Volvo's New Electric Car Can See and Talk to You

Volvo Cars has unveiled the EX60, a fully electric SUV that comes with Google's Gemini AI assistant built directly into the vehicle. Unlike voice assistants you might use on your phone or smart speaker, this AI runs right inside the car itself and can understand what the car's cameras see. Volvo and Google presented this integration at Google's I/O developer conference as the first of its kind in a production car.

The car was set to debut on January 21. What makes this different from most car AI systems today is that Gemini lives in the vehicle's own computer rather than being tethered to your phone or requiring constant internet connection to a data center.

How It Works: HuginCore as the Middleman

At the heart of the EX60's AI system is something called HuginCore, Volvo's proprietary platform. Think of HuginCore as a traffic controller that manages services from Google, NVIDIA, Qualcomm Technologies, and Volvo's own systems, making sure they all work together smoothly.

This multi-vendor approach is worth noting. Rather than relying on a single company for everything, Volvo maintains control over the overall architecture while borrowing best capabilities from different technology partners. This gives the automaker flexibility if things change in the market down the road—it's similar to how smartphone makers in the 2010s stopped depending on a single supplier for all their components and started sourcing from multiple vendors.

What the Cameras Can Do

The EX60 has cameras pointing in all directions around the vehicle. These cameras feed images directly into Gemini, allowing the AI to answer questions about what it sees around the car. This is a meaningful step beyond typical car assistants, which mainly respond to voice commands and access navigation or music functions.

In practical terms, you might ask the AI something like "What's that building on my left?" or "Is there room to park in that space?" The system processes the camera image and provides an answer. This requires the car to do some processing work locally before sending information to the cloud, because the response needs to come back quickly—a delay of even a few seconds would feel awkward in real driving situations.

Why This Matters Now

Car companies are racing to figure out how to fit artificial intelligence into vehicles, and they face constraints that don't exist in phones or home devices. Cars operate in extreme temperatures, drive through areas with weak internet signals, and must keep working even if the cloud connection drops.

Volvo chose a middle path. Tesla builds most of its AI capabilities itself. Traditional car companies often outsource everything to tech vendors. Volvo instead keeps architectural control through HuginCore while bringing in best-in-class technology from Google, NVIDIA, and Qualcomm. NVIDIA's involvement likely provides the processing power needed for analyzing camera images, while Qualcomm helps integrate everything with the car's electronic systems.

Google sees cars as an important new place to test its conversational AI. Smartphones and smart speakers aren't enough—the automotive market represents a different kind of opportunity, where Gemini can combine voice, text, and camera images to create a more natural interaction.

The Real Challenges Ahead

Building AI systems that work reliably in cars is harder than building them for phones. Vehicles deal with temperatures that swing wildly, spotty internet coverage, and strict safety and privacy regulations that don't apply to consumer electronics.

The biggest open question is whether drivers will actually want to use AI this way while driving. Voice assistants have never caught on as much in cars as they have on phones, partly because talking to a device while concentrating on the road adds mental effort rather than reducing it. The success of the EX60 will depend as much on how Volvo designs the interaction—what you say, how long the wait is, whether it feels useful—as it does on the underlying AI.

Another practical concern: processing 360-degree camera feeds constantly consumes power and generates a lot of data. Volvo's engineers had to balance AI capabilities against the EX60's driving range. An AI assistant that drains the battery faster would work against the whole point of owning an electric car.

What Comes Next

This is one of the first times we're seeing large language models like Gemini deployed in a car built for regular customers. It will likely serve as a reference point for how other automakers approach their own AI plans. Some may follow Volvo's multi-vendor model; others may build their own systems; others may stick with smartphone-based features for now.

How consumers respond to the EX60's capabilities will influence the entire industry. If people find it genuinely useful, expect to see similar AI systems in other vehicles fairly quickly. If the feature feels awkward or gimmicky, that will send a different signal about how much AI belongs in cars.