Efficient for operators
• Credible to regulators
• True to their brand promise