Google showed me the future of Android Auto — and now I dread my own car

At a glance:

• Immersive 3‑D navigation in Google Maps gives a clearer view of buildings, tunnels and overpasses.
• Gemini 3.5 AI is now callable from Android Auto, allowing voice‑only vehicle adjustments and multimodal queries.
• Prompt‑generated widgets let users stack custom weather, traffic or delivery info directly on the dashboard.

What the demo revealed

During the Android Auto showcase at Google I/O, attendees got to sit inside a Volvo EX60 equipped with Android Automotive. The new interface leans heavily on Material 3 Expressive design: rounded corners, adaptive colour theming and a three‑panel layout that blends smart‑home controls, navigation and media into a single, tablet‑like experience. A YouTube video player automatically collapses to audio‑only when the car is in motion, keeping eyes on the road while still delivering entertainment.

Immersive navigation redesign

Google Maps now renders a three‑dimensional view of the surrounding terrain. Buildings, stadiums, hills and even tunnel entrances are drawn with depth cues, helping drivers anticipate lane changes long before they appear on a flat map. The demo highlighted New York City’s maze of overpasses, where the 3‑D view made it easier to spot the correct lane for upcoming exits. This visual upgrade is expected to reduce driver stress in complex urban environments.

AI‑driven Gemini integration

The biggest headline was Gemini 3.5’s integration into Android Auto. In cars that support Android Automotive—currently models like the Volvo EX60—drivers can ask Gemini to perform vehicle‑specific actions such as dimming the sunroof, changing ambient lighting, or ordering food via DoorDash and tracking the delivery on their phone. Gemini can also tap into the car’s surrounding cameras for multimodal queries; the demo correctly identified the Transamerica Pyramid ahead and offered a short historical note. Voice prompts have been refined, delivering directions like “turn left at the intersection” instead of distance‑based cues, which research shows are easier to follow while driving.

Custom prompt‑generated widgets

Google demonstrated that developers (or even end‑users) can create widgets on the fly using natural‑language prompts. In the demo, a Google employee generated a weather widget that instantly displayed whether conditions were suitable for running or biking. Users can stack multiple widgets—traffic, bakery suggestions, delivery status—creating a personalised dashboard that adapts to each commute.

Design polish and performance tweaks

Beyond the AI and navigation upgrades, the UI now feels more cohesive. Rounded edges and colour theming make the screen resemble an Android tablet rather than a collage of disparate apps. Early testers report smoother performance, especially after applying four simple tweaks that Google shared in a separate guide (not reproduced here). The overall feel is a tighter, more responsive system that should feel familiar to long‑time Android users.

Rollout timeline and device compatibility

Google said the next Android Auto update, which bundles most of the announced features, will roll out in waves later this year. Because the Gemini functionality depends on the car’s native Android Automotive integration, rollout speed will vary by manufacturer. The author notes that his own Mazda is unlikely to receive the update soon, hinting that brands without deep Android Automotive support may lag behind early adopters like Volvo.

What to watch next

The demo suggests a future where the in‑car experience mirrors a full Android ecosystem: AI assistants handle routine tasks, custom widgets replace static shortcuts, and immersive maps become the default. Analysts will be watching how quickly OEMs adopt Android Automotive, how Gemini’s multimodal capabilities evolve, and whether drivers embrace AI‑driven voice interactions without distraction. For now, the preview offers a tantalising glimpse of a smarter, more connected cockpit.